Biodegradable polymers in dental tissue engineering and regeneration

Raffaele Conte; Anna Di Salle; Francesco Riccitiello; Orsolina Petillo; Gianfranco Peluso; Anna Calarco; Raffaele Conte; Anna Di Salle; Francesco Riccitiello; Orsolina Petillo; Gianfranco Peluso; Anna Calarco

doi:10.3934/matersci.2018.6.1073

AIMS Materials Science

2018, Volume 5, Issue 6: 1073-1101. doi: 10.3934/matersci.2018.6.1073

Previous Article Next Article

Review Topical Sections

Biodegradable polymers in dental tissue engineering and regeneration

1.
Institute of Agro-Environmental and Forest Biology, CNR, Naples, Italy
2.
Conservative Odontostomatology and Maxillofacial Surgery, University of Naples, Naples, Italy
3.
Institute of Bioscience and Bioresources, CNR, Naples, Italy

Received: 13 August 2018 Accepted: 18 October 2018 Published: 07 November 2018

Over the last decade, biodegradable polymers (BP) replaced traditional non-degradable materials in dental and maxillo-facial surgery, in particular for bone regeneration and periodontal care, due to their ability to break down and be absorbed by the body without producing harmful degradation products, along with their great potential for controlled drug delivery, wound management, dental restorations, and tissue engineering. This review described the physical–chemical characteristics of these materials, classified them based on their sensitivity on hydrolytic or enzyme degradation, and described their recent application in both restorative and regenerative dentistry.

Keywords:

Citation: Raffaele Conte, Anna Di Salle, Francesco Riccitiello, Orsolina Petillo, Gianfranco Peluso, Anna Calarco. Biodegradable polymers in dental tissue engineering and regeneration[J]. AIMS Materials Science, 2018, 5(6): 1073-1101. doi: 10.3934/matersci.2018.6.1073

Related Papers:

[1]	Lixiang Zhang, Yian Zhu, Jie Ren, Wei Lu, Ye Yao . A method for detecting abnormal behavior of ships based on multi-dimensional density distance and an abnormal isolation mechanism. Mathematical Biosciences and Engineering, 2023, 20(8): 13921-13946. doi: 10.3934/mbe.2023620
[2]	Mubashir Ahmad, Saira, Omar Alfandi, Asad Masood Khattak, Syed Furqan Qadri, Iftikhar Ahmed Saeed, Salabat Khan, Bashir Hayat, Arshad Ahmad . Facial expression recognition using lightweight deep learning modeling. Mathematical Biosciences and Engineering, 2023, 20(5): 8208-8225. doi: 10.3934/mbe.2023357
[3]	Hongwei Sun, Qian Gao, Guiming Zhu, Chunlei Han, Haosen Yan, Tong Wang . Identification of influential observations in high-dimensional survival data through robust penalized Cox regression based on trimming. Mathematical Biosciences and Engineering, 2023, 20(3): 5352-5378. doi: 10.3934/mbe.2023248
[4]	Chengjun Wang, Xingyu Yao, Fan Ding, Zhipeng Yu . A trajectory planning method for a casting sorting robotic arm based on a nature-inspired Genghis Khan shark optimized algorithm. Mathematical Biosciences and Engineering, 2024, 21(2): 3364-3390. doi: 10.3934/mbe.2024149
[5]	Wenyang Gan, Lixia Su, Zhenzhong Chu . A PSO-enhanced Gauss pseudospectral method to solve trajectory planning for autonomous underwater vehicles. Mathematical Biosciences and Engineering, 2023, 20(7): 11713-11731. doi: 10.3934/mbe.2023521
[6]	Juan Du, Jie Hou, Heyang Wang, Zhi Chen . Application of an improved whale optimization algorithm in time-optimal trajectory planning for manipulators. Mathematical Biosciences and Engineering, 2023, 20(9): 16304-16329. doi: 10.3934/mbe.2023728
[7]	Kezhou Chen, Xu Lu, Rongjun Chen, Jun Liu . Wireless wearable biosensor smart physiological monitoring system for risk avoidance and rescue. Mathematical Biosciences and Engineering, 2022, 19(2): 1496-1514. doi: 10.3934/mbe.2022069
[8]	Zhiqiang Wang, Jinzhu Peng, Shuai Ding . A Bio-inspired trajectory planning method for robotic manipulators based on improved bacteria foraging optimization algorithm and tau theory. Mathematical Biosciences and Engineering, 2022, 19(1): 643-662. doi: 10.3934/mbe.2022029
[9]	Yalong Yang, Zhen Niu, Liangliang Su, Wenjing Xu, Yuanhang Wang . Multi-scale feature fusion for pavement crack detection based on Transformer. Mathematical Biosciences and Engineering, 2023, 20(8): 14920-14937. doi: 10.3934/mbe.2023668
[10]	Rui Zhang, Shuicai Wu, Weiwei Wu, Hongjian Gao, Zhuhuang Zhou . Computer-assisted needle trajectory planning and mathematical modeling for liver tumor thermal ablation: A review. Mathematical Biosciences and Engineering, 2019, 16(5): 4846-4872. doi: 10.3934/mbe.2019244

Abstract

1. Introduction

With the popularization of various mobile terminals with high-precision positioning function, such as automobile navigation, mobile phones and wearable devices, a large number of trajectory data arose. These trajectory data contain abundant spatiotemporal and semantic information, which have been proved to be very valuable resources and have a wide range of application scenarios ^[1]. Taxi is a very important urban transportation tool. All taxis in operation frequently report their location information to the data center through installed GPS or Beidou positioning equipment. These massive trajectories data are widely distributed in the urban road network, which can well reflect the traffic conditions of the urban road network. Therefore, taxis are also known as the "flow detector" of urban traffic. The taxi trajectory data set is also widely used in urban planning, road recommendation, traffic hotspot analysis, traffic accident analysis and other fields ^[2].

The purpose of trajectory outlier detection is to identify abnormal phenomena from a large number of trajectory data, which is helpful to discover or predict potential traffic risks ^[3]. The trajectory outlier is generally divided into two categories: 1) The trajectory deviates from other trajectories in space; 2) The time sequence of trajectory points is significantly different from other trajectories. Trajectory outlier means that some behaviors deviate from expectations, usually accompanied by special or interesting events. Taking urban traffic monitoring and management as an example, in the saturated urban road network, local trajectory outliers may reflect the occurrence of traffic accidents, bad weather or road emergencies. Detection and real-time analysis of these trajectory outliers can not only provide the traffic management department with the basis for traffic situation analysis, timely find and predict the congested sections, but also provide the driver with reference for route selection.

Traditional trajectory outlier detection methods mainly focus on trajectory similarity and clustering analysis ^[4]. These methods calculate the similarity between trajectories by defining the distance metric, and then divide the trajectories into several clusters by clustering. Finally, the cluster with a small number of trajectories is identified as outliers by threshold setting. However, the efficiency of such methods is usually not high, and it cannot well meet the needs of real-time analysis of large-scale trajectory data. In the field of urban transportation, the trajectory data has the characteristics of uncertainty, sparsity, skewed distribution and continuous updating, which makes it more difficult to detect outliers in real-time ^[5]. With the rise of deep learning methods, deep neural networks have gradually been widely used in trajectory data analysis and mining ^[6]. Aiming at the problem of online real-time detection of trajectory outliers, we proposed a trajectory outlier detection model based on variational auto-encoder in this paper. The model encodes the trajectory data as parameters of distribution functions and an auto-encoder network is built and trained. Once the model training is completed, we can detect the trajectory outlier by the difference between a trajectory and the trajectory generated by the model. The main contributions include:

1) A deep learning networks based on variational auto-encoder is used to learn the trajectory characteristics and distribution. Once the model training is completed, we can detect the trajectory outlier by the difference between a trajectory and the trajectory generated by the model. Compared with the traditional density-based or classification-based trajectory outlier detection method, the calculation amount is greatly reduced which makes the model very suitable for real-time detection scenarios. Compared to other methods using variational auto-encoder, this paper performs segmentation and differential processing on trajectory data for input of the model, which make it unnecessary to eliminate noise points in trajectories and can further improve the accuracy of the model.

2) When comparing an original trajectory with the trajectory generated by the model, the difference between them is calculated from two aspects of Euclidean distance and cosine similarity, which further improve the detection effect of outliers.

3) When detecting the trajectory outliers, the distance threshold between the outliers and the normal trajectories can be set by referring to the proportion of the abnormal trajectory in the training data set, which eliminates the difficulty of setting the threshold manually and makes the model more convenient to be applied in different actual scenes.

In the remainder of this paper, we introduce related work in Section 2. After that, we give some definitions formally in Section 3. Section 4 introduces our model and methods for outlier trajectory detection. An experimental evaluation and analysis of the effectiveness and efficiency of our methods is presented in Section 5. Finally, we conclude the work and briefly discusses future work in Section 6.

2. Related work

Trajectory outlier detection methods mainly include density-based methods ^[7,8,9], classification-based methods ^[10,11,12], grid-based methods ^[13,14,15] and deep learning-based methods ^[16].

The density-based detection method uses algorithms such as k-means ^[17], DBSCAN ^[18] or improved DBSCAN ^[19] to cluster the segmented trajectories, and the sparse trajectories are determined as outliers. The commonly used definitions of distance between trajectories in these algorithms include European distance, DTW distance ^[20], LCSS distance ^[21] and Hausdorff distance ^[22]. Trajectory similarity calculation is the core of this kind of method. Recently, D. Zhang et al. studied this problem and proposed a method suitable for continuous calculation of trajectory similarity, and tried to apply this method to trajectory outlier detection ^[23]. In practical application, threshold selection is one of the difficult problems in this kind of methods. In addition, the large amount of computation makes these methods cannot meet the needs of real-time detection for large-scale trajectory data.

The idea of classification-based detection method is to establish a classifier model and train the model through the labeled trajectory data set so as to obtain normal trajectory features. Once the model training is completed, the trajectory outlier detection efficiency is very high. However, the disadvantage of this method is that it requires label data, and data annotation usually leads to a large amount of labor and time costs.

The main idea of the grid-based detection method is to map the trajectory data to the urban map grid after preprocessing, and then transform the detection of trajectory outliers into the detection of abnormal grid cell symbol sequence. This method is mainly applied to the detection of trajectory outliers in urban fixed road network environment.

In recent years, deep learning ^[24] developed rapidly and has attracted the attention of a large number of researchers. It has been widely used in almost all research fields related to feature extraction. This method has also attracted more and more researchers' attention in the field of trajectory data mining. Compared with the traditional method, the deep neural network can automatically learn the feature of the trajectory from the data set, and can achieve better results when the data set is sufficient. Document ^[25] proposes a pedestrian trajectory prediction method with dual attention fusion mechanism, which can also be used as an abnormal event detection method. In view of the problem of trajectory outlier detection caused by taxi fraud or detour, A. Belhadi et al. proposed a novel hybrid framework and two phase-based algorithms to identify trajectory outliers ^[26].

Auto-encoder (AE) ^[27] is a very important unsupervised learning method in deep learning. The auto-encoder is composed of an encoder and a decoder, which aims to learn the effective information from a large number of unmarked data, and realize the nonlinear compression and reconstruction of the input data. The goal of the traditional auto-encoder is to make the output and input as same as possible, but in the actual application, what we really care about is the hidden layer expression, so there are many improvement methods for the auto-encoder. The variational auto-encoder (VAE) ^[28] is one of the improved methods for auto-encoder. This method combines the deep learning method with Bayes on the basis of maintaining the basic function of the auto-encoder. It can be well applied to data generation ^[29] and anomaly detection in the case of no label data. Therefore, some researchers try to apply the VAE method to the field of trajectory outlier detection in different scenarios ^{[30,31,32,33]}.

3. Definitions for distances and VAE

3.1. Trajectory and distance

Trajectory is an important spatiotemporal data type, which represents the information history of the state of moving objects changing continuously with time. The trajectory can be regarded as a mapping from time to state. i.e., F: R⁺→S^d, where d is the dimension of the state space. For trajectory outlier detection, some preliminary definitions are discussed below.

Definition 1: A trajectory T is a sequence of time-ordered points, denoted by T = (p₀, p₂, …, p_i, …, p_n-1), where ${p}_{i}\in {\mathbb{R}}^{2}$ is the physical location (i.e., latitude and longitude), denoted by p_i = (lon_i, lat_i), p₁ and p_n-1 are the start and end points of the trajectory respectively.

Definition 2: A trajectory segment refers to the segment formed by the connection of adjacent points in the trajectory, denoted by seg_i(T) = (s_i, e_i), where s_i and is e_i are the start and end points of the segment.

Typically, measurement of distance between two trajectory segments include vertical distance (d_⊥), parallel distance(d_∥) and angular distance(d_θ) ^[18]. For two trajectory segments seg_i(T) = (s_i, e_i) and seg_j(T) = (s_j, e_j), the different distances between them are defined in Eqs (1)–(3), as shown in Figure 1.

${d_ \bot }(se{g_i}, se{g_j}) = \frac{{l_{ \bot 1}^2 + l_{ \bot 2}^2}}{{{l_{ \bot 1}} + {l_{ \bot 2}}}}$

(1)

${d_\parallel }(se{g_i}, se{g_j}) = \min ({l_{\parallel 1}}, {l_{\parallel 2}})$

(2)

${d_\theta }(se{g_i}, se{g_j}) = \left\{ \begin{array}{l} \left\| {se{g_j}} \right\| \times \sin \theta , {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} 0 \leqslant \theta < {90^ \circ } \hfill \\ \left\| {se{g_j}} \right\|, {\kern 1pt} {\kern 1pt} {90^ \circ } \leqslant \theta < {180^ \circ }{\kern 1pt} {\kern 1pt} \hfill \end{array} \right.{\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt}$

(3)

Figure 1. Space distance between trajectory segments.

DownLoad: Full-Size Img PowerPoint

3.2. Variational auto-encoder

Auto-encoder is a neural network that tries to copy the input to the output, that is, reproduce the original data as much as possible. The auto-encoder includes two parts: encoder and decoder.

The encoder converts the input signal into a hidden layer expression through a certain mapping to learn the characteristics of the data, as shown in Eq (4). The decoder tries to remap the hidden layer expression into the original input signal through the learned feature expression, as shown in Eq (5).

$z = {\sigma _e}({W_1}T + {b_1})$

(4)

$T' = {\sigma _d}({W_2}z + {b_2})$

(5)

T and T' represent the trajectory input and output space, ${W_1}$ and ${b_1}$ are the weights and offsets of the encoding stage, ${W_2}$ and ${b_2}$ are the weights and offsets of the decoding stage, ${\sigma _e}$ and ${\sigma _d}$ are the nonlinear transformations. The objective of auto-encoder optimization is to minimize the error between T and T' as much as possible.

Variational auto-encoder (VAE) is an improved model of auto-encoder proposed by Kingma et al. ^[16] in 2014, which is mainly used for data generation. The basic idea of VAE is to assume that all data are generated by statistical process, and the distribution characteristics of data should be considered in the process of encoding and decoding. Therefore, the difference between VAE and traditional auto-encoder is that the auto-encoder compresses the input data into a fixed code in the hidden space, while VAE converts the input data into parameters of statistical distribution function, i.e., mean and standard deviation (μ, σ). Finally, the hidden space distribution parameters are optimized through network training to maximize the generation probability of the original input data during decoding. The basic principle of VAE is shown in Figure 2.

Figure 2. Variational auto-encoder probability model.

DownLoad: Full-Size Img PowerPoint

In this paper, we reconstruct the trajectory through VAE. The VAE model is trained with the actual trajectory data set T, and finally the output trajectory data T' is generated by the hidden variable Z. T→Z is the feature extraction and recognition model ${q_\phi }(Z|T)$ , which is completed by the encoding process of the auto-encoder. Z→T' is the generation model ${p_\theta }(T'|Z)$ , which is completed by the decoding process of the auto-encoder. Assuming that the trajectory data T = [T₁, T₂, …, T_N] are all independent and identically distributed, the maximum likelihood method is used for parameter estimation, as shown in Eq (6).

$\log {P_\theta }({T_1}, {T_2}, ..., {T_N}) = \sum\limits_{i = 1}^N {\log {P_\theta }({T_i})}$

(6)

VAE uses feature extraction and recognition model ${q_\phi }(Z|{T_i})$ to approximate the real posterior probability ${p_\theta }(Z|{T_i})$ , and uses KL divergence to measure the similarity of the two distributions, as shown in Eq (7).

$\begin{array}{l} KL({q_\phi }(Z|{T_i})||{p_\theta }(Z|{T_i})) = {E_{{q_\phi }(Z|{T_i})}}\log \frac{{{q_\phi }(Z|{T_i})}}{{{p_\theta }(Z|{T_i})}} \hfill \\ = {E_{{q_\phi }(Z|{T_i})}}(\log \frac{{{q_\phi }(Z|{T_i})}}{{{p_\theta }(Z, {T_i})}} + \log {p_\theta }({T_i})) \hfill \end{array}$

(7)

Assuming that the sample input conforms to the normal distribution, we set two encoders in VAE, one for calculating the mean and the other for calculating the variance. In the mean value calculation network, the robustness of the result is improved by adding "Gaussian noise", and the encoder is regularized by KL loss to ensure that the encoder result has zero mean value. The variance calculation network is used to dynamically adjust the noise intensity. When the reconstruction result error is large (greater than the KL threshold), the noise is appropriately reduced. When the reconstruction result error is small (less than the KL threshold), the noise is appropriately increased and the generation ability of the decoder is improved through training. The essence of VAE is to find a suitable probability distribution parameter θ = (μ, σ) for each input X_i through continuous training. The overall neural network structure of VAE is shown in Figure 3.

Figure 3. The neural network architecture of VAE.

DownLoad: Full-Size Img PowerPoint

4. Trajectory outlier detection method based on variational auto-encoder

4.1. Model

Normal trajectories are generally continuous and smooth, so their trajectories reconstructed by VAE should maintain good consistency with the original trajectories. Abnormal trajectories are generally not smooth, usually manifested as position drift, sharp changes in speed or motion direction, so their trajectories reconstructed by VAE should be greatly different from the original trajectories. Under the guidance of this basic idea, a trajectory outlier detection model based on variational auto-encoder is proposed. The model consists of three parts: the trajectory data preprocessing module, the trajectory data generation module and the trajectory outlier determination module, as shown in Figure 4.

Figure 4. Trajectory outlier detection model based on VAE.

DownLoad: Full-Size Img PowerPoint

In this paper, Long Short-Term Memory (LSTM) network is used as the basic unit of encoder and decoder in VAE Network. LSTM is a classical sequence modeling network. It solves the problem of long sequence dependence through forgetting gate, control gate and output gate. The current output is determined according to the output at the previous time and the input at this time. The basic structure of LSTM we used in this paper is shown in Figure 5.

Figure 5. Basic LSTM structure of Encoder/Decoder.

DownLoad: Full-Size Img PowerPoint

4.1.1. Preprocessing of trajectory data

In general, the trajectories in the dataset are sequences of different lengths, while the input format required by the VAE is equal length sequence. Therefore, it is necessary to divide the trajectories in the data set into equal length sub trajectories first. In this paper, the sliding window method is used to process the original trajectory, which not only guarantees the integrity of the original trajectory, but also preserves the dependence of the trajectory points on the time series.

Let N multi-dimensional trajectory sequence samples T = [T₁, T₂, …, T_N], where a single sample with length M is represented as ${T^n} = (p_0^n, p_1^n, ..., p_i^n, ..., p_{M - 1}^n)$ , and $p_i^n$ represents the trajectory point at the ith time of the nth sample. We use longitude and latitude data to represent a trajectory point. When the sliding window size is w and the time step is s, the sample Tⁿ is segmented as shown in Eq (8).

${T^{{\text{s}}n}} = f({T^n}) = \left[ {\begin{array}{*{20}{c}} {T_1^n} \\ \begin{array}{l} T_2^n \hfill \\ \cdots \hfill \end{array} \\ \begin{array}{l} T_i^n \hfill \\ \cdots \hfill \end{array} \\ {T_d^n} \end{array}} \right] = \left[ {\begin{array}{*{20}{c}} \begin{array}{l} \left[ {p_0^n, \cdots , p_{w - 1}^n} \right] \hfill \\ \left[ {p_s^n, \cdots , p_{s{\text{ + }}w - 1}^n} \right] \hfill \\ \cdots \hfill \end{array} \\ {\left[ {p_{(i - 1)s}^n, \cdots , p_{{\text{(i - 1)}}s{\text{ + }}w - 1}^n} \right]} \\ \cdots \\ {\left[ {p_{(d - 1)s}^n, \cdots , p_{M - 1}^n} \right]} \end{array}} \right]$

(8)

After the actual trajectory data set is processed by sliding window, the equal length sub trajectory set can be obtained, which meets the equal length requirement of VAE network for input sequence. However, the spatial position coordinates of the trajectory points in the sub trajectory set do not meet the distribution consistency, which will cause the model to be biased. Therefore, the first-order difference method is used to further process the equal length sub trajectories. By calculating the first-order difference of the coordinates of the adjacent trajectory points in the trajectory segment, the position deviation P of the trajectory points at each time can be obtained and then the coordinate data of the trajectory points can be limited to [-max(P), +max(P)], which can ensure the consistence of the distribution of the point data.

4.1.2. VAE network training

The input for VAE network training is the trajectory data set of equal length and equal distribution after preprocessing, and the output is the distribution parameters of encoder and decoder. In the encoding stage, the parameters describing the distribution of each dimension in the hidden space are output through the neural network. Assuming that the trajectory data conforms to the normal distribution a priori, the mean and variance describing the distribution of the hidden state are output. In the decoding stage, the reparameterization trick is used to combine the mean and variance of the output in the encoding stage, sample from the standard normal distribution, generate the hidden variables through the neural network and reconstruct the original input and measure the distribution similarity through the divergence of Eq (7). The specific process of VAE network training is shown in Algorithm 1. Where, $\phi$ and $\theta$ represent parameters of Gaussian distribution of encoder and decoder respectively. In the process of model training, the learning rate is set to 0.005 and the convergence condition is that the training error is less than 0.01.

Algorithm 1 VAE model training
Input: T is the trajectory dataset, N is the size of dataset
Output: probabilistic encoder, probabilistic decoder
1: $\phi, \theta$ ← initialization parameter
2: repeat
3: for i = 1 to N do
4: get L samples from $\varepsilon \sim N(0, 1)$
5: ${z^{(i, l)}} = {h_\phi }({\varepsilon ^{(i)}}, {t^{(i)}}), {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} i = 1, ..., N$ // Generate hidden variables
6: end for
7: ${\text{E = }}\sum\limits_{i = 1}^N { - KL({q_\phi }(z\|{t^{(i)}})\|\|{p_\theta }(z)) + \frac{1}{L}\sum\limits_{i = 1}^L {(\log {p_\theta }({t^{(i)}}\|{z^{(i, l)}}))} }$
8: $\phi, \theta$ ← use gradients of E to update parameter
9: until $\phi, \theta$ parameter convergence

4.1.3. Trajectory outlier detection

After the training, the VAE network acquired the trajectory point distribution characteristics. For normal trajectories, the trajectories reconstructed by VAE should be consistent with the original trajectories. Therefore, outliers can be found through the differences between the original trajectories and the reconstructed trajectories. For moving objects or vehicles, sudden change in speed or direction during driving means abnormal occurrence. Considering the spatial distance, direction and time factors, the difference between the original trajectory and the reconstructed trajectory is expressed as the vertical distance (d_⊥), the parallel distance (d_∥) and the angular distance (d_θ), as shown in Eq (9).

$diff(se{g_i}, se{g_j}) = {\alpha _1}{d_ \bot }(se{g_i}, se{g_j}) + {\alpha _2}{d_\parallel }(se{g_i}, se{g_j}) + {\alpha _3}{d_\theta }(se{g_i}, se{g_j})$

(9)

α₁, α₂, α₃ are weight coefficients of vertical distance, parallel distance and angular distance respectively, α₁ + α₂ + α₃ = 1. If the difference between the original trajectory and the generated trajectory is greater than a certain threshold τ, the trajectory should be determined to be an outlier. The selection of threshold τ has an important impact on the detection results. In general, the threshold τ should be customized according to the experience in different scenarios, which is a difficult problem in practical application. In order to solve this problem, this paper introduces the ratio parameter λ of abnormal trajectory, and adjusts the threshold τ adaptively through the parameter λ to meet the actual application needs. Step 1, the trajectory data set to be detected is used as input to train the VAE model using Algorithm 1 to obtain the distribution parameters of encoder and decoder. Step 2, take each trajectory in the trajectory dataset to be detected as input, and use the decoder to regenerate the trajectory to obtain the generated trajectory set. Step 3, Calculate the distance between each trajectory in the validation dataset and its generated trajectory, and sort it in descending order. Step 4, Use the ratio λ of abnormal trajectories in the validation dataset as a condition to intercept the distance set and obtain the distance threshold τ of abnormal trajectories. Step 5, Determine whether each trajectory in the detected trajectory dataset is an abnormal trajectory based on the distance threshold τ. The specific trajectory outlier detection process is shown in Algorithm 2.

Algorithm 2 trajectory outlier detection
Input: T^N is the trajectory dataset for detecting, T^V is the validation dataset, λ is the abnormal proportion in T^V, α₁, α₂, α₃ are distance coefficients
Output: T^D is the abnormal trajectory data set detected
1: Encoder, Decoder ← VAEModelTrain (T^N, size(T^N))
2: refactoringSet ← Decoder.predict (T^V)
3: difference ←diff (T^V, refactoringSet, α₁, α₂, α₃)
4: τ ← reverseOrder (difference) [size(T^V)*λ]
5: for t in T^N do
6: if (diff (Decoder.predict (t), t) > τ)
7: T^D.append(t)
8: end if
9: end for
10: return T^D

5. Experiments and analysis

5.1. Experimental environment and dataset processing

The hardware platform of this experiment is CPU Intel (R) core (TM) i7-7700, 12 G memory, GPU Intel (R) HD graphics 630. The software platform of the experiment is operating system Windows10, machine learning framework Keras and algorithm implementation language Python.

The experimental data set is the public data set of taxis from Porto, Portugal, from early July 2013 to the end of June 2014 (https://www.kaggle.com/c/pkdd-15-predict-taxi-service-trajectory-i/data). The data set contains 1,710,670 tracks of 442 taxis in the whole year. The sampling frequency of trajectory points is 15 seconds, and some trajectories are shown in Figure 6.

Figure 6. Partial trajectories of experimental data set.

DownLoad: Full-Size Img PowerPoint

First, the data set is filtered to remove the discontinuous trajectories, and then the sliding window method is used to acquire trajectory segments with equal length. The window size is set to w = 25 and the step size is set to s = 1. Then, select the verification set for labeling, and divide the trajectories into normal trajectories and abnormal trajectories. The criterion for judging whether the trajectory is normal or not is whether it includes position drift, sharp turn or rapid acceleration. Figure 7 shows several typical normal and abnormal trajectories. Figure 7(a) shows the situation where the trajectory deviates from most normal trajectories, as such sharp turns cannot occur on normal trajectories, and even exceed the limits of taxi turning operation. Figure 7(b) shows time sequence of trajectory points is significantly different from normal trajectories or there may be abnormal acceleration in the trajectory. Finally, through the first-order difference processing and normalization processing, the regular trajectory data set is obtained for model training and verification.

Figure 7. Typical normal and abnormal trajectories.

DownLoad: Full-Size Img PowerPoint

5.2. Evaluation metrics

In this paper, precision, recall and F1 values are selected to evaluate the effect of the method we proposed. The calculation methods of the three indicators are shown in Eqs (10)–(12).

$Precision = \frac{{TP}}{{TP + FP}}$

(10)

$Recall = \frac{{TP}}{{TP + FN}}$

(11)

${F_1} = \frac{{2 \times Precision \times Recall}}{{Precision + Recall}}$

(12)

TP represents the number of outliers correctly identified as abnormal trajectories, FN represents the number of outliers incorrectly identified as normal trajectories, FP represents the number of normal trajectories incorrectly identified as outliers and TN represents the number of normal trajectories correctly identified as normal trajectories. In addition, the AUC value of ROC curve is used to evaluate the comprehensive recognition ability of the model for positive and negative samples.

5.3. Analysis of model generalization effect

5.3.1. Influence of training parameter epoch

In the experiment, 2000 trajectories were randomly selected from the preprocessed trajectory data set as the training set, and 500 trajectories were selected as the verification set, which including 400 normal tracks and 100 outliers. The threshold value λ in the trajectory outlier detection algorithm is set to 0.2. When the training parameter epochs = [200, 1000, 2000, 5000, 10,000], the outlier detection experiment was carried out separately. The effect of the model on each evaluation index is shown in Figure 8.

Figure 8. Model effect under different epochs.

DownLoad: Full-Size Img PowerPoint

Overall, the model has a high recognition effect, and the AUC value of the ROC curve always remains above 0.90. The difference of model effect is small under different training epochs, which reflects that the model has good convergence and stability. With the increase of the training epoch, the precision, recall and F1 value will increase first and then decrease slightly. The descending process shows that the trajectory reconstruction model may have a certain overfitting phenomenon in the process of learning the features. However, the problem of overfitting is not serious.

5.3.2. Influence of training set size train_size

This section analyzes the influence of the training set size on the model effect. Under the same verification set conditions as the previous section, set epochs = 200 and set the training set size parameter train_size = [1000, 1500, 2000, 3000, 5000], respectively carry out experiments on the model, and the performance of each evaluation index of the model is shown in Figure 9.

Figure 9. Model effect under different train_size.

DownLoad: Full-Size Img PowerPoint

From the experimental results, it can be seen that with the increase of the training set size, the trajectory reconstruction model learns more and more fully the normal trajectory distribution features, and the recognition effect of outliers is also improved. When the training set size is increased to a certain extent, the model effect enters a relatively stable convergence state. In addition, the model is not very sensitive to the size of the training set, and it can still obtain good results even when the training set is small (The prerequisite is that the trajectory distribution characteristics are sufficient). For the experimental data set in this paper, when the size of the training set reaches about 2000, the model can achieve good results and enter the convergence state.

5.3.3. Influence of threshold λ in outlier detection algorithm

In this paper, in order to solve the difficult problem of manually setting the distance threshold when judging the outliers, the distance threshold is automatically adjusted by the parameter λ, which is the ratio of the abnormal trajectory. This section analyzes the effect of parameter λ on the trajectory outlier detection algorithm through experiments.

We first get the trajectory generation model when the training epochs = 200 and the training set size train_size = 2000. 500 trajectories including 86 abnormal trajectories are randomly selected from the trajectory data set as the verification set, which the ratio is 0.172. When different thresholds λ = [0.10, 0.15, 0.20, 0.25, 0.40, 0.50] are set in the trajectory outlier detection algorithm, the effect of the algorithm is shown in Figure 10.

Figure 10. Model effect under different threshold λ.

DownLoad: Full-Size Img PowerPoint

From the experimental results, it can be seen that when the threshold λ of the trajectory outlier detection algorithm is set between 0.15 and 0.25, the algorithm achieves good results and the recall, precision and F1 value all remain at a high level. When λ is set to 0.2, the effect of the algorithm is optimal, and the recall, precision and F1 value are all kept above 95%. When λ is set to be greater than or equal to 0.4, the recall rate will reach 100%, but the error rate will increase, which will lead to a decrease in the accuracy rate. The experimental results show that the algorithm achieves good results when the value of λ is near ratio of abnormal trajectory in the verification set. From this, we can learn that in practical application, if we pursue high precision, we can obtain the approximate proportion of outliers according to the samples in the data set, so as to provide a reference for setting the parameter λ of the algorithm. According to different application requirements, if we pursue higher recall, we can appropriately increase the λ value in the trajectory outlier detection algorithm.

5.4. Effect comparison

We select three typical types of methods as benchmarks. One is the density-based method, one is the classification-based method and the other is the deep learning based method.

1) Density-based methods: DTW distance-based clustering and Hausdorff distance based clustering.

2) Classification-based methods: KNN and deep clustering method ^[3].

3) Deep learning based methods: Sequence autoencoder (SAE) model ^[32] and GM-VSAE model ^[33].

For density-based method, we implement the two methods using the DBSCAN algorithm interface provided by Scikit learn library. The distance threshold value T₀ for the trajectory outlier of the clustering algorithm is 0.1, 0.3, 0.5, 0.7 and 0.9 respectively, and the quantity threshold value T₁ is 3, 5, 7, 9 and 11 respectively. The results of the optimal accuracy are used as the comparison benchmark.

For KNN and deep clustering methods, we implement the two methods based on literature ^[3]. The public data set of taxis from Porto, Portugal is used in literature ^[3] which is the same as that in this paper.

For deep learning-based method, we implement SAE and GM-VSAE algorithms based on literature ^[33]. The basic data sets used in the experiment are all also from the public data set of taxis from Porto, Portugal as the same as this paper. Literature ^[33] mainly focuses on two types of anomaly detection: detour and route switching. Therefore, in literature ^[33], drift points and abnormal turning points in trajectories are eliminated in data set preprocessing and then two different perturbation schemes are used to generate two types of anomalous trajectories. This paper does not distinguish between different types of abnormal trajectories, so it is not necessary to eliminate these abnormal points in the data preprocessing stage. On the contrary, we think that these outliers are very important and may reflect the occurrence of some special events. For SAE and GM-VSAE models, we take the best results under different parameter conditions for comparison.

The comparison indexes are still precision, recall, F1 value and PR-AUC value. The experimental results of trajectory outlier detection by different methods are shown in Table 1.

Table 1. Test results of different methods.

Methods	Precision	Recall	F1	PR-AUC
DTW	0.649	0.726	0.685	0.711
Hausdorff	0.876	0.953	0.916	0.836
KNN ^[3]	0.713	0.601	0.656	0.64
Deep clustering ^[3]	0.862	0.783	0.821	0.84
SAE ^[32]	0.812	0.884	0.846	0.829
GM-VSAE ^[33]	0.832	0.941	0.883	0.871
VAE (ours)	0.960	0.965	0.963	0.943

| Show Table

DownLoad: CSV

From the experimental results in Table 1, it can be seen that the trajectory outlier detection method based on variational auto-encoder VAE proposed in this paper is superior to the reference methods in all indicators.

The DTW based clustering method can coordinate the time alignment between trajectory points, and can achieve better results for clustering of unequal length trajectories. In this paper, the influence of unequal length trajectories is eliminated through data preprocessing. Therefore, the method based on DTW clustering cannot show advantages. The data set selected in this paper is the all-weather trajectory data of taxis. There is a large speed difference between the trajectory points in different periods. The clustering method based on Hausdorff distance can locally optimize the dislocation alignment between the trajectory points, and has a certain coordination for the speed difference between trajectory points. It can achieve good results when used in the clustering of urban vehicle trajectories in different periods in theory. However, the clustering method based on Hausdorff distance is sensitive to local outliers, which will increase the false positive rate to a certain extent. In the experimental results, the precision of the clustering method based on Hausdorff distance is lower than the recall rate, which also shows this. KNN also needs to calculate the distance in the clustering process, so the effect is lower than DTW and Hausdorff which are based on density clustering. Deep clustering method trained a binary pairwise deep neural network to cluster the sequence of trajectory represented as trips. Dynamic Time Warping (DTW) is used to calculate the distance between two ordered degree sequences. This method has achieved good results, but road network information needs to be used in trajectory data preprocessing.

The PR-AUC value of method SEA and method GM-VSAE exceeds 0.8 under the best parameters, which is a good result. However, the premise of such a good result is that it is oriented to specific types of anomalies diagnosis. When these two methods are used to diagnose another type of anomalies, they may not be suitable. For example, when these two methods are used to diagnose route switching anomalies, the PR-AUC value drops to about 0.7.

In this paper, we use the VAE model to obtain the distribution characteristics of trajectory points, which can also eliminate the influence of velocity difference between different trajectories through trajectory reconstruction. It is also less affected by local outliers, so it performs better in accuracy. In addition, the method in this paper obtains the distribution characteristics of normal trajectory points through the first-order difference method in the learning stage, and it is not necessary to distinguish different types of anomalies in the diagnosis stage.

5.5. Efficiency analysis

In practical application, when the trajectory outlier detection method achieves more than 90% recall and precision, it will have a good usability. In the field of urban traffic management under the background of the current big data era, the demand for real-time trajectory outlier detection is increasingly urgent ^[2]. This paper also compares and analyze the efficiency of the proposed VAE model to the baseline methods selected as above.

Clustering algorithms based on DTW distance and Hausdorff distance generally divide N trajectories into several clusters through DBSCAN clustering algorithm. Finally, clusters with less than the threshold T_c are detected as outliers. In the clustering stage, the distance between trajectories needs to be calculated and the clustering process needs to be carried out. Therefore, the theoretical time complexity of this kind of methods is O(N²), and it can reach O(NlogN) after optimization. In the case of real-time detection, the distance between a given trajectory and all trajectories in the set needs to be calculated in the detection stage of these methods. Without considering the trajectory segmentation, the time consumption will increase linearly with the increase of the amount of trajectories in the set. For KNN and deep clustering methods, it is also necessary to calculate the distance in the clustering process, so its efficiency should be equal to or lower than DTW and Hausdorff.

For machine learning based methods, the training time of the model depends on the super parameters of the model, training set size and the convergence rate. Once the training is completed, only the distance between the trajectory itself and the reconstructed trajectory needs to be calculated in the detection phase, so the time complexity of the detection phase is O(1). Because the VAE model proposed in this paper can be trained by sampling data set, the training data set is generally small and the training time is relatively short. While models SAE and GM-VSAE require training of all samples, so the training time is relatively long.

Under the same experiment conditions as in the previous section, set epochs = 200 and set the data set size N = [1000, 2000, 4000], we compare the clustering or training time and detection time of the baseline methods under different data sizes. For the clustering-based method, the training time refers to the calculation of the distance between trajectories and the clustering time; for the VAE method, the training time refers to the time required for the model training to converge. Detection time refers to the time taken to detect a single trajectory after clustering or model training is completed.

In the experiment, the DTW distance and Hausdorff distance were calculated by 32 threads in parallel. The experimental results are shown in Table 2. The training time and detection time in the table are the average of the experimental results.

Table 2. Computation efficiency of different methods (unit: second).

Methods	N = 1000		N = 2000		N = 4000		N > 100,000
Methods	Training	Detection	Training	Detection	Training	Detection	Training	Detection
DTW	31.65	0.35	135.376	1.304	513.228	5.446	-	-
Hausdorff	29.9053	1.62	126.713	10.696	514.683	61.490	-	-
SAE ^[32]	123.22	0.01	144.33	0.01	245.69	0.01	> 1000	< = 0.01
GM-VSAE ^[33]	124.39	0.01	146.82	0.01	268.24	0.01	> 1000	< = 0.01
VAE (ours)	110.17	0.01	130.32	0.01	180.38	0.01	< 1000	< = 0.01

| Show Table

DownLoad: CSV

It can be seen from the experimental results in Table 2 that when n = 1000, the VAE method consumes more training time than DTW and Hausdorff; When n = 2000, the VAE method takes about the same time as DTW and Hausdorff method; When n = 4000, the time-consuming of VAE method is significantly less than that of DTW and Hausdorff methods. This shows that the VAE method proposed in this paper has good convergence in the model training stage. The training time increases slowly with the growth of the data set size, so it can be applied to large-scale data set. In detection time, the VAE, SAE and GM-VSAE methods are all far less time-consuming than DTW and Hausdorff method, which can be ignored in practical application, because deep learning based method only needs to calculate the distance between the trajectory itself and the reconstructed trajectory. This again shows that the VAE based method proposed in this paper is very suitable for real-time detection of large-scale data set.

6. Conclusions

This paper introduces a trajectory outlier detection model based on variational auto-encoder in detail. Based on the statistical characteristics of the normal urban traffic trajectory data, the model converts the trajectory data into a distribution function using a variational auto-encoder, and optimizes the distribution parameters through historical data training to ensure that the generation probability of the normal original trajectory data is maximized when decoding. Finally, the outlier is detected by calculating the difference between the generated trajectory by the trained model and the original trajectory. The biggest advantage of the model proposed in this paper is that the outlier detection only needs to calculate the difference between the original trajectory and the generated trajectory by the model. Compared with the density-based methods and classification-based methods, the calculation amount is greatly reduced, which makes it very suitable for the real-time detection in large-scale data environments. In addition, the model can use the ratio of outliers in the verification data set to define the detection threshold of outliers, which can eliminate the difficulty of setting the distance threshold artificially, and make the applicable scenarios of the model more abundant. The experimental results on the real urban traffic trajectory data set show that the model proposed in this paper is very suitable for large-scale data real-time detection scenarios. In terms of effect, the precision and recall of the proposed model are over 95%, which is better than the methods we selected for comparison; in terms of efficiency, the model has good convergence in the training stage. The training time of the model increases slowly with the size of the data. The time consumption in the detection stage is a constant level, which is far better than some reference methods.

This paper verifies the effectiveness and efficiency of a trajectory outlier detection model based on variational auto-encoder. The data set used in the model training is urban traffic all-weather trajectory data. In view of the strong correlation between urban traffic trajectories and space-time, in practical applications, the data set can be spatiotemporal divided according to specific application needs, and model training can be carried out according to different spatiotemporal data set to obtain trajectory distribution characteristics under specific space-time, which can further improve the effect and efficiency of the model. This issue is worth further research in the future.

Use of AI tools declaration

The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

Acknowledgments

This work was supported in part by National Natural Science Foundation of China (No. 61801373) and Silk Road Research (2019YA07).

Conflict of interest

The authors declare there is no conflict of interest.

References

[1]	Holtfreter B, Kocher T, Hoffmann T, et al. (2010) Prevalence of periodontal disease and treatment demands based on a German dental survey (DMS IV). J Clin Periodontol 37: 211–219. doi: 10.1111/j.1600-051X.2009.01517.x
[2]	Fujita T, Yamamoto S, Ota M, et al. (2011) Coverage of gingival recession defects using guided tissue regeneration with and without adjunctive enamel matrix derivative in a dog model. Int J Periodont Rest 31: 247–253.
[3]	Sun HH, Qu TJ, Zhang XH, et al. (2012) Designing biomaterials for in situ periodontal tissue regeneration. Biotechnol Progr 28: 3–20. doi: 10.1002/btpr.698
[4]	Lloyd AW (2002) Interfacial bioengineering to enhance surface biocompatibility. Med Device Technol 13: 18–21.
[5]	Schakenraad JM, Hardonk MJ, Feijen J, et al. (1990) Enzymatic activity toward poly(L-lactic acid) implants. J Biomed Mater Res 24: 529–545. doi: 10.1002/jbm.820240502
[6]	Finer Y, Santerre JP (2004) Salivary esterase activity and its association with the biodegradation of dental composites. J Dent Res 83: 22–26. doi: 10.1177/154405910408300105
[7]	Rokaya D, Srimaneepong V, Sapkota J, et al. (2018) Polymeric materials and films in dentistry: An overview. J Adv Res 14: 25–34. doi: 10.1016/j.jare.2018.05.001
[8]	Yuan W, Ji J, Fu J, et al. (2008) A facile method to construct hybrid multilayered films as a strong and multifunctional antibacterial coating. J Biomed Mater Res B 85: 556–563.
[9]	Travan A, Marsich E, Donati I, et al. (2011) Silver–polysaccharide nanocomposite antimicrobial coatings for methacrylic thermosets. Acta Biomater 7: 337–346. doi: 10.1016/j.actbio.2010.07.024
[10]	Mironov V, Visconti RP, Markwald RR (2004) What is regenerative medicine? Emergence of applied stem cell and developmental biology. Expert Opin Biol Th 4: 773–781.
[11]	Rajeswari R, Clarisse CHN, Susan L, et al. (2012) Biomimetic surface modification of titanium surfaces for early cell capture by advanced electrospinning. Biomed Mater 7: 015001. doi: 10.1088/1748-6041/7/1/015001
[12]	Riccitiello F, De Luise A, Conte R, et al. (2018) Effect of resveratrol release kinetic from electrospun nanofibers on osteoblast and osteoclast differentiation. Eur Polym J 99: 289–297. doi: 10.1016/j.eurpolymj.2017.12.035
[13]	Coulembier O, Degée P, Hedrick JL, et al. (2006) From controlled ring-opening polymerization to biodegradable aliphatic polyester: Especially poly(β-malic acid) derivatives. Prog Polym Sci 31: 723–747. doi: 10.1016/j.progpolymsci.2006.08.004
[14]	Maurus PB, Kaeding CC (2004) Bioabsorbable implant material review. Oper Techn Sport Med 12: 158–160. doi: 10.1053/j.otsm.2004.07.015
[15]	Gunatillake P, Mayadunne R, Adhikari R (2006) Recent developments in biodegradable synthetic polymers. Biotechnol Annu Rev 12: 301–347. doi: 10.1016/S1387-2656(06)12009-8
[16]	Katz AR, Turner RJ (1970) Evaluation of tensile and absorption properties of polyglycolic acid sutures. Surg Gynecol Obstet 131: 701–716.
[17]	Reed TM (1999) Allofix freeze-dried cortical bone pins as an alternative to synthetic absorbable polymeric pins: a preliminary study in short Z bunionectomies. J Foot Ankle Surg 38: 14–23. doi: 10.1016/S1067-2516(99)80083-0
[18]	Knecht S, Erggelet C, Endres M, et al. (2007) Mechanical testing of fixation techniques for scaffold-based tissue-engineered grafts. J Biomed Mater Res B 83: 50–57.
[19]	Frisbie DD, Lu Y, Kawcak CE, et al. (2009) In vivo evaluation of autologous cartilage fragment-loaded scaffolds implanted into equine articular defects and compared with autologous chondrocyte implantation. Am J Sport Med 37: 71–80. doi: 10.1177/0363546509348478
[20]	Xu L, Cao D, Liu W, et al. (2010) In vivo engineering of a functional tendon sheath in a hen model. Biomaterials 31: 3894–3902. doi: 10.1016/j.biomaterials.2010.01.106
[21]	Sayasneh A, Johnson H (2010) Risk factors for mesh erosion complicating vaginal reconstructive surgery. J Obstet Gynaecol 30: 721–724. doi: 10.3109/01443615.2010.501921
[22]	Aysan E, Bektas H, Ersoz F, et al. (2010) A novel colonic anastomosis technique involving fixed polyglycolic acid mesh. Int J Clin Exp Med 3: 341–346.
[23]	Dai T, Jiang Z, Li S, et al. (2010) Reconstruction of lymph vessel by lymphatic endothelial cells combined with polyglycolic acid scaffolds: a pilot study. J Biotechnol 150: 182–189.
[24]	Abbushi A, Endres M, Cabraja M, et al. (2008) Regeneration of intervertebral disc tissue by resorbable cell-free polyglycolic acid-based implants in a rabbit model of disc degeneration. Spine 33: 1527–1532. doi: 10.1097/BRS.0b013e3181788760
[25]	Ohara T, Itaya T, Usami K, et al. (2010) Evaluation of scaffold materials for tooth tissue engineering. J Biomed Mater Res A 94: 800–805.
[26]	Chang HH, Wang YL, Chiang YC, et al. (2014) A novel chitosan-γPGA polyelectrolyte complex hydrogel promotes early new bone formation in the alveolar socket following tooth extraction. PLoS One 9: e92362. doi: 10.1371/journal.pone.0092362
[27]	Middleton JC, Tipton AJ (2000) Synthetic biodegradable polymers as orthopedic devices. Biomaterials 21: 2335–2346. doi: 10.1016/S0142-9612(00)00101-0
[28]	Suuronen R, Pohjonen T, Hietanen J, et al. (1998) A 5-year in vitro and in vivo study of the biodegradation of polylactide plates. J Oral Maxil Surg 56: 604–614. doi: 10.1016/S0278-2391(98)90461-X
[29]	Loo JS, Ooi CP, Boey FY (2005) Degradation of poly(lactide-co-glycolide) (PLGA) and poly(L-lactide) (PLLA) by electron beam radiation. Biomaterials 26: 1359–1367. doi: 10.1016/j.biomaterials.2004.05.001
[30]	Shim IK, Jung MR, Kim KH, et al. (2010) Novel three-dimensional scaffolds of poly(L-lactic acid) microfibers using electrospinning and mechanical expansion: Fabrication and bone regeneration. J Biomed Mater Res B 95: 150–160.
[31]	Tanaka Y, Yamaoka H, Nishizawa S, et al. (2010) The optimization of porous polymeric scaffolds for chondrocyte/atelocollagen based tissue-engineered cartilage. Biomaterials 31: 4506–4516. doi: 10.1016/j.biomaterials.2010.02.028
[32]	Inui A, Kokubu T, Makino T, et al. (2010) Potency of double-layered poly L-lactic acid scaffold in tissue engineering of tendon tissue. Int Orthop 34: 1327–1332. doi: 10.1007/s00264-009-0917-8
[33]	Wang HB, Mullins ME, Cregg JM, et al. (2010) Varying the diameter of aligned electrospun fibers alters neurite outgrowth and Schwann cell migration. Acta Biomater 6: 2970–2978. doi: 10.1016/j.actbio.2010.02.020
[34]	Francois S, Chakfe N, Durand B, et al. (2009) A poly(L-lactic acid) nanofibre mesh scaffold for endothelial cells on vascular prostheses. Acta Biomater 5: 2418–2428. doi: 10.1016/j.actbio.2009.03.013
[35]	Tang L, Zhao C, Xiong Y, et al. (2010) Preparation, antibacterial properties and biocompatibility studies on vancomycin-poly(D,L)-lactic loaded plates. Int Orthop 34: 755–759. doi: 10.1007/s00264-009-0808-z
[36]	Carletti E, Endogan T, Hasirci N, et al. (2011) Microfabrication of PDLLA scaffolds. J Tissue Eng Regen M 5: 569–577. doi: 10.1002/term.349
[37]	Lee SH, Kim IY, Song WS (2014) Biodegradation of polylactic acid (PLA) fibers using different enzymes. Macromol Res 22: 657–663. doi: 10.1007/s13233-014-2107-9
[38]	Tomlin EM, Nelson SJ, Rossmann JA (2014) Ridge preservation for implant therapy: a review of the literature. Open Dent J 8: 66–76. doi: 10.2174/1874210601408010066
[39]	Thomas N, Sanil G, Rajmohan G, et al. (2011) Fabrication and anti-microbial evaluation of drug loaded polylactide space filler intended for ridge preservation following tooth extraction. J Indian Soc Periodontol 15: 260–264. doi: 10.4103/0972-124X.85671
[40]	Serino G, Biancu S, Iezzi G, et al. (2003) Ridge preservation following tooth extraction using a polylactide and polyglycolide sponge as space filler: a clinical and histological study in humans. Clin Oral Implan Res 14: 651–658. doi: 10.1034/j.1600-0501.2003.00970.x
[41]	Makadia HK, Siegel SJ (2011) Poly lactic-co-glycolic acid (PLGA) as biodegradable controlled drug delivery carrier. Polymers 3: 1377–1397. doi: 10.3390/polym3031377
[42]	Conn Jr. J, Oyasu R, Welsh M, et al. (1974) Vicryl (polyglactin 910) synthetic absorbable sutures. Am J Surg 128: 19–23. doi: 10.1016/0002-9610(74)90228-1
[43]	Liu J, Qiu Z, Wang S, et al. (2010) A modified double-emulsion method for the preparation of daunorubicin-loaded polymeric nanoparticle with enhanced in vitro anti-tumor activity. Biomed Mater 5: 065002. doi: 10.1088/1748-6041/5/6/065002
[44]	Ye M, Kim S, Park K (2010) Issues in long-term protein delivery using biodegradable microparticles. J Control Release 146: 241–260. doi: 10.1016/j.jconrel.2010.05.011
[45]	Jiang W, Schwendeman SP (2008) Stabilization of tetanus toxoid encapsulated in PLGA microspheres. Mol Pharmaceut 5: 808–817. doi: 10.1021/mp800027f
[46]	Jhunjhunwala S, Raimondi G, Thomson AW, et al. (2009) Delivery of rapamycin to dendritic cells using degradable microparticles. J Control Release 133: 191–197. doi: 10.1016/j.jconrel.2008.10.011
[47]	Vega E, Gamisans F, Garcia ML, et al. (2008) PLGA nanospheres for the ocular delivery of flurbiprofen: drug release and interactions. J Pharm Sci 97: 5306–5317. doi: 10.1002/jps.21383
[48]	Eperon S, Bossy-Nobs L, Petropoulos IK, et al. (2008) A biodegradable drug delivery system for the treatment of postoperative inflammation. Int J Pharmaceut 352: 240–247. doi: 10.1016/j.ijpharm.2007.10.054
[49]	Murata N, Takashima Y, Toyoshima K, et al. (2008) Anti-tumor effects of anti-VEGF siRNA encapsulated with PLGA microspheres in mice. J Control Release 126: 246–254. doi: 10.1016/j.jconrel.2007.11.017
[50]	Levin BP (2013) Alveolar ridge augmentation: combining bioresorbable scaffolds with osteoinductive bone grafts in atrophic sites. A follow-up to an evolving technique. Compend Contin Educ Dent 34: 178–186.
[51]	Hassan KS (2009) Autogenous bone graft combined with polylactic polyglycolic acid polymer for treatment of dehiscence around immediate dental implants. Oral Surg Oral Med O 108: e19–e25.
[52]	Jones AA, Buser D, Schenk R, et al. (2006) The effect of rhBMP-2 around endosseous implants with and without membranes in the canine model. J Periodontol 77: 1184–1193. doi: 10.1902/jop.2006.050337
[53]	Wu Z, Liu C, Zang G, et al. (2008) The effect of simvastatin on remodelling of the alveolar bone following tooth extraction. Int J Oral Max Surg 37: 170–176. doi: 10.1016/j.ijom.2007.06.018
[54]	Sousa FFO, Luzardo-Álvarez A, Pérez-Estévéz A, et al. (2010) Development of a novel AMX-loaded PLGA/zein microsphere for root canal disinfection. Biomed Mater 5: 055008. doi: 10.1088/1748-6041/5/5/055008
[55]	Zhao H, Wu B, Wu H, et al. (2006) Protective immunity in rats by intranasal immunization with Streptococcus mutans glucan-binding protein D encapsulated into chitosan-coated poly(lactic-co-glycolic acid) microspheres. Biotechnol Lett 28: 1299–1304. doi: 10.1007/s10529-006-9086-7
[56]	Jamal T, Rahman A, Mirza A, et al. (2012) Formulation, antimicrobial and toxicity evaluation of bioceramic based ofloxacin loaded biodegradable microspheres for periodontal infection. Curr Drug Deliv 9: 515–526. doi: 10.2174/156720112802650644
[57]	Chang PC, Dovban AS, Lim LP, et al. (2013) Dual delivery of PDGF and simvastatin to accelerate periodontal regeneration in vivo. Biomaterials 34: 9990–9997. doi: 10.1016/j.biomaterials.2013.09.030
[58]	Brown A, Zaky S, Ray H, et al. (2015) Porous magnesium/PLGA composite scaffolds for enhanced bone regeneration following tooth extraction. Acta Biomater 11: 543–553. doi: 10.1016/j.actbio.2014.09.008
[59]	Van Manen EHC, Zhang W, Walboomers XF, et al. (2014) The influence of electrospun fibre scaffold orientation and nano-hydroxyapatite content on the development of tooth bud stem cells in vitro. Odontology 102: 14–21. doi: 10.1007/s10266-012-0087-9
[60]	Shirakata Y, Nakamura T, Shinohara Y, et al. (2014) An exploratory study on the efficacy of rat dedifferentiated fat cells (rDFATs) with a poly lactic-co-glycolic acid/hydroxylapatite (PLGA/HA) composite for bone formation in a rat calvarial defect model. J Mater Sci-Mater M 25: 899–908. doi: 10.1007/s10856-013-5124-x
[61]	Marei MK, Saad MM, El-Ashwah AM, et al. (2009) Experimental formation of periodontal structure around titanium implants utilizing bone marrow mesenchymal stem cells: a pilot study. J Oral Implantol 35: 106–129. doi: 10.1563/1548-1336-35.3.106
[62]	Akita D, Morokuma M, Saito Y, et al. (2014) Periodontal tissue regeneration by transplantation of rat adipose-derived stromal cells in combination with PLGA-based solid scaffolds. Biomed Res 35: 91–103. doi: 10.2220/biomedres.35.91
[63]	Zhang H, Wang J, Ma H, et al. (2016) Bilayered PLGA/wool keratin composite membranes support periodontal regeneration in beagle dogs. ACS Biomater Sci Eng 2: 2162–2175. doi: 10.1021/acsbiomaterials.6b00357
[64]	Naito Y, Terukina T, Galli S, et al. (2014) The effect of simvastatin-loaded polymeric microspheres in a critical size bone defect in the rabbit calvaria. Int J Pharmaceut 461: 157–162. doi: 10.1016/j.ijpharm.2013.11.046
[65]	Yonamine Y, Matsuyama T, Sonomura T, et al. (2010) Effectable application of vascular endothelial growth factor to critical sized rat calvaria defects. Oral Surg Oral Med O 109: 225–231. doi: 10.1016/j.tripleo.2009.09.010
[66]	Son JS, Choi YA, Park EK, et al. (2013) Drug delivery from hydroxyapatite-coated titanium surfaces using biodegradable particle carriers. J Biomed Mater Res B 101: 247–257.
[67]	Wang F, Song YL, Li CX, et al. (2010) Sustained release of insulin-like growth factor-1 from poly(lactide-co-glycolide) microspheres improves osseointegration of dental implants in type 2 diabetic rats. Eur J Pharmacol 640: 226–232. doi: 10.1016/j.ejphar.2010.04.024
[68]	Han Y, Zeng Q, Lingling E, et al. (2012) Sustained topical delivery of insulin from fibrin gel loaded with poly(lactic-co-glycolic Acid) microspheres improves the biomechanical retention of titanium implants in type 1 diabetic rats. J Oral Maxil Surg 70: 2299–2308. doi: 10.1016/j.joms.2012.05.028
[69]	Gentile P, Frongia ME, Cardellach M, et al. (2015) Functionalised nanoscale coatings using layer-by-layer assembly for imparting antibacterial properties to polylactide-co-glycolide surfaces. Acta Biomater 21: 35–43. doi: 10.1016/j.actbio.2015.04.009
[70]	Tay FR, Pashley DH, Yiu CKY, et al. (2005) Susceptibility of a polycaprolactone-based root canal filling material to degradation. II. Gravimetric evaluation of enzymatic hydrolysis. J Endodont 31: 737–741.
[71]	Patlolla A, Collins G, Arinzeh TL (2010) Solvent-dependent properties of electrospun fibrous composites for bone tissue regeneration. Acta Biomater 6: 90–101. doi: 10.1016/j.actbio.2009.07.028
[72]	Nobile MR, Lucia G, Santella M, et al. (2015) Biodegradable compounds: Rheological, mechanical and thermal properties. AIP Conf Proc 1695: 020058.
[73]	Danhier F, Vroman B, Lecouturier N, et al. (2009) Targeting of tumor endothelium by RGD-grafted PLGA-nanoparticles loaded with paclitaxel. J Control Release 140: 166–173. doi: 10.1016/j.jconrel.2009.08.011
[74]	Chung S, Ingle NP, Montero GA, et al. (2010) Bioresorbable elastomeric vascular tissue engineering scaffolds via melt spinning and electrospinning. Acta Biomater 6: 1958–1967. doi: 10.1016/j.actbio.2009.12.007
[75]	Guarino V, Ambrosio L (2008) The synergic effect of polylactide fiber and calcium phosphate particle reinforcement in poly epsilon-caprolactone-based composite scaffolds. Acta Biomater 4: 1778–1787. doi: 10.1016/j.actbio.2008.05.013
[76]	Zuo Y, Yang F, Wolke JG, et al. (2010) Incorporation of biodegradable electrospun fibers into calcium phosphate cement for bone regeneration. Acta Biomater 6: 1238–1247. doi: 10.1016/j.actbio.2009.10.036
[77]	Hayami JW, Surrao DC, Waldman SD, et al. (2010) Design and characterization of a biodegradable composite scaffold for ligament tissue engineering. J Biomed Mater Res A 92: 1407–1420.
[78]	Li WJ, Chiang H, Kuo TF, et al. (2009) Evaluation of articular cartilage repair using biodegradable nanofibrous scaffolds in a swine model: a pilot study. J Tissue Eng Regen M 3: 1–10. doi: 10.1002/term.127
[79]	Nisbet DR, Rodda AE, Horne MK, et al. (2009) Neurite infiltration and cellular response to electrospun polycaprolactone scaffolds implanted into the brain. Biomaterials 30: 4573–4580. doi: 10.1016/j.biomaterials.2009.05.011
[80]	Heydarkhan-Hagvall S, Schenke-Layland K, Dhanasopon AP, et al. (2008) Three-dimensional electrospun ECM-based hybrid scaffolds for cardiovascular tissue engineering. Biomaterials 29: 2907–2914. doi: 10.1016/j.biomaterials.2008.03.034
[81]	Huang MH, Chou AH, Lien SP, et al. (2009) Formulation and immunological evaluation of novel vaccine delivery systems based on bioresorbable poly(ethylene glycol)-block-poly(lactide-co-ε-caprolactone). J Biomed Mater Res B 90: 832–841.
[82]	Louvrier A, Euvrard E, Nicod L, et al. (2017) Odontoblastic differentiation of dental pulp stem cells from healthy and carious teeth on an original PCL-based 3D scaffold. Int Endod J 51: e252–e263.
[83]	Chuenjitkuntaworn B, Osathanon T, Nowwarote N, et al. (2016) The efficacy of polycaprolactone/hydroxyapatite scaffold in combination with mesenchymal stem cells for bone tissue engineering. J Biomed Mater Res A 104: 264–271. doi: 10.1002/jbm.a.35558
[84]	Baranowska-Korczyc A, Warowicka A, Jasiurkowska-Delaporte M, et al. (2016) Antimicrobial electrospun poly(ɛ-caprolactone) scaffolds for gingival fibroblast growth. RSC Adv 6: 19647–19656. doi: 10.1039/C6RA02486F
[85]	He S, Timmer MD, Yaszemski MJ, et al. (2001) Synthesis of biodegradable poly(propylene fumarate) networks with poly(propylene fumarate)-diacrylate macromers as crosslinking agents and characterization of their degradation products. Polymer 42: 1251–1260. doi: 10.1016/S0032-3861(00)00479-1
[86]	Christenson EM, Soofi W, Holm JL, et al. (2007) Biodegradable fumarate-based polyHIPEs as tissue engineering scaffolds. Biomacromolecules 8: 3806–3814. doi: 10.1021/bm7007235
[87]	Hacker MC, Haesslein A, Ueda H, et al. (2009) Biodegradable fumarate-based drug-delivery systems for ophthalmic applications. J Biomed Mater Res A 88: 976–989.
[88]	Jayabalan M, Shalumon KT, Mitha MK, et al. (2010) Effect of hydroxyapatite on the biodegradation and biomechanical stability of polyester nanocomposites for orthopaedic applications. Acta Biomater 6: 763–775. doi: 10.1016/j.actbio.2009.09.015
[89]	Mistry AS, Pham QP, Schouten C, et al. (2010) In vivo bone biocompatibility and degradation of porous fumarate-based polymer/alumoxane nanocomposites for bone tissue engineering. J Biomed Mater Res A 92: 451–462.
[90]	Trachtenberg JE, Placone JK, Smith BT, et al. (2017) Extrusion-based 3D printing of poly(propylene fumarate) scaffolds with hydroxyapatite gradients. J Biomat Sci-Polym E 28: 532–554. doi: 10.1080/09205063.2017.1286184
[91]	Alge DL, Bennet J, Treasure T, et al. (2012) Poly(propylene fumarate) reinforced dicalcium phosphate dihydrate cement composites for bone tissue engineering. J Biomed Mater Res A 100: 1792–1802.
[92]	Shahabi S, Rezaei Y, Moztarzadeh F, et al. (2016) In vitro degradation and bioactivity of poly(propylene fumarate)/bioactive glass sintered microsphere scaffolds for bone tissue engineering. Sci Eng Compos Mater 23: 245–256.
[93]	Agueros M, Zabaleta V, Espuelas S, et al. (2010) Increased oral bioavailability of paclitaxel by its encapsulation through complex formation with cyclodextrins in poly(anhydride) nanoparticles. J Control Release 145: 2–8. doi: 10.1016/j.jconrel.2010.03.012
[94]	Krasko MY, Golenser J, Nyska A, et al. (2007) Gentamicin extended release from an injectable polymeric implant. J Control Release 117: 90–96. doi: 10.1016/j.jconrel.2006.10.010
[95]	Tamayo I, Irache JM, Mansilla C, et al. (2010) Poly(anhydride) nanoparticles act as active Th1 adjuvants through Toll-like receptor exploitation. Clin Vaccine Immunol 17: 1356–1362. doi: 10.1128/CVI.00164-10
[96]	Petersen LK, Sackett CK, Narasimhan B (2010) High-throughput analysis of protein stability in polyanhydride nanoparticles. Acta Biomater 6: 3873–3881. doi: 10.1016/j.actbio.2010.04.004
[97]	Weiner AA, Shuck DM, Bush JR, et al. (2007) In vitro degradation characteristics of photocrosslinked anhydride systems for bone augmentation applications. Biomaterials 28: 5259–5270. doi: 10.1016/j.biomaterials.2007.08.022
[98]	Hasturk H, Kantarci A, Ghattas M, et al. (2014) The use of light/chemically hardened polymethylmethacrylate, polyhydroxylethylmethacrylate, and calcium hydroxide graft material in combination with polyanhydride around implants and extraction sockets in minipigs: Part II: histologic and micro-CT evaluations. J Periodontol 85: 1230–1239. doi: 10.1902/jop.2014.120424
[99]	Uhrich KE (2010) Polyanhydride linkers for production of drug polymers and drug polymer compositions produced thereby. U.S. Patent.
[100]	Zhang Z, Kuijer R, Bulstra SK, et al. (2006) The in vivo and in vitro degradation behavior of poly(trimethylene carbonate). Biomaterials 27: 1741–1748. doi: 10.1016/j.biomaterials.2005.09.017
[101]	Habraken WJ, Zhang Z, Wolke JG, et al. (2008) Introduction of enzymatically degradable poly(trimethylene carbonate) microspheres into an injectable calcium phosphate cement. Biomaterials 29: 2464–2476. doi: 10.1016/j.biomaterials.2008.02.012
[102]	Kluin OS, van der Mei HC, Busscher HJ, et al. (2009) A surface-eroding antibiotic delivery system based on poly-(trimethylene carbonate). Biomaterials 30: 4738–4742. doi: 10.1016/j.biomaterials.2009.05.012
[103]	Bat E, Feijen J, Grijpma DW (2010) Biodegradable elastomeric networks: highly efficient cross-linking of poly(trimethylene carbonate) by gamma irradiation in the presence of pentaerythritol triacrylate. Biomacromolecules 11: 2692–2699. doi: 10.1021/bm1007234
[104]	Chen W, Meng F, Li F, et al. (2009) pH-responsive biodegradable micelles based on acid-labile polycarbonate hydrophobe: synthesis and triggered drug release. Biomacromolecules 10: 1727–1735. doi: 10.1021/bm900074d
[105]	Amsden BG, Timbart L, Marecak D, et al. (2010) VEGF-induced angiogenesis following localized delivery via injectable, low viscosity poly(trimethylene carbonate). J Control Release 145: 109–115. doi: 10.1016/j.jconrel.2010.03.029
[106]	Welle A, Kroger M, Doring M, et al. (2007) Electrospun aliphatic polycarbonates as tailored tissue scaffold materials. Biomaterials 28: 2211–2219. doi: 10.1016/j.biomaterials.2007.01.024
[107]	Suriano F, Pratt R, Tan JP, et al. (2010) Synthesis of a family of amphiphilic glycopolymers via controlled ring-opening polymerization of functionalized cyclic carbonates and their application in drug delivery. Biomaterials 31: 2637–2645. doi: 10.1016/j.biomaterials.2009.12.022
[108]	Pronych GJ, Sutow EJ, Sykora O (2003) Dimensional stability and dehydration of a thermoplastic polycarbonate-based and two PMMA-based denture resins. J Oral Rehabil 30: 1157–1161. doi: 10.1111/j.1365-2842.2003.01189.x
[109]	Tanimoto Y, Inami T, Yamaguchi M, et al. (2015) Preparation, mechanical, and in vitro properties of glass fiber-reinforced polycarbonate composites for orthodontic application. J Biomed Mater Res B 103: 743–750. doi: 10.1002/jbm.b.33245
[110]	Zhang W, Zhang Z, Chen S, et al. (2016) Mandibular jaw bone regeneration using human dental cell-seeded tyrosine-derived polycarbonate scaffolds. Tissue Eng A 22: 985–993. doi: 10.1089/ten.tea.2016.0166
[111]	Yilgor I, Yilgor E (2007) Structure‐morphology-property behavior of segmented thermoplastic polyurethanes and polyureas prepared without chain extenders. Polym Rev 47: 487–510. doi: 10.1080/15583720701638260
[112]	Asai T, Lee MH, Arrecubieta C, et al. (2007) Cellular coating of the left ventricular assist device textured polyurethane membrane reduces adhesion of Staphylococcus aureus. J Thorac Cardiov Sur 133: 1147–1153. doi: 10.1016/j.jtcvs.2006.10.084
[113]	Uttayarat P, Perets A, Li M, et al. (2010) Micropatterning of three-dimensional electrospun polyurethane vascular grafts. Acta Biomater 6: 4229–4237. doi: 10.1016/j.actbio.2010.06.008
[114]	Backman S, Bjorling G, Johansson UB, et al. (2009) Material wear of polymeric tracheostomy tubes: a six-month study. Laryngoscope 119: 657–664. doi: 10.1002/lary.20048
[115]	Lavorgna M, Cerruti P, Casula G, et al. (2007) Curing characteristics and mechanical properties of carbon fiber-interlayered fabric composites based on a polyurethane matrix. Adv Polym Tech 26: 132–145. doi: 10.1002/adv.20094
[116]	Kim JH, Kim KB, Kim WC, et al. (2014) Accuracy and precision of polyurethane dental arch models fabricated using a three-dimensional subtractive rapid prototyping method with an intraoral scanning technique. Korean J Orthod 44: 69–76. doi: 10.4041/kjod.2014.44.2.69
[117]	Mampieri G, Giancotti A (2013) Invisalign technique in the treatment of adults with pre-restorative concerns. Prog Orthod 14: 40–40. doi: 10.1186/2196-1042-14-40
[118]	Lee JH, Cho SA (2015) Altered polyurethane cast for a partial removable dental prosthesis. J Prosthet Dent 114: 305–306. doi: 10.1016/j.prosdent.2015.02.011
[119]	Selten MH, Broekema FI, Zuidema J, et al. (2013) Modified polyurethane foam as a local hemostatic agent after dental extractions. Ned Tijdschr Tandheelkd 120: 378–382.
[120]	Ohta T, Tani A, Kimbara K, et al. (2005) A novel nicotinoprotein aldehyde dehydrogenase involved in polyethylene glycol degradation. Appl Microbiol Biot 68: 639–646. doi: 10.1007/s00253-005-1936-z
[121]	Kabanov AV, Lemieux P, Vinogradov S, et al. (2002) Pluronic^® block copolymers: novel functional molecules for gene therapy. Adv Drug Deliver Rev 54: 223–233. doi: 10.1016/S0169-409X(02)00018-2
[122]	Aka-Any-Grah A, Bouchemal K, Koffi A, et al. (2010) Formulation of mucoadhesive vaginal hydrogels insensitive to dilution with vaginal fluids. Eur J Pharm Biopharm 76: 296–303. doi: 10.1016/j.ejpb.2010.07.004
[123]	Vashi AV, Keramidaris E, Abberton KM, et al. (2008) Adipose differentiation of bone marrow-derived mesenchymal stem cells using Pluronic F-127 hydrogel in vitro. Biomaterials 29: 573–579. doi: 10.1016/j.biomaterials.2007.10.017
[124]	Zhang X, He H, Yen C, et al. (2008) A biodegradable, immunoprotective, dual nanoporous capsule for cell-based therapies. Biomaterials 29: 4253–4259. doi: 10.1016/j.biomaterials.2008.07.032
[125]	Buxadera-Palomero J, Canal C, Torrent-Camarero S, et al. (2015) Antifouling coatings for dental implants: Polyethylene glycol-like coatings on titanium by plasma polymerization. Biointerphases 10: 029505. doi: 10.1116/1.4913376
[126]	Dabbagh A, Abu Kasim NH, Bakri MM, et al. (2014) Polyethylene-glycol coated maghemite nanoparticles for treatment of dental hypersensitivity. Mater Lett 121: 89–92. doi: 10.1016/j.matlet.2014.01.120
[127]	Mei M, Li QL, Chu C (2016) Inhibition of cariogenic plaque formation on root surface with polydopamine-induced-polyethylene glycol coating. Materials 9: 414. doi: 10.3390/ma9060414
[128]	Thoma DS, Jung UW, Park JY, et al. (2017) Bone augmentation at peri-implant dehiscence defects comparing a synthetic polyethylene glycol hydrogel matrix vs. standard guided bone regeneration techniques. Clin Oral Implan Res 28: e76–e83.
[129]	Walters NJ, Xia W, Salih V, et al. (2016) Poly(propylene glycol) and urethane dimethacrylates improve conversion of dental composites and reveal complexity of cytocompatibility testing. Dent Mater 32: 264–277. doi: 10.1016/j.dental.2015.11.017
[130]	Münchow EA, da Silva AF, da Silveira Lima G, et al. (2015) Polypropylene glycol phosphate methacrylate as an alternative acid-functional monomer on self-etching adhesives. J Dent 43: 94–102. doi: 10.1016/j.jdent.2014.11.005
[131]	Diniz IMA, Chen C, Xu X, et al. (2015) Pluronic F-127 hydrogel as a promising scaffold for encapsulation of dental-derived mesenchymal stem cells. J Mater Sci-Mater M 26: 153. doi: 10.1007/s10856-015-5493-4
[132]	Mogen AB, Chen F, Ahn SJ, et al. (2015) Pluronics-formulated farnesol promotes efficient killing and demonstrates novel interactions with Streptococcus mutans biofilms. PLoS One 10: e0133886. doi: 10.1371/journal.pone.0133886
[133]	Ramshaw JA, Shah NK, Brodsky B (1998) Gly-X-Y tripeptide frequencies in collagen: a context for host-guest triple-helical peptides. J Struct Biol 122: 86–91. doi: 10.1006/jsbi.1998.3977
[134]	Krane SM (2008) The importance of proline residues in the structure, stability and susceptibility to proteolytic degradation of collagens. Amino Acids 35: 703–710. doi: 10.1007/s00726-008-0073-2
[135]	Yilmaz N, Inal S, Muglali M, et al. (2010) Effects of polyglecaprone 25, silk and catgut suture materials on oral mucosa wound healing in diabetic rats: an evaluation of nitric oxide dynamics. Med Oral Patol Oral 15: e526–e530.
[136]	Friberg O, Dahlin LG, Kallman J, et al. (2009) Collagen-gentamicin implant for prevention of sternal wound infection; long-term follow-up of effectiveness. Interact Cardiov Th 9: 454–458. doi: 10.1510/icvts.2009.207514
[137]	Holladay C, Keeney M, Greiser U, et al. (2009) A matrix reservoir for improved control of non-viral gene delivery. J Control Release 136: 220–225. doi: 10.1016/j.jconrel.2009.02.006
[138]	Vinas-Castells R, Holladay C, di Luca A, et al. (2009) Snail1 down-regulation using small interfering RNA complexes delivered through collagen scaffolds. Bioconjugate Chem 20: 2262–2269. doi: 10.1021/bc900241w
[139]	Maehara H, Sotome S, Yoshii T, et al. (2010) Repair of large osteochondral defects in rabbits using porous hydroxyapatite/collagen (HAp/Col) and fibroblast growth factor-2 (FGF-2). J Orthop Res 28: 677–686.
[140]	Baik SH, Kim JH, Cho HH, et al. (2010) Development and analysis of a collagen-based hemostatic adhesive. J Surg Res 164: e221–e228. doi: 10.1016/j.jss.2010.08.004
[141]	Wang L, Stegemann JP (2010) Thermogelling chitosan and collagen composite hydrogels initiated with beta-glycerophosphate for bone tissue engineering. Biomaterials 31: 3976–3985. doi: 10.1016/j.biomaterials.2010.01.131
[142]	Yan LP, Wang YJ, Ren L, et al. (2010) Genipin-cross-linked collagen/chitosan biomimetic scaffolds for articular cartilage tissue engineering applications. J Biomed Mater Res A 95: 465–475.
[143]	Casper CL, Yang W, Farach-Carson MC, et al. (2007) Coating electrospun collagen and gelatin fibers with perlecan domain I for increased growth factor binding. Biomacromolecules 8: 1116–1123. doi: 10.1021/bm061003s
[144]	Ciardelli G, Gentile P, Chiono V, et al. (2010) Enzymatically crosslinked porous composite matrices for bone tissue regeneration. J Biomed Mater Res A 92: 137–151.
[145]	Kim JY, Xin X, Moioli EK, et al. (2010) Regeneration of dental-pulp-like tissue by chemotaxis-induced cell homing. Tissue Eng A 16: 3023–3031. doi: 10.1089/ten.tea.2010.0181
[146]	Tatakis DN, Promsudthi A, Wikesjö UME (1999) Devices for periodontal regeneration. Periodontol 2000 19: 59–73. doi: 10.1111/j.1600-0757.1999.tb00147.x
[147]	Chatzistavrou X, Rao RR, Caldwell DJ, et al. (2016) Collagen/fibrin microbeads as a delivery system for Ag-doped bioactive glass and DPSCs for potential applications in dentistry. J Non-Cryst Solids 432: 143–149. doi: 10.1016/j.jnoncrysol.2015.03.024
[148]	Kothapalli CR, Ramamurthi A (2010) Induced elastin regeneration by chronically activated smooth muscle cells for targeted aneurysm repair. Acta Biomater 6: 170–178. doi: 10.1016/j.actbio.2009.06.006
[149]	Mithieux SM, Rasko JE, Weiss AS (2004) Synthetic elastin hydrogels derived from massive elastic assemblies of self-organized human protein monomers. Biomaterials 25: 4921–4927. doi: 10.1016/j.biomaterials.2004.01.055
[150]	Kaufmann D, Weberskirch R (2006) Efficient synthesis of protein-drug conjugates using a functionalizable recombinant elastin-mimetic polypeptide. Macromol Biosci 6: 952–958. doi: 10.1002/mabi.200600117
[151]	Tu Y, Wise SG, Weiss AS (2010) Stages in tropoelastin coalescence during synthetic elastin hydrogel formation. Micron 41: 268–272. doi: 10.1016/j.micron.2009.11.003
[152]	Gurumurthy B, Bierdeman PC, Janorkar AV (2016) Composition of elastin like polypeptide–collagen composite scaffold influences in vitro osteogenic activity of human adipose derived stem cells. Dent Mater 32: 1270–1280. doi: 10.1016/j.dental.2016.07.009
[153]	Amruthwar SS, Janorkar AV (2013) In vitro evaluation of elastin-like polypeptide–collagen composite scaffold for bone tissue engineering. Dent Mater 29: 211–220. doi: 10.1016/j.dental.2012.10.003
[154]	Shaikh FM, Callanan A, Kavanagh EG, et al. (2008) Fibrin: a natural biodegradable scaffold in vascular tissue engineering. Cells Tissues Organs 188: 333–346. doi: 10.1159/000139772
[155]	Choukroun J, Diss A, Simonpieri A, et al. (2006) Platelet-rich fibrin (PRF): a second-generation platelet concentrate. Part IV: clinical effects on tissue healing. Oral Surg Oral Med O 101: e56–60.
[156]	Li Q, Pan S, Dangaria SJ, et al. (2013) Platelet-rich fibrin promotes periodontal regeneration and enhances alveolar bone augmentation. BioMed Res Int 2013: 638043.
[157]	Marenzi G, Riccitiello F, Tia M, et al. (2015) Influence of leukocyte- and platelet-rich fibrin (L-PRF) in the healing of simple postextraction sockets: a split-mouth study. BioMed Res Int 2015: 369273.
[158]	Keswani D, Pandey RK (2013) Revascularization of an immature tooth with a necrotic pulp using platelet-rich fibrin: a case report. Int Endod J 46: 1096–1104.
[159]	Zhao YH, Zhang M, Liu NX, et al. (2013) The combined use of cell sheet fragments of periodontal ligament stem cells and platelet-rich fibrin granules for avulsed tooth reimplantation. Biomaterials 34: 5506–5520. doi: 10.1016/j.biomaterials.2013.03.079
[160]	Dohan Ehrenfest DM, Del Corso M, Diss A, et al. (2010) Three-dimensional architecture and cell composition of a Choukroun's platelet-rich fibrin clot and membrane. J Periodontol 81: 546–555. doi: 10.1902/jop.2009.090531
[161]	Portilla-Arias JA, Camargo B, Garcia-Alvarez M, et al. (2009) Nanoparticles made of microbial poly(gamma-glutamate)s for encapsulation and delivery of drugs and proteins. J Biomat Sci Polym E 20: 1065–1079. doi: 10.1163/156856209X444420
[162]	Takehara M, Hibino A, Saimura M, et al. (2010) High-yield production of short chain length poly(epsilon-L-lysine) consisting of 5–20 residues by Streptomyces aureofaciens, and its antimicrobial activity. Biotechnol Lett 32: 1299–1303. doi: 10.1007/s10529-010-0294-9
[163]	Tang CK, Sheng KC, Pouniotis D, et al. (2008) Oxidized and reduced mannan mediated MUC1 DNA immunization induce effective anti-tumor responses. Vaccine 26: 3827–3834. doi: 10.1016/j.vaccine.2008.05.008
[164]	Kim SS, Seo M, Chung JW, et al. (2009) Polymeric calcium phosphate cements incorporated with poly-γ-glutamic acid: Comparative study of poly-γ-glutamic acid and citric acid. J Appl Polym Sci 113: 1223–1231. doi: 10.1002/app.30037
[165]	Varoni E, Canciani E, Palazzo B, et al. (2015) Effect of Poly-L-Lysine coating on titanium osseointegration: from characterization to in vivo studies. J Oral Implantol 41: 626–631. doi: 10.1563/AAID-JOI-D-13-00036
[166]	Walters NJ, Palmer G, Ashley PF, et al. (2015) Monocalcium phosphate induces greater hydroxyapatite mineral formation than tricalcium phosphate on dental composites containing ε-poly-L-lysine. TERMIS-WC 2015, Proceedings of Tissue Engineering & Regenerative Medicine International Society World Congress, Boston, MA, USA.
[167]	Chen L, Tian H, Chen J, et al. (2010) Multi-armed poly(L-glutamic acid)-graft-oligoethylenimine copolymers as efficient nonviral gene delivery vectors. J Gene Med 12: 64–76. doi: 10.1002/jgm.1405
[168]	Pilbat AM, Ball V, Schaaf P, et al. (2006) Partial poly(glutamic acid) ↔ poly(aspartic acid) exchange in layer-by-layer polyelectrolyte films. Structural alterations in the three-component architectures. Langmuir 22: 5753–5759.
[169]	Karal-Yilmaz O, Kayaman-Apohan N, Misirli Z, et al. (2006) Synthesis and characterization of poly(L-lactic acid-co-ethylene oxide-co-aspartic acid) and its interaction with cells. J Mater Sci-Mater M 17: 213–227. doi: 10.1007/s10856-006-7307-1
[170]	Benkirane-Jessel N, Palomares CM, Fioretti F (2014) Compound comprising alpha-msh for use in endodontic regeneration. U.S. Patent.
[171]	Osorio R, Sauro S (2015) Polyaspartic acid enhances dentine remineralization bonded with a zinc-doped Portland-based resin cement. Int Endod J 49: 874–883.
[172]	Wang X, Ji J (2009) Postdiffusion of oligo-peptide within exponential growth multilayer films for localized peptide delivery. Langmuir 25: 11664–11671. doi: 10.1021/la9013575
[173]	Pasquinelli G, Orrico C, Foroni L, et al. (2008) Mesenchymal stem cell interaction with a non-woven hyaluronan-based scaffold suitable for tissue repair. J Anat 213: 520–530. doi: 10.1111/j.1469-7580.2008.00974.x
[174]	Temiz A, Kazikdas KC, Ergur B, et al. (2010) Esterified hyaluronic acid improves cartilage viability in experimental tracheal reconstruction with an auricular graft. Otolaryng Head Neck 143: 772–778. doi: 10.1016/j.otohns.2010.07.007
[175]	Grigolo B, Lisignoli G, Desando G, et al. (2009) Osteoarthritis treated with mesenchymal stem cells on hyaluronan-based scaffold in rabbit. Tissue Eng C 15: 647–658. doi: 10.1089/ten.tec.2008.0569
[176]	Pandis L, Zavan B, Abatangelo G, et al. (2010) Hyaluronan-based scaffold for in vivo regeneration of the rat vena cava: Preliminary results in an animal model. J Biomed Mater Res A 93: 1289–1296.
[177]	Zavan B, Abatangelo G, Mazzoleni F, et al. (2008) New 3D hyaluronan-based scaffold for in vitro reconstruction of the rat sciatic nerve. Neurol Res 30: 190–196. doi: 10.1179/174313208X281082
[178]	El-Sayed KMF, Dahaba MA, Aboul-Ela S, et al. (2012) Local application of hyaluronan gel in conjunction with periodontal surgery: a randomized controlled trial. Clin Oral Invest 16: 1229–1236. doi: 10.1007/s00784-011-0630-z
[179]	Gontiya G, Galgali SR (2012) Effect of hyaluronan on periodontitis: A clinical and histological study. J Indian Soc Periodontol 16: 184–192. doi: 10.4103/0972-124X.99260
[180]	Chrepa V, Austah O, Diogenes A (2017) Evaluation of a commercially available hyaluronic acid hydrogel (restylane) as injectable scaffold for dental pulp regeneration: an in vitro evaluation. J Endodont 43: 257–262. doi: 10.1016/j.joen.2016.10.026
[181]	Ferroni L, Gardin C, Sivolella S, et al. (2015) A hyaluronan-based scaffold for the in vitro construction of dental pulp-like tissue. Int J Mol Sci 16: 4666. doi: 10.3390/ijms16034666
[182]	Tan L, Wang J, Yin S, et al. (2015) Regeneration of dentin-pulp-like tissue using an injectable tissue engineering technique. RSC Adv 5: 59723–59737. doi: 10.1039/C5RA06481C
[183]	Xia W, Liu P, Liu J (2008) Advance in chitosan hydrolysis by non-specific cellulases. Bioresource Technol 99: 6751–6762. doi: 10.1016/j.biortech.2008.01.011
[184]	Li DH, Liu LM, Tian KL, et al. (2007) Synthesis, biodegradability and cytotoxicity of water-soluble isobutylchitosan. Carbohyd Polym 67: 40–45. doi: 10.1016/j.carbpol.2006.04.022
[185]	Park CJ, Gabrielson NP, Pack DW, et al. (2009) The effect of chitosan on the migration of neutrophil-like HL60 cells, mediated by IL-8. Biomaterials 30: 436–444. doi: 10.1016/j.biomaterials.2008.09.060
[186]	Burkatovskaya M, Tegos GP, Swietlik E, et al. (2006) Use of chitosan bandage to prevent fatal infections developing from highly contaminated wounds in mice. Biomaterials 27: 4157–4164. doi: 10.1016/j.biomaterials.2006.03.028
[187]	Wu L, Li H, Li S, et al. (2010) Composite fibrous membranes of PLGA and chitosan prepared by coelectrospinning and coaxial electrospinning. J Biomed Mater Res A 92: 563–574.
[188]	Azab AK, Orkin B, Doviner V, et al. (2006) Crosslinked chitosan implants as potential degradable devices for brachytherapy: in vitro and in vivo analysis. J Control Release 111: 281–289. doi: 10.1016/j.jconrel.2005.12.014
[189]	Wang J, Feng SS, Wang S, et al. (2010) Evaluation of cationic nanoparticles of biodegradable copolymers as siRNA delivery system for hepatitis B treatment. Int J Pharmaceut 400: 194–200. doi: 10.1016/j.ijpharm.2010.08.026
[190]	Sarukawa J, Takahashi M, Abe M, et al. (2011) Effects of chitosan-coated fibers as a scaffold for three-dimensional cultures of rabbit fibroblasts for ligament tissue engineering. J Biomat Sci Polym E 22: 717–732. doi: 10.1163/092050610X491067
[191]	Busscher HJ, Engels E, Dijkstra RJB, et al. (2008) Influence of a chitosan on oral bacterial adhesion and growth in vitro. Eur J Oral Sci 116: 493–495. doi: 10.1111/j.1600-0722.2008.00568.x
[192]	Barreras US, Mendez FT, Martinez RE, et al. (2016) Chitosan nanoparticles enhance the antibacterial activity of chlorhexidine in collagen membranes used for periapical guided tissue regeneration. Mat Sci Eng C-Mater 58: 1182–1187. doi: 10.1016/j.msec.2015.09.085
[193]	Elsaka SE (2012) Antibacterial activity and adhesive properties of a chitosan-containing dental adhesive. Quintessence Int 43: 603–613.
[194]	Mota J, Yu N, Caridade SG, et al. (2012) Chitosan/bioactive glass nanoparticle composite membranes for periodontal regeneration. Acta Biomater 8: 4173–4180. doi: 10.1016/j.actbio.2012.06.040
[195]	Tan CS, Jejurikar A, Rai B, et al. (2009) Encapsulation of a glycosaminoglycan in hydroxyapatite/alginate capsules. J Biomed Mater Res A 91: 866–877.
[196]	Liu XY, Nothias JM, Scavone A, et al. (2010) Biocompatibility investigation of polyethylene glycol and alginate-poly-L-lysine for islet encapsulation. ASAIO J 56: 241–245. doi: 10.1097/MAT.0b013e3181d7b8e3
[197]	Giovagnoli S, Tsai T, DeLuca PP (2010) Formulation and release behavior of doxycycline-alginate hydrogel microparticles embedded into pluronic F127 thermogels as a potential new vehicle for doxycycline intradermal sustained delivery. AAPS PharmSciTech 11: 212–220. doi: 10.1208/s12249-009-9361-8
[198]	Jay SM, Shepherd BR, Andrejecsk JW, et al. (2010) Dual delivery of VEGF and MCP-1 to support endothelial cell transplantation for therapeutic vascularization. Biomaterials 31: 3054–3062. doi: 10.1016/j.biomaterials.2010.01.014
[199]	Qi J, Chen A, You H, et al. (2011) Proliferation and chondrogenic differentiation of CD105-positive enriched rat synovium-derived mesenchymal stem cells in three-dimensional porous scaffolds. Biomed Mater 6: 015006. doi: 10.1088/1748-6041/6/1/015006
[200]	Meng X, Li P, Wei Q, et al. (2011) pH sensitive alginate-chitosan hydrogel beads for carvedilol delivery. Pharm Dev Technol 16: 22–28. doi: 10.3109/10837450903479947
[201]	Wittmer CR, Phelps JA, Lepus CM, et al. (2008) Multilayer nanofilms as substrates for hepatocellular applications. Biomaterials 29: 4082–4090. doi: 10.1016/j.biomaterials.2008.06.027
[202]	Lee JW, Park YJ, Lee SJ, et al. (2010) The effect of spacer arm length of an adhesion ligand coupled to an alginate gel on the control of fibroblast phenotype. Biomaterials 31: 5545–5551. doi: 10.1016/j.biomaterials.2010.03.063
[203]	Moshaverinia A, Chen C, Xu X, et al. (2014) Bone regeneration potential of stem cells derived from periodontal ligament or gingival tissue sources encapsulated in RGD-modified alginate scaffold. Tissue Eng A 20: 611–621. doi: 10.1089/ten.tec.2013.0408
[204]	Srinivasan S, Jayasree R, Chennazhi KP, et al. (2012) Biocompatible alginate/nano bioactive glass ceramic composite scaffolds for periodontal tissue regeneration. Carbohyd Polym 87: 274–283. doi: 10.1016/j.carbpol.2011.07.058
[205]	Mattamal GJ (2008) US FDA perspective on the regulations of medical-grade polymers: cyanoacrylate polymer medical device tissue adhesives. Expert Rev Med Devic 5: 41–49. doi: 10.1586/17434440.5.1.41

This article has been cited by:

1.	Jonathan Kabala Mbuya, Dieter Pfoser, Antonios Anastasopoulos, 2024, Trajectory Anomaly Detection with Language Models, 9798400711077, 208, 10.1145/3678717.3691257
2.	Mingqi Qiu, Shuhua Mao, Jiangbin Zhu, Yingjie Yang, Spatiotemporal multi-feature fusion vehicle trajectory anomaly detection for intelligent transportation: An improved method combining autoencoders and dynamic Bayesian networks, 2025, 211, 00014575, 107911, 10.1016/j.aap.2024.107911

Reader Comments

Your name:*

Email:*
© 2018 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)

通讯作者: 陈斌, bchen63@163.com

1.
沈阳化工大学材料科学与工程学院沈阳 110142

AIMS Materials Science

1.8 3.7

Metrics

Article views(10130) PDF downloads(1843) Cited by(26)

Preview PDF

Download XML

Export Citation

Article outline

Show full outline

Figures and Tables

Figures(4) / Tables(2)

AIMS Materials Science

Biodegradable polymers in dental tissue engineering and regeneration

Related Papers:

Abstract

1. Introduction

2. Related work

3. Definitions for distances and VAE

3.1. Trajectory and distance

3.2. Variational auto-encoder

4. Trajectory outlier detection method based on variational auto-encoder

4.1. Model

4.1.1. Preprocessing of trajectory data

4.1.2. VAE network training

4.1.3. Trajectory outlier detection

5. Experiments and analysis

5.1. Experimental environment and dataset processing

5.2. Evaluation metrics

5.3. Analysis of model generalization effect

5.3.1. Influence of training parameter epoch

5.3.2. Influence of training set size train_size

5.3.3. Influence of threshold λ in outlier detection algorithm

5.4. Effect comparison

5.5. Efficiency analysis

6. Conclusions

Use of AI tools declaration

Acknowledgments

Conflict of interest

References

This article has been cited by:

Reader Comments

通讯作者: 陈斌, bchen63@163.com

Metrics

Figures and Tables

Other Articles By Authors

Catalog

AIMS Materials Science

Biodegradable polymers in dental tissue engineering and regeneration

Related Papers:

Abstract

1. Introduction

2. Related work

3. Definitions for distances and VAE

3.1. Trajectory and distance

3.2. Variational auto-encoder

4. Trajectory outlier detection method based on variational auto-encoder

4.1. Model

4.1.1. Preprocessing of trajectory data

4.1.2. VAE network training

4.1.3. Trajectory outlier detection

5. Experiments and analysis

5.1. Experimental environment and dataset processing

5.2. Evaluation metrics

5.3. Analysis of model generalization effect

5.3.1. Influence of training parameter epoch

5.3.2. Influence of training set size train_size

5.3.3. Influence of threshold λ in outlier detection algorithm

5.4. Effect comparison

5.5. Efficiency analysis

6. Conclusions

Use of AI tools declaration

Acknowledgments

Conflict of interest

References

This article has been cited by:

Reader Comments

通讯作者: 陈斌, bchen63@163.com

Metrics

Figures and Tables

Other Articles By Authors

Related pages

Tools

Export File

Citation

Format

Content

Catalog