1
|
Shiri I, Salimi Y, Sirjani N, Razeghi B, Bagherieh S, Pakbin M, Mansouri Z, Hajianfar G, Avval AH, Askari D, Ghasemian M, Sandoughdaran S, Sohrabi A, Sadati E, Livani S, Iranpour P, Kolahi S, Khosravi B, Bijari S, Sayfollahi S, Atashzar MR, Hasanian M, Shahhamzeh A, Teimouri A, Goharpey N, Shirzad-Aski H, Karimi J, Radmard AR, Rezaei-Kalantari K, Oghli MG, Oveisi M, Vafaei Sadr A, Voloshynovskiy S, Zaidi H. Differential privacy preserved federated learning for prognostic modeling in COVID-19 patients using large multi-institutional chest CT dataset. Med Phys 2024. [PMID: 38335175 DOI: 10.1002/mp.16964] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/05/2023] [Revised: 01/10/2024] [Accepted: 01/21/2024] [Indexed: 02/12/2024] Open
Abstract
BACKGROUND Notwithstanding the encouraging results of previous studies reporting on the efficiency of deep learning (DL) in COVID-19 prognostication, clinical adoption of the developed methodology still needs to be improved. To overcome this limitation, we set out to predict the prognosis of a large multi-institutional cohort of patients with COVID-19 using a DL-based model. PURPOSE This study aimed to evaluate the performance of deep privacy-preserving federated learning (DPFL) in predicting COVID-19 outcomes using chest CT images. METHODS After applying inclusion and exclusion criteria, 3055 patients from 19 centers, including 1599 alive and 1456 deceased, were enrolled in this study. Data from all centers were split (randomly with stratification respective to each center and class) into a training/validation set (70%/10%) and a hold-out test set (20%). For the DL model, feature extraction was performed on 2D slices, and averaging was performed at the final layer to construct a 3D model for each scan. The DensNet model was used for feature extraction. The model was developed using centralized and FL approaches. For FL, we employed DPFL approaches. Membership inference attack was also evaluated in the FL strategy. For model evaluation, different metrics were reported in the hold-out test sets. In addition, models trained in two scenarios, centralized and FL, were compared using the DeLong test for statistical differences. RESULTS The centralized model achieved an accuracy of 0.76, while the DPFL model had an accuracy of 0.75. Both the centralized and DPFL models achieved a specificity of 0.77. The centralized model achieved a sensitivity of 0.74, while the DPFL model had a sensitivity of 0.73. A mean AUC of 0.82 and 0.81 with 95% confidence intervals of (95% CI: 0.79-0.85) and (95% CI: 0.77-0.84) were achieved by the centralized model and the DPFL model, respectively. The DeLong test did not prove statistically significant differences between the two models (p-value = 0.98). The AUC values for the inference attacks fluctuate between 0.49 and 0.51, with an average of 0.50 ± 0.003 and 95% CI for the mean AUC of 0.500 to 0.501. CONCLUSION The performance of the proposed model was comparable to centralized models while operating on large and heterogeneous multi-institutional datasets. In addition, the model was resistant to inference attacks, ensuring the privacy of shared data during the training process.
Collapse
Affiliation(s)
- Isaac Shiri
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
| | - Yazdan Salimi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
| | - Nasim Sirjani
- Research and Development Department, Med Fanavarn Plus Co, Karaj, Iran
| | - Behrooz Razeghi
- Department of Computer Science, University of Geneva, Geneva, Switzerland
| | - Sara Bagherieh
- School of Medicine, Isfahan University of Medical Sciences, Isfahan, Iran
| | - Masoumeh Pakbin
- Imaging Department, Qom University of Medical Sciences, Qom, Iran
| | - Zahra Mansouri
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
| | - Ghasem Hajianfar
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
| | | | - Dariush Askari
- Department of Radiology Technology, Shahid Beheshti University of Medical Sciences, Tehran, Iran
| | - Mohammadreza Ghasemian
- Department of Radiology, Shahid Beheshti Hospital, Qom University of Medical Sciences, Qom, Iran
| | - Saleh Sandoughdaran
- Department of Clinical Oncology, Royal Surrey County Hospital, Guildford, UK
| | - Ahmad Sohrabi
- Radin Makian Azma Mehr Ltd., Radinmehr Veterinary Laboratory, Iran University of Medical Sciences, Gorgan, Iran
| | - Elham Sadati
- Department of Medical Physics, Faculty of Medical Sciences, Tarbiat Modares University, Tehran, Iran
| | - Somayeh Livani
- Clinical Research Development Unit (CRDU), Sayad Shirazi Hospital, Golestan University of Medical Sciences, Gorgan, Iran
| | - Pooya Iranpour
- Medical Imaging Research Center, Department of Radiology, Shiraz University of Medical Sciences, Shiraz, Iran
| | - Shahriar Kolahi
- Department of Radiology, School of Medicine, Advanced Diagnostic and Interventional Radiology Research Center (ADIR), Imam Khomeini Hospital, Tehran University of Medical Sciences, Tehran, Iran
| | - Bardia Khosravi
- Digestive Diseases Research Center, Digestive Diseases Research Institute, Tehran University of Medical Sciences, Tehran, Iran
| | - Salar Bijari
- Department of Medical Physics, Faculty of Medical Sciences, Tarbiat Modares University, Tehran, Iran
| | - Sahar Sayfollahi
- Department of Neurosurgery, Faculty of Medical Sciences, Iran University of Medical Sciences, Tehran, Iran
| | - Mohammad Reza Atashzar
- Department of Immunology, School of Medicine, Fasa University of Medical Sciences, Fasa, Iran
| | - Mohammad Hasanian
- Department of Radiology, Arak University of Medical Sciences, Arak, Iran
| | - Alireza Shahhamzeh
- Clinical research development center, Qom University of Medical Sciences, Qom, Iran
| | - Arash Teimouri
- Medical Imaging Research Center, Department of Radiology, Shiraz University of Medical Sciences, Shiraz, Iran
| | - Neda Goharpey
- Department of radiation oncology, Shohada-e Tajrish Hospital, Shahid Beheshti University of Medical Sciences, Tehran, Iran
| | | | - Jalal Karimi
- Department of Infectious Disease, School of Medicine, Fasa University of Medical Sciences, Fasa, Iran
| | - Amir Reza Radmard
- Department of Radiology, Shariati Hospital, Tehran University of Medical Sciences, Tehran, Iran
| | - Kiara Rezaei-Kalantari
- Rajaie Cardiovascular, Medical & Research Center, Iran University of Medical Science, Tehran, Iran
| | | | - Mehrdad Oveisi
- Department of Computer Science, University of British Columbia, Vancouver, British Columbia, Canada
| | - Alireza Vafaei Sadr
- Department of Public Health Sciences, College of Medicine, Pennsylvania State University, Hershey, Pennsylvania, USA
| | | | - Habib Zaidi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
- Department of Nuclear Medicine and Molecular Imaging, University of Groningen, University Medical Center Groningen, Groningen, Netherlands
- Department of Nuclear Medicine, University of Southern Denmark, Odense, Denmark
- University Research and Innovation Center, Óbuda University, Budapest, Hungary
| |
Collapse
|
2
|
Kinakh V, Belousov Y, Quétant G, Drozdova M, Holotyak T, Schaerer D, Voloshynovskiy S. Hubble Meets Webb: Image-to-Image Translation in Astronomy. Sensors (Basel) 2024; 24:1151. [PMID: 38400307 PMCID: PMC10892943 DOI: 10.3390/s24041151] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 01/01/2024] [Revised: 02/04/2024] [Accepted: 02/07/2024] [Indexed: 02/25/2024]
Abstract
This work explores the generation of James Webb Space Telescope (JWSP) imagery via image-to-image translation from the available Hubble Space Telescope (HST) data. Comparative analysis encompasses the Pix2Pix, CycleGAN, TURBO, and DDPM-based Palette methodologies, assessing the criticality of image registration in astronomy. While the focus of this study is not on the scientific evaluation of model fairness, we note that the techniques employed may bear some limitations and the translated images could include elements that are not present in actual astronomical phenomena. To mitigate this, uncertainty estimation is integrated into our methodology, enhancing the translation's integrity and assisting astronomers in distinguishing between reliable predictions and those of questionable certainty. The evaluation was performed using metrics including MSE, SSIM, PSNR, LPIPS, and FID. The paper introduces a novel approach to quantifying uncertainty within image translation, leveraging the stochastic nature of DDPMs. This innovation not only bolsters our confidence in the translated images but also provides a valuable tool for future astronomical experiment planning. By offering predictive insights when JWST data are unavailable, our approach allows for informed preparatory strategies for making observations with the upcoming JWST, potentially optimizing its precious observational resources. To the best of our knowledge, this work is the first attempt to apply image-to-image translation for astronomical sensor-to-sensor translation.
Collapse
Affiliation(s)
- Vitaliy Kinakh
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (V.K.); (Y.B.); (G.Q.); (M.D.); (T.H.)
| | - Yury Belousov
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (V.K.); (Y.B.); (G.Q.); (M.D.); (T.H.)
| | - Guillaume Quétant
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (V.K.); (Y.B.); (G.Q.); (M.D.); (T.H.)
| | - Mariia Drozdova
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (V.K.); (Y.B.); (G.Q.); (M.D.); (T.H.)
| | - Taras Holotyak
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (V.K.); (Y.B.); (G.Q.); (M.D.); (T.H.)
| | - Daniel Schaerer
- Department of Astronomy, University of Geneva, 1290 Versoix, Switzerland;
| | - Slava Voloshynovskiy
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (V.K.); (Y.B.); (G.Q.); (M.D.); (T.H.)
| |
Collapse
|
3
|
Shiri I, Razeghi B, Ferdowsi S, Salimi Y, Gündüz D, Teodoro D, Voloshynovskiy S, Zaidi H. PRIMIS: Privacy-preserving medical image sharing via deep sparsifying transform learning with obfuscation. J Biomed Inform 2024; 150:104583. [PMID: 38191010 DOI: 10.1016/j.jbi.2024.104583] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/27/2023] [Revised: 11/19/2023] [Accepted: 01/02/2024] [Indexed: 01/10/2024]
Abstract
OBJECTIVE The primary objective of our study is to address the challenge of confidentially sharing medical images across different centers. This is often a critical necessity in both clinical and research environments, yet restrictions typically exist due to privacy concerns. Our aim is to design a privacy-preserving data-sharing mechanism that allows medical images to be stored as encoded and obfuscated representations in the public domain without revealing any useful or recoverable content from the images. In tandem, we aim to provide authorized users with compact private keys that could be used to reconstruct the corresponding images. METHOD Our approach involves utilizing a neural auto-encoder. The convolutional filter outputs are passed through sparsifying transformations to produce multiple compact codes. Each code is responsible for reconstructing different attributes of the image. The key privacy-preserving element in this process is obfuscation through the use of specific pseudo-random noise. When applied to the codes, it becomes computationally infeasible for an attacker to guess the correct representation for all the codes, thereby preserving the privacy of the images. RESULTS The proposed framework was implemented and evaluated using chest X-ray images for different medical image analysis tasks, including classification, segmentation, and texture analysis. Additionally, we thoroughly assessed the robustness of our method against various attacks using both supervised and unsupervised algorithms. CONCLUSION This study provides a novel, optimized, and privacy-assured data-sharing mechanism for medical images, enabling multi-party sharing in a secure manner. While we have demonstrated its effectiveness with chest X-ray images, the mechanism can be utilized in other medical images modalities as well.
Collapse
Affiliation(s)
- Isaac Shiri
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland; Department of Cardiology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland
| | - Behrooz Razeghi
- Department of Computer Science, University of Geneva, Switzerland; Idiap Research Institute, Switzerland
| | - Sohrab Ferdowsi
- Department of Radiology and Medical Informatics, University of Geneva, Switzerland
| | - Yazdan Salimi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
| | - Deniz Gündüz
- Department of Electrical and Electronic Engineering, Imperial College London, UK
| | - Douglas Teodoro
- Department of Radiology and Medical Informatics, University of Geneva, Switzerland
| | | | - Habib Zaidi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland; Department of Nuclear Medicine and Molecular Imaging, University of Groningen, The Netherlands; Department of Nuclear Medicine, University of Southern Denmark, Denmark; University Research and Innovation Center, Óbuda University, Budapest, Hungary.
| |
Collapse
|
4
|
Shiri I, Salimi Y, Maghsudi M, Jenabi E, Harsini S, Razeghi B, Mostafaei S, Hajianfar G, Sanaat A, Jafari E, Samimi R, Khateri M, Sheikhzadeh P, Geramifar P, Dadgar H, Bitrafan Rajabi A, Assadi M, Bénard F, Vafaei Sadr A, Voloshynovskiy S, Mainta I, Uribe C, Rahmim A, Zaidi H. Differential privacy preserved federated transfer learning for multi-institutional 68Ga-PET image artefact detection and disentanglement. Eur J Nucl Med Mol Imaging 2023; 51:40-53. [PMID: 37682303 PMCID: PMC10684636 DOI: 10.1007/s00259-023-06418-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/14/2023] [Accepted: 08/24/2023] [Indexed: 09/09/2023]
Abstract
PURPOSE Image artefacts continue to pose challenges in clinical molecular imaging, resulting in misdiagnoses, additional radiation doses to patients and financial costs. Mismatch and halo artefacts occur frequently in gallium-68 (68Ga)-labelled compounds whole-body PET/CT imaging. Correcting for these artefacts is not straightforward and requires algorithmic developments, given that conventional techniques have failed to address them adequately. In the current study, we employed differential privacy-preserving federated transfer learning (FTL) to manage clinical data sharing and tackle privacy issues for building centre-specific models that detect and correct artefacts present in PET images. METHODS Altogether, 1413 patients with 68Ga prostate-specific membrane antigen (PSMA)/DOTA-TATE (TOC) PET/CT scans from 3 countries, including 8 different centres, were enrolled in this study. CT-based attenuation and scatter correction (CT-ASC) was used in all centres for quantitative PET reconstruction. Prior to model training, an experienced nuclear medicine physician reviewed all images to ensure the use of high-quality, artefact-free PET images (421 patients' images). A deep neural network (modified U2Net) was trained on 80% of the artefact-free PET images to utilize centre-based (CeBa), centralized (CeZe) and the proposed differential privacy FTL frameworks. Quantitative analysis was performed in 20% of the clean data (with no artefacts) in each centre. A panel of two nuclear medicine physicians conducted qualitative assessment of image quality, diagnostic confidence and image artefacts in 128 patients with artefacts (256 images for CT-ASC and FTL-ASC). RESULTS The three approaches investigated in this study for 68Ga-PET imaging (CeBa, CeZe and FTL) resulted in a mean absolute error (MAE) of 0.42 ± 0.21 (CI 95%: 0.38 to 0.47), 0.32 ± 0.23 (CI 95%: 0.27 to 0.37) and 0.28 ± 0.15 (CI 95%: 0.25 to 0.31), respectively. Statistical analysis using the Wilcoxon test revealed significant differences between the three approaches, with FTL outperforming CeBa and CeZe (p-value < 0.05) in the clean test set. The qualitative assessment demonstrated that FTL-ASC significantly improved image quality and diagnostic confidence and decreased image artefacts, compared to CT-ASC in 68Ga-PET imaging. In addition, mismatch and halo artefacts were successfully detected and disentangled in the chest, abdomen and pelvic regions in 68Ga-PET imaging. CONCLUSION The proposed approach benefits from using large datasets from multiple centres while preserving patient privacy. Qualitative assessment by nuclear medicine physicians showed that the proposed model correctly addressed two main challenging artefacts in 68Ga-PET imaging. This technique could be integrated in the clinic for 68Ga-PET imaging artefact detection and disentanglement using multicentric heterogeneous datasets.
Collapse
Affiliation(s)
- Isaac Shiri
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva, Switzerland
- Department of Cardiology, Inselspital, University of Bern, Bern, Switzerland
| | - Yazdan Salimi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva, Switzerland
| | - Mehdi Maghsudi
- Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Sciences, Tehran, Iran
| | - Elnaz Jenabi
- Research Center for Nuclear Medicine, Shariati Hospital, Tehran University of Medical Sciences, Tehran, Iran
| | - Sara Harsini
- BC Cancer Research Institute, Vancouver, BC, Canada
| | - Behrooz Razeghi
- Department of Computer Science, University of Geneva, Geneva, Switzerland
| | - Shayan Mostafaei
- Division of Clinical Geriatrics, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm, Sweden
- Department of Medical Epidemiology and Biostatistics, Karolinska Institute, Stockholm, Sweden
| | - Ghasem Hajianfar
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva, Switzerland
| | - Amirhossein Sanaat
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva, Switzerland
| | - Esmail Jafari
- The Persian Gulf Nuclear Medicine Research Center, Department of Nuclear Medicine, Molecular Imaging, and Theranostics, Bushehr Medical University Hospital, School of Medicine, Bushehr University of Medical Sciences, Bushehr, Iran
| | - Rezvan Samimi
- Department of Medical Radiation Engineering, Shahid Beheshti University, Tehran, Iran
| | - Maziar Khateri
- Department of Medical Radiation Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran
| | - Peyman Sheikhzadeh
- Department of Nuclear Medicine, Imam Khomeini Hospital Complex, Tehran University of Medical Sciences, Tehran, Iran
| | - Parham Geramifar
- Research Center for Nuclear Medicine, Shariati Hospital, Tehran University of Medical Sciences, Tehran, Iran
| | - Habibollah Dadgar
- Cancer Research Center, Razavi Hospital, Imam Reza International University, Mashhad, Iran
| | - Ahmad Bitrafan Rajabi
- Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Sciences, Tehran, Iran
- Echocardiography Research Center, Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Sciences, Tehran, Iran
| | - Majid Assadi
- The Persian Gulf Nuclear Medicine Research Center, Department of Nuclear Medicine, Molecular Imaging, and Theranostics, Bushehr Medical University Hospital, School of Medicine, Bushehr University of Medical Sciences, Bushehr, Iran
| | - François Bénard
- BC Cancer Research Institute, Vancouver, BC, Canada
- Department of Radiology, University of British Columbia, Vancouver, BC, Canada
| | - Alireza Vafaei Sadr
- Institute of Pathology, RWTH Aachen University Hospital, Aachen, Germany
- Department of Public Health Sciences, College of Medicine, The Pennsylvania State University, Hershey, PA, 17033, USA
| | | | - Ismini Mainta
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva, Switzerland
| | - Carlos Uribe
- Department of Radiology, University of British Columbia, Vancouver, BC, Canada
- Molecular Imaging and Therapy, BC Cancer, Vancouver, BC, Canada
- Department of Integrative Oncology, BC Cancer Research Institute, Vancouver, BC, Canada
| | - Arman Rahmim
- Department of Radiology, University of British Columbia, Vancouver, BC, Canada
- Department of Integrative Oncology, BC Cancer Research Institute, Vancouver, BC, Canada
- Department of Physics and Astronomy, University of British Columbia, Vancouver, Canada
| | - Habib Zaidi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva, Switzerland.
- Geneva University Neuro Center, Geneva University, Geneva, Switzerland.
- Department of Nuclear Medicine and Molecular Imaging, University of Groningen, University Medical Center Groningen, Groningen, Netherlands.
- Department of Nuclear Medicine, University of Southern Denmark, Odense, Denmark.
| |
Collapse
|
5
|
Quétant G, Belousov Y, Kinakh V, Voloshynovskiy S. TURBO: The Swiss Knife of Auto-Encoders. Entropy (Basel) 2023; 25:1471. [PMID: 37895592 PMCID: PMC10606332 DOI: 10.3390/e25101471] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 09/20/2023] [Revised: 10/13/2023] [Accepted: 10/18/2023] [Indexed: 10/29/2023]
Abstract
We present a novel information-theoretic framework, termed as TURBO, designed to systematically analyse and generalise auto-encoding methods. We start by examining the principles of information bottleneck and bottleneck-based networks in the auto-encoding setting and identifying their inherent limitations, which become more prominent for data with multiple relevant, physics-related representations. The TURBO framework is then introduced, providing a comprehensive derivation of its core concept consisting of the maximisation of mutual information between various data representations expressed in two directions reflecting the information flows. We illustrate that numerous prevalent neural network models are encompassed within this framework. The paper underscores the insufficiency of the information bottleneck concept in elucidating all such models, thereby establishing TURBO as a preferable theoretical reference. The introduction of TURBO contributes to a richer understanding of data representation and the structure of neural network models, enabling more efficient and versatile applications.
Collapse
Affiliation(s)
- Guillaume Quétant
- Centre Universitaire d’Informatique, Université de Genève, Route de Drize 7, CH-1227 Carouge, Switzerland; (Y.B.); (V.K.)
| | | | | | - Slava Voloshynovskiy
- Centre Universitaire d’Informatique, Université de Genève, Route de Drize 7, CH-1227 Carouge, Switzerland; (Y.B.); (V.K.)
| |
Collapse
|
6
|
Shiri I, Razeghi B, Vafaei Sadr A, Amini M, Salimi Y, Ferdowsi S, Boor P, Gündüz D, Voloshynovskiy S, Zaidi H. Multi-institutional PET/CT image segmentation using federated deep transformer learning. Comput Methods Programs Biomed 2023; 240:107706. [PMID: 37506602 DOI: 10.1016/j.cmpb.2023.107706] [Citation(s) in RCA: 3] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 01/09/2023] [Accepted: 07/02/2023] [Indexed: 07/30/2023]
Abstract
BACKGROUND AND OBJECTIVE Generalizable and trustworthy deep learning models for PET/CT image segmentation necessitates large diverse multi-institutional datasets. However, legal, ethical, and patient privacy issues challenge sharing of datasets between different centers. To overcome these challenges, we developed a federated learning (FL) framework for multi-institutional PET/CT image segmentation. METHODS A dataset consisting of 328 FL (HN) cancer patients who underwent clinical PET/CT examinations gathered from six different centers was enrolled. A pure transformer network was implemented as fully core segmentation algorithms using dual channel PET/CT images. We evaluated different frameworks (single center-based, centralized baseline, as well as seven different FL algorithms) using 68 PET/CT images (20% of each center data). In particular, the implemented FL algorithms include clipping with the quantile estimator (ClQu), zeroing with the quantile estimator (ZeQu), federated averaging (FedAvg), lossy compression (LoCo), robust aggregation (RoAg), secure aggregation (SeAg), and Gaussian differentially private FedAvg with adaptive quantile clipping (GDP-AQuCl). RESULTS The Dice coefficient was 0.80±0.11 for both centralized and SeAg FL algorithms. All FL approaches achieved centralized learning model performance with no statistically significant differences. Among the FL algorithms, SeAg and GDP-AQuCl performed better than the other techniques. However, there was no statistically significant difference. All algorithms, except the center-based approach, resulted in relative errors less than 5% for SUVmax and SUVmean for all FL and centralized methods. Centralized and FL algorithms significantly outperformed the single center-based baseline. CONCLUSIONS The developed FL-based (with centralized method performance) algorithms exhibited promising performance for HN tumor segmentation from PET/CT images.
Collapse
Affiliation(s)
- Isaac Shiri
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
| | - Behrooz Razeghi
- Department of Computer Science, University of Geneva, Geneva, Switzerland
| | - Alireza Vafaei Sadr
- Institute of Pathology, RWTH Aachen University Hospital, Aachen, Germany; Department of Public Health Sciences, College of Medicine, The Pennsylvania State University, Hershey, PA 17033, USA
| | - Mehdi Amini
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
| | - Yazdan Salimi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland
| | - Sohrab Ferdowsi
- Department of Computer Science, University of Geneva, Geneva, Switzerland
| | - Peter Boor
- Institute of Pathology, RWTH Aachen University Hospital, Aachen, Germany
| | - Deniz Gündüz
- Department of Electrical and Electronic Engineering, Imperial College London, UK
| | | | - Habib Zaidi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, Geneva, Switzerland; Geneva University Neurocenter, University of Geneva, Geneva, Switzerland; Department of Nuclear Medicine and Molecular Imaging, University of Groningen, Groningen, The Netherlands; Department of Nuclear Medicine, University of Southern Denmark, Odense, Denmark.
| |
Collapse
|
7
|
Taran O, Tutt J, Holotyak T, Chaban R, Bonev S, Voloshynovskiy S. Mobile authentication of copy detection patterns. EURASIP J Inf Secur 2023; 2023:4. [PMID: 37292064 PMCID: PMC10244288 DOI: 10.1186/s13635-023-00140-5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/22/2022] [Accepted: 05/23/2023] [Indexed: 06/10/2023] Open
Abstract
In the recent years, the copy detection patterns (CDP) attracted a lot of attention as a link between the physical and digital worlds, which is of great interest for the internet of things and brand protection applications. However, the security of CDP in terms of their reproducibility by unauthorized parties or clonability remains largely unexplored. In this respect, this paper addresses a problem of anti-counterfeiting of physical objects and aims at investigating the authentication aspects and the resistances to illegal copying of the modern CDP from machine learning perspectives. A special attention is paid to a reliable authentication under the real-life verification conditions when the codes are printed on an industrial printer and enrolled via modern mobile phones under regular light conditions. The theoretical and empirical investigation of authentication aspects of CDP is performed with respect to four types of copy fakes from the point of view of (i) multi-class supervised classification as a baseline approach and (ii) one-class classification as a real-life application case. The obtained results show that the modern machine-learning approaches and the technical capacities of modern mobile phones allow to reliably authenticate CDP on end-user mobile phones under the considered classes of fakes.
Collapse
Affiliation(s)
- Olga Taran
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, 1227 Carouge, Switzerland
| | - Joakim Tutt
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, 1227 Carouge, Switzerland
| | - Taras Holotyak
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, 1227 Carouge, Switzerland
| | - Roman Chaban
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, 1227 Carouge, Switzerland
| | - Slavi Bonev
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, 1227 Carouge, Switzerland
| | - Slava Voloshynovskiy
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, 1227 Carouge, Switzerland
| |
Collapse
|
8
|
Ullmann D, Taran O, Voloshynovskiy S. Multivariate Time Series Information Bottleneck. Entropy (Basel) 2023; 25:e25050831. [PMID: 37238586 DOI: 10.3390/e25050831] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 02/22/2023] [Revised: 05/10/2023] [Accepted: 05/17/2023] [Indexed: 05/28/2023]
Abstract
Time series (TS) and multiple time series (MTS) predictions have historically paved the way for distinct families of deep learning models. The temporal dimension, distinguished by its evolutionary sequential aspect, is usually modeled by decomposition into the trio of "trend, seasonality, noise", by attempts to copy the functioning of human synapses, and more recently, by transformer models with self-attention on the temporal dimension. These models may find applications in finance and e-commerce, where any increase in performance of less than 1% has large monetary repercussions, they also have potential applications in natural language processing (NLP), medicine, and physics. To the best of our knowledge, the information bottleneck (IB) framework has not received significant attention in the context of TS or MTS analyses. One can demonstrate that a compression of the temporal dimension is key in the context of MTS. We propose a new approach with partial convolution, where a time sequence is encoded into a two-dimensional representation resembling images. Accordingly, we use the recent advances made in image extension to predict an unseen part of an image from a given one. We show that our model compares well with traditional TS models, has information-theoretical foundations, and can be easily extended to more dimensions than only time and space. An evaluation of our multiple time series-information bottleneck (MTS-IB) model proves its efficiency in electricity production, road traffic, and astronomical data representing solar activity, as recorded by NASA's interface region imaging spectrograph (IRIS) satellite.
Collapse
Affiliation(s)
- Denis Ullmann
- Faculty of Science, University of Geneva, CUI, 1227 Carouge, Switzerland
| | - Olga Taran
- Faculty of Science, University of Geneva, CUI, 1227 Carouge, Switzerland
| | | |
Collapse
|
9
|
Shiri I, Vafaei Sadr A, Akhavan A, Salimi Y, Sanaat A, Amini M, Razeghi B, Saberi A, Arabi H, Ferdowsi S, Voloshynovskiy S, Gündüz D, Rahmim A, Zaidi H. Decentralized collaborative multi-institutional PET attenuation and scatter correction using federated deep learning. Eur J Nucl Med Mol Imaging 2023; 50:1034-1050. [PMID: 36508026 PMCID: PMC9742659 DOI: 10.1007/s00259-022-06053-8] [Citation(s) in RCA: 10] [Impact Index Per Article: 10.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/13/2022] [Accepted: 11/18/2022] [Indexed: 12/15/2022]
Abstract
PURPOSE Attenuation correction and scatter compensation (AC/SC) are two main steps toward quantitative PET imaging, which remain challenging in PET-only and PET/MRI systems. These can be effectively tackled via deep learning (DL) methods. However, trustworthy, and generalizable DL models commonly require well-curated, heterogeneous, and large datasets from multiple clinical centers. At the same time, owing to legal/ethical issues and privacy concerns, forming a large collective, centralized dataset poses significant challenges. In this work, we aimed to develop a DL-based model in a multicenter setting without direct sharing of data using federated learning (FL) for AC/SC of PET images. METHODS Non-attenuation/scatter corrected and CT-based attenuation/scatter corrected (CT-ASC) 18F-FDG PET images of 300 patients were enrolled in this study. The dataset consisted of 6 different centers, each with 50 patients, with scanner, image acquisition, and reconstruction protocols varying across the centers. CT-based ASC PET images served as the standard reference. All images were reviewed to include high-quality and artifact-free PET images. Both corrected and uncorrected PET images were converted to standardized uptake values (SUVs). We used a modified nested U-Net utilizing residual U-block in a U-shape architecture. We evaluated two FL models, namely sequential (FL-SQ) and parallel (FL-PL) and compared their performance with the baseline centralized (CZ) learning model wherein the data were pooled to one server, as well as center-based (CB) models where for each center the model was built and evaluated separately. Data from each center were divided to contribute to training (30 patients), validation (10 patients), and test sets (10 patients). Final evaluations and reports were performed on 60 patients (10 patients from each center). RESULTS In terms of percent SUV absolute relative error (ARE%), both FL-SQ (CI:12.21-14.81%) and FL-PL (CI:11.82-13.84%) models demonstrated excellent agreement with the centralized framework (CI:10.32-12.00%), while FL-based algorithms improved model performance by over 11% compared to CB training strategy (CI: 22.34-26.10%). Furthermore, the Mann-Whitney test between different strategies revealed no significant differences between CZ and FL-based algorithms (p-value > 0.05) in center-categorized mode. At the same time, a significant difference was observed between the different training approaches on the overall dataset (p-value < 0.05). In addition, voxel-wise comparison, with respect to reference CT-ASC, exhibited similar performance for images predicted by CZ (R2 = 0.94), FL-SQ (R2 = 0.93), and FL-PL (R2 = 0.92), while CB model achieved a far lower coefficient of determination (R2 = 0.74). Despite the strong correlations between CZ and FL-based methods compared to reference CT-ASC, a slight underestimation of predicted voxel values was observed. CONCLUSION Deep learning-based models provide promising results toward quantitative PET image reconstruction. Specifically, we developed two FL models and compared their performance with center-based and centralized models. The proposed FL-based models achieved higher performance compared to center-based models, comparable with centralized models. Our work provided strong empirical evidence that the FL framework can fully benefit from the generalizability and robustness of DL models used for AC/SC in PET, while obviating the need for the direct sharing of datasets between clinical imaging centers.
Collapse
Affiliation(s)
- Isaac Shiri
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland
| | - Alireza Vafaei Sadr
- Department of Theoretical Physics and Center for Astroparticle Physics, University of Geneva, Geneva, Switzerland.,Institute of Pathology, RWTH Aachen University Hospital, Aachen, Germany
| | - Azadeh Akhavan
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland
| | - Yazdan Salimi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland
| | - Amirhossein Sanaat
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland
| | - Mehdi Amini
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland
| | - Behrooz Razeghi
- Department of Computer Science, University of Geneva, Geneva, Switzerland
| | - Abdollah Saberi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland
| | - Hossein Arabi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland
| | | | | | - Deniz Gündüz
- Department of Electrical and Electronic Engineering, Imperial College London, London, UK
| | - Arman Rahmim
- Departments of Radiology and Physics, University of British Columbia, Vancouver, Canada.,Department of Integrative Oncology, BC Cancer Research Institute, Vancouver, BC, Canada
| | - Habib Zaidi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland. .,Geneva University Neurocenter, Geneva University, Geneva, Switzerland. .,Department of Nuclear Medicine and Molecular Imaging, University of Groningen, University Medical Center Groningen, Groningen, Netherlands. .,Department of Nuclear Medicine, University of Southern Denmark, Odense, Denmark.
| |
Collapse
|
10
|
Rezaeifar S, Voloshynovskiy S, Asgari Jirhandeh M, Kinakh V. Privacy-Preserving Image Template Sharing Using Contrastive Learning. Entropy 2022; 24:e24050643. [PMID: 35626528 PMCID: PMC9141880 DOI: 10.3390/e24050643] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 01/04/2022] [Revised: 04/08/2022] [Accepted: 04/20/2022] [Indexed: 12/10/2022]
Abstract
With the recent developments of Machine Learning as a Service (MLaaS), various privacy concerns have been raised. Having access to the user’s data, an adversary can design attacks with different objectives, namely, reconstruction or attribute inference attacks. In this paper, we propose two different training frameworks for an image classification task while preserving user data privacy against the two aforementioned attacks. In both frameworks, an encoder is trained with contrastive loss, providing a superior utility-privacy trade-off. In the reconstruction attack scenario, a supervised contrastive loss was employed to provide maximal discrimination for the targeted classification task. The encoded features are further perturbed using the obfuscator module to remove all redundant information. Moreover, the obfuscator module is jointly trained with a classifier to minimize the correlation between private feature representation and original data while retaining the model utility for the classification. For the attribute inference attack, we aim to provide a representation of data that is independent of the sensitive attribute. Therefore, the encoder is trained with supervised and private contrastive loss. Furthermore, an obfuscator module is trained in an adversarial manner to preserve the privacy of sensitive attributes while maintaining the classification performance on the target attribute. The reported results on the CelebA dataset validate the effectiveness of the proposed frameworks.
Collapse
|
11
|
Ullmann D, Rezaeifar S, Taran O, Holotyak T, Panos B, Voloshynovskiy S. Information Bottleneck Classification in Extremely Distributed Systems. Entropy (Basel) 2020; 22:e22111237. [PMID: 33287005 PMCID: PMC7711965 DOI: 10.3390/e22111237] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 08/28/2020] [Revised: 10/24/2020] [Accepted: 10/26/2020] [Indexed: 06/01/2023]
Abstract
We present a new decentralized classification system based on a distributed architecture. This system consists of distributed nodes, each possessing their own datasets and computing modules, along with a centralized server, which provides probes to classification and aggregates the responses of nodes for a final decision. Each node, with access to its own training dataset of a given class, is trained based on an auto-encoder system consisting of a fixed data-independent encoder, a pre-trained quantizer and a class-dependent decoder. Hence, these auto-encoders are highly dependent on the class probability distribution for which the reconstruction distortion is minimized. Alternatively, when an encoding-quantizing-decoding node observes data from different distributions, unseen at training, there is a mismatch, and such a decoding is not optimal, leading to a significant increase of the reconstruction distortion. The final classification is performed at the centralized classifier that votes for the class with the minimum reconstruction distortion. In addition to the system applicability for applications facing big-data communication problems and or requiring private classification, the above distributed scheme creates a theoretical bridge to the information bottleneck principle. The proposed system demonstrates a very promising performance on basic datasets such as MNIST and FasionMNIST.
Collapse
|
12
|
Voloshynovskiy S, Taran O, Kondah M, Holotyak T, Rezende D. Variational Information Bottleneck for Semi-Supervised Classification. Entropy (Basel) 2020; 22:E943. [PMID: 33286710 PMCID: PMC7597214 DOI: 10.3390/e22090943] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 07/22/2020] [Revised: 08/24/2020] [Accepted: 08/24/2020] [Indexed: 11/17/2022]
Abstract
In this paper, we consider an information bottleneck (IB) framework for semi-supervised classification with several families of priors on latent space representation. We apply a variational decomposition of mutual information terms of IB. Using this decomposition we perform an analysis of several regularizers and practically demonstrate an impact of different components of variational model on the classification accuracy. We propose a new formulation of semi-supervised IB with hand crafted and learnable priors and link it to the previous methods such as semi-supervised versions of VAE (M1 + M2), AAE, CatGAN, etc. We show that the resulting model allows better understand the role of various previously proposed regularizers in semi-supervised classification task in the light of IB framework. The proposed IB semi-supervised model with hand-crafted and learnable priors is experimentally validated on MNIST under different amount of labeled data.
Collapse
Affiliation(s)
- Slava Voloshynovskiy
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (O.T.); (M.K.); (T.H.)
| | - Olga Taran
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (O.T.); (M.K.); (T.H.)
| | - Mouad Kondah
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (O.T.); (M.K.); (T.H.)
| | - Taras Holotyak
- Department of Computer Science, University of Geneva, 1227 Carouge, Switzerland; (O.T.); (M.K.); (T.H.)
| | | |
Collapse
|
13
|
Taran O, Rezaeifar S, Holotyak T, Voloshynovskiy S. Machine learning through cryptographic glasses: combating adversarial attacks by key-based diversified aggregation. EURASIP J Inf Secur 2020; 2020:10. [PMID: 32685910 PMCID: PMC7357678 DOI: 10.1186/s13635-020-00106-x] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/04/2019] [Accepted: 04/22/2020] [Indexed: 12/03/2022] Open
Abstract
In recent years, classification techniques based on deep neural networks (DNN) were widely used in many fields such as computer vision, natural language processing, and self-driving cars. However, the vulnerability of the DNN-based classification systems to adversarial attacks questions their usage in many critical applications. Therefore, the development of robust DNN-based classifiers is a critical point for the future deployment of these methods. Not less important issue is understanding of the mechanisms behind this vulnerability. Additionally, it is not completely clear how to link machine learning with cryptography to create an information advantage of the defender over the attacker. In this paper, we propose a key-based diversified aggregation (KDA) mechanism as a defense strategy in a gray- and black-box scenario. KDA assumes that the attacker (i) knows the architecture of classifier and the used defense strategy, (ii) has an access to the training data set, but (iii) does not know a secret key and does not have access to the internal states of the system. The robustness of the system is achieved by a specially designed key-based randomization. The proposed randomization prevents the gradients’ back propagation and restricts the attacker to create a “bypass” system. The randomization is performed simultaneously in several channels. Each channel introduces its own randomization in a special transform domain. The sharing of a secret key between the training and test stages creates an information advantage to the defender. Finally, the aggregation of soft outputs from each channel stabilizes the results and increases the reliability of the final score. The performed experimental evaluation demonstrates a high robustness and universality of the KDA against state-of-the-art gradient-based gray-box transferability attacks and the non-gradient-based black-box attacks (The results reported in this paper have been partially presented in CVPR 2019 (Taran et al., Defending against adversarial attacks by randomized diversification, 2019) & ICIP 2019 (Taran et al., Robustification of deep net classifiers by key-based diversified aggregation with pre-filtering, 2019)).
Collapse
Affiliation(s)
- Olga Taran
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, Carouge, GE Switzerland
| | - Shideh Rezaeifar
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, Carouge, GE Switzerland
| | - Taras Holotyak
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, Carouge, GE Switzerland
| | - Slava Voloshynovskiy
- Stochastic Information Processing Group, Department of Computer Science, University of Geneva, 7 Route de Drize, Carouge, GE Switzerland
| |
Collapse
|
14
|
Nowak J, Holotyak T, Korytkowski M, Scherer R, Voloshynovskiy S. Fingerprinting of URL Logs: Continuous User Authentication from Behavioural Patterns. Lecture Notes in Computer Science 2020. [PMCID: PMC7303709 DOI: 10.1007/978-3-030-50423-6_14] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Indexed: 11/03/2022]
Abstract
Security of computer systems is now a critical and evolving issue. Current trends try to use behavioural biometrics for continuous authorization. Our work is intended to strengthen network user authentication by a software interaction analysis. In our research, we use HTTP request (URLs) logs that network administrators collect. We use a set of full-convolutional autoencoders and one authentication (one-class) convolutional neural network. The proposed method copes with extensive data from many users and allows to add new users in the future. Moreover, the system works in a real-time manner, and the proposed deep learning framework can use other user behaviour- and software interaction-related features.
Collapse
|