The greedy block Kaczmarz (GBK) method has been successfully applied in areas such as data mining, image reconstruction, and large-scale image restoration. However, the computation of pseudo-inverses in each iterative step of the GBK method not only complicates the computation and slows down the convergence rate, but it is also ill-suited for distributed implementation. The leverage score sampling free pseudo-inverse GBK algorithm proposed in this paper demonstrated significant potential in the field of image reconstruction. By ingeniously transforming the problem framework, the algorithm not only enhanced the efficiency of processing systems of linear equations with multiple solution vectors but also optimized specifically for applications in image reconstruction. A methodology that combined theoretical and experimental approaches has validated the robustness and practicality of the algorithm, providing valuable insights for technical advancements in related disciplines.
Citation: Wenya Shi, Xinpeng Yan, Zhan Huan. Faster free pseudoinverse greedy block Kaczmarz method for image recovery[J]. Electronic Research Archive, 2024, 32(6): 3973-3988. doi: 10.3934/era.2024178
[1] | Valeria Paradies, Filippo Masi, Francesco Bartolomucci, Alfredo Marchese, Armando Liso, Fabrizio Resta, Stefano Favale, Martino Pepe . No Reflow-phenomenon: from Current State of the Art to Future Perspectives. AIMS Medical Science, 2015, 2(4): 374-395. doi: 10.3934/medsci.2015.4.374 |
[2] | Feldberg Carolina, Hermida Paula D, Maria Florencia Tartaglini, Stefani Dorina, Somale Verónica, Allegri Ricardo F . Cognitive Reserve in Patients with Mild Cognitive Impairment: The Importance of Occupational Complexity as a Buffer of Declining Cognition in Older Adults. AIMS Medical Science, 2016, 3(1): 77-95. doi: 10.3934/medsci.2016.1.77 |
[3] | Luis Miguel Juárez-Salcedo, Luis Manuel González, Samir Dalia . Immunotherapy for diffuse large B-cell lymphoma: current use of immune checkpoint inhibitors therapy. AIMS Medical Science, 2023, 10(3): 259-272. doi: 10.3934/medsci.2023020 |
[4] | Emma Altobelli, Maurizio Amichetti, Alessio Langiu, Francesca Marzi, Filippo Mignosi, Pietro Pisciotta, Giuseppe Placidi, Fabrizio Rossi, Giorgio Russo, Marco Schwarz, Stefano Smriglio, Sabina Vennarini . Combinatorial optimisation in radiotherapy treatment planning. AIMS Medical Science, 2018, 5(3): 204-223. doi: 10.3934/medsci.2018.3.204 |
[5] | Ying Chen, Sok Kean Khoo, Richard Leach, Kai Wang . MTA3 Regulates Extravillous Trophoblast Invasion Through NuRD Complex. AIMS Medical Science, 2017, 4(1): 17-27. doi: 10.3934/medsci.2017.1.17 |
[6] | Joseph L. Amaral, Veronica Romero, Heidi Kloos, Michael J. Richardson . Complexity and Autism Spectrum Disorder: Exploring Hysteresis in A Grasping Task. AIMS Medical Science, 2017, 4(1): 113-130. doi: 10.3934/medsci.2017.1.113 |
[7] | Heliya Bandehagh, Farnaz Gozalpour, Ali Mousavi, Mahdi Hemmati Ghavshough . Effects of melatonin on the management of multiple sclerosis: A scoping review on animal studies. AIMS Medical Science, 2024, 11(2): 137-156. doi: 10.3934/medsci.2024012 |
[8] | Renjith Parameswaran Nair, Gulshan Sunavala-Dossabhoy . Promising Gene Therapeutics for Salivary Gland Radiotoxicity. AIMS Medical Science, 2016, 3(4): 329-344. doi: 10.3934/medsci.2016.4.329 |
[9] | Ilana S. Golub, Angela Misic, Lucia P. Schroeder, Jairo Aldana-Bitar, Srikanth Krishnan, Sina Kianoush, Travis Benzing, Keishi Ichikawa, Matthew J. Budoff . Calcific coronary lesions: management, challenges, and a comprehensive review. AIMS Medical Science, 2024, 11(3): 292-317. doi: 10.3934/medsci.2024021 |
[10] | Benjamin P Jones, Srdjan Saso, Timothy Bracewell-Milnes, Jen Barcroft, Jane Borley, Teodor Goroszeniuk, Kostas Lathouras, Joseph Yazbek, J Richard Smith . Laparoscopic uterosacral nerve block: A fertility preserving option in chronic pelvic pain. AIMS Medical Science, 2019, 6(4): 260-267. doi: 10.3934/medsci.2019.4.260 |
The greedy block Kaczmarz (GBK) method has been successfully applied in areas such as data mining, image reconstruction, and large-scale image restoration. However, the computation of pseudo-inverses in each iterative step of the GBK method not only complicates the computation and slows down the convergence rate, but it is also ill-suited for distributed implementation. The leverage score sampling free pseudo-inverse GBK algorithm proposed in this paper demonstrated significant potential in the field of image reconstruction. By ingeniously transforming the problem framework, the algorithm not only enhanced the efficiency of processing systems of linear equations with multiple solution vectors but also optimized specifically for applications in image reconstruction. A methodology that combined theoretical and experimental approaches has validated the robustness and practicality of the algorithm, providing valuable insights for technical advancements in related disciplines.
A 38-year-old librarian with confirmed Carney complex (PRKAR1a mutation) was referred for further evaluation of ACTH-independent Cushing’s syndrome. Previously, she was known to have schwannoma (excised), adenomyoepithelioma and normal annual echocardiograms. She had a borderline high prolactin of 789 mIU (normal range < 500 mIU) with regular menstrual cycle and magnetic resonance imaging of the pituitary suggestive of 4 mm microadenoma. Over three years prior to current presentation, she had become aware of coarse hair on her chin and abdomen, as well as centripetal weight gain. She had noticed subtle but definite reduction in her girdle muscle strength. She had acquired some mood changes atypical of her personality, and had developed an interrupted sleep pattern.
On examination she had the classic dermatological signs associated with the Carney complex. Her BMI was just over the normal Caucasian range, with minimal increase in abdominal girth (relative to her), small amount of dorsocervical fat and mild hirsutism. She did not however show evidence of a typical cushingoid phenotype. Her blood pressure was 158/108.
Twenty-four hour urinary free cortisol measurements on three occasions six months apart were 63,176 and 96 nmol/L (normal range < 120 nmol/L). Serum total cortisol prior to and after low dose dexamethasone suppression was 292 and 357 nmol/L, and after high dose dexamethasone suppression test was 321 and 457 nmol/L, respectively (normal < 50 nmol/L post suppression). Adrenocorticotropic hormone (ACTH) prior to dexamethasone administration was undetectable. Salivary cortisol levels on three separate mornings on waking were all 4 nmol/L, and on the preceding nights were 4, 3 and 4 nmol/L. Computed tomography performed for ACTH-independent Cushing’s syndrome was reported to show two normal adrenal glands.
Carney complex or Swiss syndrome [1], a rare condition of autosomal dominant inheritance, has a varied presentation that involves tumours of multiple organ systems including the skin, heart, usually more than one endocrine organ including the adrenal, pituitary and thyroid glands; male and female reproductive organs, breast tissue and neural tissue [2]. Classically it was known as a complex of spotty skin pigmentation, myxomas, endocrine overactivity and schwannomas [3]. The presence of specific diagnostic criteria in a given patient must prompt consideration of the condition [4]. Most of the individual tumors appear to be benign on histopathological examination [2], although recently a case of adrenal cancer has been described [5]. Cutaneous manifestations may include pigmented lesions unrelated to sun exposure and/or myxomas usually involving the face and upper trunk but may occur elsewhere and tend to develop early in life [2]. Cardiac myxomas may be single or multifocal and may have fatal consequences if left untreated [2]. The endocrinopathies associated with this condition are often multiple, and include non-ACTH dependent hypercortisolism, growth hormone- or prolactin- secreting (usually asymptomatic) pituitary adenomas, thyroid adenomas and rarely carcinomas, large-cell calcifying Sertoli cell tumours of the testes and ovarian cysts [4]. The genetic defect in PRKAR1a located on chromosome 17 is seen in approximately 40% of kindreds studied in a large case series [4].
Primary pigmented nodular adrenocortical disease (PPNAD) most commonly seen in this context, describes the histologic appearance of adrenal glands with small-pigmented nodules interspersed between usually atrophic cortex [2] and is believed to be almost universal [2,4]. It manifests as autonomous ACTH-independent hypercortisolism, which may remain subclinical, be cyclic, or present with the Cushing phenotype [2] and rarely, may regress spontaneously [6]. It generally manifests in the second decade of life, rarely diagnosed after the fifth. Laboratory diagnosis can be challenging as it takes years to develop. Tests used conventionally may be confirmatory as in any other non-ACTH dependent Cushing’s syndrome, such as elevated urinary free cortisol, failure of dexamethasone to suppress cortisol, with no response of cortisol or ACTH to corticotropin-releasing hormone. 6-day Liddle test leading to a paradoxical increase of 50% or more in urine free cortisol excretion has been recommended as the preferred test in this group of patients, although the mechanism is poorly understood [7]. Imaging of adrenal glands is often unhelpful.
The classic pattern of HPA activity in PPNAD is complete loss of circadian rhythm of cortisol, with the absolute levels generally within or just over the normal range [8]. Hence a physiological approach to treatment would be to lower cortisol levels in the hours immediately surrounding sleep in an effort to restore normality of the circadian rhythm. The recommended treatment for PPNAD is bilateral adrenalectomy [8], but although it can remit physical features [9], it has significant morbidity and mortality [10]. Of the therapeutic agents available, a relatively short-acting option to reduce cortisol levels or block its effects during the latter part of the evening and early sleep hours would potentially be able to achieve this. The likely candidates are adrenal inhibitors (e.g. Ketoconazole, Metyrapone, Aminoglutethimide, Mitotane) or the glucocorticoid receptor antagonist Mifepristone. Adrenal inhibitors are not without significant side effects [9]. Mifepristone (RU38486 or RU486) is a promising alternative that has been used less frequently in the treatment of Cushing’s syndrome and although it has fewer side effects [11], biochemical monitoring of cortisol levels during treatment is unreliable [12].
We chose to treat our patient with low dose Mifepristone (200 mg) in the evening, which she has been taking for the past 6 months. Within weeks, her sleep pattern had normalised. She used to be wide-awake after 2-3 hours of sleep resulting in fatigue and lower productivity during the day. She now has more energy and interest in day-to-day activities. Her irritability has abated, and her calm persona restored. She has lost 6kg in weight within these 6 months without any lifestyle modification, and more importantly for her the fat around her abdomen has reduced. There is a marked subjective improvement in the strength of shoulder and hip muscles. Her blood pressure at the last clinic visit had fallen to 140/91. She has reported no side-effects during subsequent visits. She has been counselled to avoid pregnancy whilst on treatment, and she has no current plans.
To our knowledge, this is the first published report of circadian RU486 therapy for PPNAD in a patient with Carney complex. It may have been possible to restore low levels surrounding the midnight hours using other agents, but the side-effect profile and lack of significantly elevated levels of cortisol made them less favourable options. With Mifepristone, it was possible to switch-off the glucocorticoid receptor mediated transcriptional activity at night to mimic the normal nocturnal fall in HPA activity. It is the lack of this nocturnal nadir of glucocorticoid activity with continuous activation of the glucocorticoid receptor throughout the 24 hours that is a major factor in the aetiology of glucocorticoid-mediated symptoms [13]. Monitoring the efficacy of treatment is by clinical review including history and clinical examination. Since daytime glucocorticoid activity is not compromised due to the short half-life of mifepristone, there is little danger of hypoadrenal complications. We propose this circadian therapy with mifepristone as a novel approach to treatment of PPNAD with autonomous ACTH-independent hypercortisolism, by restoring circadian rhythm with low dose RU486 administration in the evening. Its potential effectiveness in other secretory adrenal incidentalomas may also be worth considering.
The authors claim no conflict of interest.
[1] | A. C. Kak, M. Slaney, Principles of computerized tomographic imaging, Society for Industrial and Applied Mathematics, New York, 2001. |
[2] |
J. A. Fessler, B. P. Sutton, Nonuniform fast Fourier transforms using min-max interpolation, IEEE T. Signal Proces., 51 (2003), 560–574. https://doi.org/10.1109/TSP.2002.807005 doi: 10.1109/TSP.2002.807005
![]() |
[3] |
S. F. Gull, G. J. Daniell, Image reconstruction from incomplete and noisy data, Nature, 272 (1978), 686–690. https://doi.org/10.1038/272686a0 doi: 10.1038/272686a0
![]() |
[4] |
X. L. Zhao, T. Z. Huang, X. M. Gu, L. J. Deng, Vector extrapolation based Landweber method for discrete ill-posed problems, Math. Probl. Eng., 2017 (2017), 1375716. https://doi.org/10.1155/2017/1375716 doi: 10.1155/2017/1375716
![]() |
[5] |
S. C. Park, M. K. Park, M. G. Kang, Super-resolution image reconstruction: a technical overview, IEEE Signal Proc. Mag., 20 (2003), 21–36. https://doi.org/10.1109/MSP.2003.120320 doi: 10.1109/MSP.2003.120320
![]() |
[6] | F. Natterer, F. Wübbeling, Mathematical methods in image reconstruction, Society for Industrial and Applied Mathematics, New York, 2001. |
[7] |
G. L. Zeng, Image reconstruction—a tutorial, Comput. Med. Imag. Grap., 25 (2001), 97–103. https://doi.org/10.1016/S0895-6111(00)00059-8 doi: 10.1016/S0895-6111(00)00059-8
![]() |
[8] | J. I. Goldstein, D. E. Newbury, J. R. Michael, N. W. M. Ritchie, J. H. J. Scott, D. C. Joy, X-Rays, in Scanning Electron Microscopy and X-Ray Microanalysis, Springer, (2018), 39–63. https://doi.org/10.1007/978-1-4939-6676-9_4 |
[9] | P. Toft, The Radon Transform-Theory and Implementation, Ph.D thesis, Technical University of Denmark in Copenhagen, 1996. |
[10] | A. Rahman, System of linear equations in Computed Tomography (CT), Bachelor's thesis, Brac University in Dacca, 2018. |
[11] | G. T. Herman, Fundamentals of Computerized Tomography: Image Reconstruction from Projections, Springer Science & Business Media, Berlin, 2009. |
[12] |
Y. Censor, G. T. Herman, M. Jiang, A note on the behavior of the randomized Kaczmarz algorithm of Strohmer and Vershynin, J. Fourier Anal. Appl., 15 (2009), 431–436. https://doi.org/10.1007/s00041-009-9077-x doi: 10.1007/s00041-009-9077-x
![]() |
[13] | O. Axelsson, Iterative Solution Methods, Cambridge University Press, Cambridge, 1996. |
[14] | I. Gohberg, I. A. Fel_dman, Convolution Equations and Projection Methods for Their Solution, American Mathematical Soc., Providence, 2005. |
[15] | Z. Z. Bai, C. H. Jin, Column-decomposed relaxation methods for the overdetermined systems of linear equations, Int. J. Appl. Mat. Com.-Pol., 13 (2003), 71–82. |
[16] | S. Kaczmarz, Angenäherte auflösung von systemen linearer glei-chungen, Bull. Int. Acad. Pol. Sci. Lett. Class. Sci. Math. Nat., (1937), 355–357. |
[17] | M. Brooks, A Survey of Algebraic Algorithms in Computerized Tomography, Master's Thesis, University of Ontario Institute of Technology in Oshawa, 2010. |
[18] |
Y. Censor, Parallel application of block-iterative methods in medical imaging and radiation therapy, Math. Program., 42 (1988), 307–325. https://doi.org/10.1007/BF01589408 doi: 10.1007/BF01589408
![]() |
[19] |
C. Byrne, A unified treatment of some iterative algorithms in signal processing and image reconstruction, Inverse Probl., 20 (20031), 103. https://doi.org/10.1088/0266-5611/20/1/006 doi: 10.1088/0266-5611/20/1/006
![]() |
[20] | D. A. Lorenz, S. Wenger, F. Schöpfer, M. Magnor, A sparse Kaczmarz solver and a linearized Bregman method for online compressed sensing, in 2014 IEEE International Conference on Image Processing (ICIP), (2014), 1347–1351. https://doi.org/10.1109/ICIP.2014.7025269 |
[21] |
J. D. Moorman, T. K. Tu, D. Molitor, D. Needell, Randomized Kaczmarz with averaging, BIT Numer. Math., 61 (2021), 337–359. https://doi.org/10.1007/s10543-020-00824-1 doi: 10.1007/s10543-020-00824-1
![]() |
[22] |
T. Strohmer, R. Vershynin, A randomized Kaczmarz algorithm with exponential convergence, J. Fourier Anal. Appl., 15 (2009), 262–278. https://doi.org/10.1007/s00041-008-9030-4 doi: 10.1007/s00041-008-9030-4
![]() |
[23] | A. Zouzias, N. M. Freris, Randomized extended Kaczmarz for solving least squares, SIAM J. Matrix Anal. A., 34 (2013), 773–793. |
[24] | K. Du, X. H. Sun, Pseudoinverse-free randomized block iterative algorithms for consistent and inconsistent linear systems, preprint, arXiv: 2011.10353. |
[25] |
D. Needell, J. A. Tropp, Paved with good intentions: analysis of a randomized block Kaczmarz method, Linear Algebra Appl., 441 (2014), 199–221. https://doi.org/10.1016/j.laa.2012.12.022 doi: 10.1016/j.laa.2012.12.022
![]() |
[26] |
Z Z. Bai, W. T. Wu, On relaxed greedy randomized Kaczmarz methods for solving large sparse linear systems, Appl. Math. Lett., 83 (2018), 21–26. https://doi.org/10.1016/j.aml.2018.03.008 doi: 10.1016/j.aml.2018.03.008
![]() |
[27] |
Z Z. Bai, W. T. Wu, On greedy randomized Kaczmarz method for solving large sparse linear systems, SIAM J. Sci. Comput., 40 (2018), A592–A606. https://doi.org/10.1137/17M1137747 doi: 10.1137/17M1137747
![]() |
[28] |
E. Rebrova, D. Needell, On block Gaussian sketching for the Kaczmarz method, Numer. Algor., 86 (2021), 443–473. https://doi.org/10.1007/s11075-020-00895-9 doi: 10.1007/s11075-020-00895-9
![]() |
[29] |
Y. Q. Niu, B. Zheng, A greedy block Kaczmarz algorithm for solving large-scale linear systems, Appl. Math. Lett., 104 (2020), 106294. https://doi.org/10.1016/j.aml.2020.106294 doi: 10.1016/j.aml.2020.106294
![]() |
[30] |
I. Necoara, Faster randomized block Kaczmarz algorithms, SIAM J. Matrix Anal. A., 40 (2019), 1425–1452. https://doi.org/10.1137/19M1251643 doi: 10.1137/19M1251643
![]() |
[31] |
T. Elfving, Block-iterative methods for consistent and inconsistent linear equations, Numer. Math., 35 (1980), 1–12. https://doi.org/10.1007/BF01396365 doi: 10.1007/BF01396365
![]() |
[32] |
D. P. Woodruff, Sketching as a tool for numerical linear algebra, Found. Trends Theor. Comput. Sci., 10 (2014), 1–157. http://doi.org/10.1561/0400000060 doi: 10.1561/0400000060
![]() |
[33] |
Y. Zhang, H. Li, L. Tang, Greedy randomized sampling nonlinear Kaczmarz methods, Calcolo, 61 (2024). https://doi.org/10.1007/s10092-024-00577-1 doi: 10.1007/s10092-024-00577-1
![]() |
[34] |
J. Zhang, Y. Wang, J. Zhao, On maximum residual nonlinear Kaczmarz-type algorithms for large nonlinear systems of equations, J. Comput. Appl. Math., 425 (2023), 115065. https://doi.org/10.1016/j.cam.2023.115065 doi: 10.1016/j.cam.2023.115065
![]() |
[35] |
T. Li, T. J. Kao, D. Isaacson, J. C. Newell, G. J. Saulnier, Adaptive Kaczmarz method for image reconstruction in electrical impedance tomography, Physiol. Meas., 34 (2013), 595. https://doi.org/10.1088/0967-3334/34/6/595 doi: 10.1088/0967-3334/34/6/595
![]() |
[36] | M. B. Cohen, C. Musco, C. Musco, Input sparsity time low-rank approximation via ridge leverage score sampling, in Proceedings of the Twenty-Eighth Annual ACM-SIAM Symposium on Discrete Algorithms, (2017), 1758–1777. https://doi.org/10.1137/1.9781611974782.115 |
[37] | A. Rudi, D. Calandriello, L. Carratino, L. Rosasco, On fast leverage score sampling and optimal learning, Adv. Neural Inform. Process. Syst., 31 (2018). |
[38] |
Y. Zhang, H. Li, A count sketch maximal weighted residual Kaczmarz method for solving highly overdetermined linear systems, Appl. Math. Comput., 410 (2021), 126486. https://doi.org/10.1016/j.amc.2021.126486 doi: 10.1016/j.amc.2021.126486
![]() |
[39] |
Y. Jiang, G. Wu, L. Jiang, A semi-randomized Kaczmarz method with simple random sampling for large-scale linear systems, Adv. Comput. Math., 49 (2023). https://doi.org/10.1007/s10444-023-10018-2 doi: 10.1007/s10444-023-10018-2
![]() |
[40] | G. Wu, Q. Chang, A semi-randomized block Kaczmarz method with simple random sampling for large-scale linear systems with multiple right-hand sides, preprint, arXiv: 2212.08797. |