Joachim A. Behar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Joachim A. Behar has a total of 296 co-authors. Here are the top co-authors by number of shared publications:
- Jan Van Eijgen
KU Leuven
7 shared publications - Yael Yaniv
Laboratory of Bioelectric and Bioenergetic Systems, Faculty of Biomedical Engineering, Technion-IIT, Haifa 3200003, Israel
7 shared publications - Eran Zvuloni
Faculty of Biomedical Engineering, Technion-IIT, Haifa, Israel
6 shared publications - Ingeborg Stalmans
KU Leuven
6 shared publications Jonathan Fhima
Technion, Israel Institute of Technology
6 shared publications- Jonathan Sobel
Infection Control Program and WHO Collaborating Center, Geneva University Hospitals and Faculty of Medicine, Geneva, Switzerland
6 shared publications
Latest Publications
- Open Access
- 26 October 2026
- Published by IOP Publishing in Machine Learning: Health
- Vol. 2 (2) , 025022
Enhancing retinal vessel segmentation generalization via layout-aware generative modelling
- 26 October 2026
- Published by IOP Publishing in Machine Learning: Health
- Vol. 2 (2) , 025022
Open AccessEnhancing retinal vessel segmentation generalization via layout-aware generative modelling
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Engineering
- Vol. 73 (10) , 3567-3575
uPVC-Net: A Universal Premature Ventricular Contraction Detection Deep Learning Algorithm
Premature Ventricular Contractions (PVCs) are common cardiac arrhythmias originating from the ventricles. Accurate detection remains challenging due to variability in electrocardiogram (ECG) waveforms caused by differences in lead placement, recording conditions, and population demographics. We developed uPVC-Net, a universal deep learning model to detect PVCs from any single-lead ECG recordings. The model is developed on four independent ECG datasets comprising a total of 8.3 million beats collected from Holter monitors and a modern wearable ECG patch. uPVC-Net employs a custom architecture and a multi-source, multi-lead training strategy. For each experiment, one dataset is held out to evaluate out-of-distribution (OOD) generalization. uPVC-Net achieved an AUC between 97.8% and 99.1% on the held-out datasets. Notably, performance on wearable single-lead ECG data reached an AUC of 99.1%. uPVC-Net exhibits strong generalization across diverse lead configurations and populations, highlighting its potential for robust, real-world clinical deployment.
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Engineering
- Vol. 73 (10) , 3567-3575
uPVC-Net: A Universal Premature Ventricular Contraction Detection Deep Learning Algorithm
Premature Ventricular Contractions (PVCs) are common cardiac arrhythmias originating from the ventricles. Accurate detection remains challenging due to variability in electrocardiogram (ECG) waveforms caused by differences in lead placement, recording conditions, and population demographics. We developed uPVC-Net, a universal deep learning model to detect PVCs from any single-lead ECG recordings. The model is developed on four independent ECG datasets comprising a total of 8.3 million beats collected from Holter monitors and a modern wearable ECG patch. uPVC-Net employs a custom architecture and a multi-source, multi-lead training strategy. For each experiment, one dataset is held out to evaluate out-of-distribution (OOD) generalization. uPVC-Net achieved an AUC between 97.8% and 99.1% on the held-out datasets. Notably, performance on wearable single-lead ECG data reached an AUC of 99.1%. uPVC-Net exhibits strong generalization across diverse lead configurations and populations, highlighting its potential for robust, real-world clinical deployment.
- Open Access
- 30 September 2026
- Published by IOP Publishing in Physiological Measurement
- Vol. 47 (9) , 095011
Single-lead ECG biomarkers for cardiovascular and mortality risk prediction
Objective. Cardiovascular disease remains one of the leading causes of morbidity and mortality worldwide, highlighting the need for accurate risk prediction. Atrial fibrillation burden (AFB), premature ventricular contraction burden (PVCB), and T-wave alternans (TWA) have been individually associated with adverse outcomes. We hypothesized that these three digital ECG biomarkers provide complementary prognostic information for major cardiovascular endpoints and all-cause mortality (ACM), and developed a machine learning approach combining these biomarkers for risk prediction. Approach. We analyzed 81 362 Holter recordings from 54 395 individuals across 20 primary care centers in Israel. A random forest model using AFB, PVCB, TWA, and age was trained to predict 5 year risk of heart failure (HF), ischemic stroke (IS), and ACM. Main results. On the test set, best AUROCs were for HF 0.75 [95% CI: 0.74–0.77] ( n pos = 622 ), for IS 0.69 [0.66–0.71] ( n pos = 365 ), and for ACM 0.79 [0.77–0.80] ( n pos = 1060 ). Combining the three ECG-derived biomarkers showed complementary predictive value and improved discrimination by up to 10% over age alone in individuals aged < 75 years. Significance. AFB, PVCB, and TWA contribute complementary prognostic information for risk stratification. When combined with age, these biomarkers improve predictive performance compared with age alone.
- 30 September 2026
- Published by IOP Publishing in Physiological Measurement
- Vol. 47 (9) , 095011
Open AccessSingle-lead ECG biomarkers for cardiovascular and mortality risk prediction
Objective. Cardiovascular disease remains one of the leading causes of morbidity and mortality worldwide, highlighting the need for accurate risk prediction. Atrial fibrillation burden (AFB), premature ventricular contraction burden (PVCB), and T-wave alternans (TWA) have been individually associated with adverse outcomes. We hypothesized that these three digital ECG biomarkers provide complementary prognostic information for major cardiovascular endpoints and all-cause mortality (ACM), and developed a machine learning approach combining these biomarkers for risk prediction. Approach. We analyzed 81 362 Holter recordings from 54 395 individuals across 20 primary care centers in Israel. A random forest model using AFB, PVCB, TWA, and age was trained to predict 5 year risk of heart failure (HF), ischemic stroke (IS), and ACM. Main results. On the test set, best AUROCs were for HF 0.75 [95% CI: 0.74–0.77] ( n pos = 622 ), for IS 0.69 [0.66–0.71] ( n pos = 365 ), and for ACM 0.79 [0.77–0.80] ( n pos = 1060 ). Combining the three ECG-derived biomarkers showed complementary predictive value and improved discrimination by up to 10% over age alone in individuals aged < 75 years. Significance. AFB, PVCB, and TWA contribute complementary prognostic information for risk stratification. When combined with age, these biomarkers improve predictive performance compared with age alone.
- Open Access
- 27 June 2026
- Published by Oxford University Press (OUP) in SLEEP
- Vol. 49 (8)
The Quest for Automated Pediatric Sleep Scoring: Are We There Yet?
- 27 June 2026
- Published by Oxford University Press (OUP) in SLEEP
- Vol. 49 (8)
Open AccessThe Quest for Automated Pediatric Sleep Scoring: Are We There Yet?
- Open Access
- 24 June 2026
- Published by IOP Publishing in Machine Learning: Health
- Vol. 2 (2) , 025011
Benchmarking AI Against Expert Ophthalmologists for Clinical Detection of Referable Eye Diseases
The past few years have witnessed a rapid proliferation of AI systems for the automated interpretation of digital fundus images (DFIs). Although these systems have achieved strong performance on retrospective datasets and several have recently obtained FDA approval, their comparative performance against human expert interpretation remains insufficiently characterized. The aim of this study was to compare the diagnostic performance of an AI system, Lirot.ai, with that of board-certified ophthalmologists in detecting referable eye diseases from single, color non-mydriatic DFIs. This study evaluated the detection of three vision threatening ophthalmic diseases: (1) referable diabetic retinopathy (rDR), (2) referable age-related macular degeneration (rAMD), and (3) glaucomatous optic neuropathy (GON). Primary endpoints included the sensitivity and specificity for detection of each referable eye disease by Lirot.ai versus human readers. Secondary endpoints include inter-reader agreement (κ), human-hours required to annotate, and subset analysis of detection performance on images marked ungradable by human readers. A total of 10 ophthalmologists were recruited for the study: 80% of whom were specialists and 80% of whom had ⩾6 years of clinical experience. Fleiss’ k among readers was 0.401 for GON detection, 0.663 for rAMD detection, and 0.719 for rDR detection. Lirot.ai’s sensitivities across GON, rAMD, and rDR detection were the maximum among readers while maintaining specificities within the observed range of reader performance. Lirot.ai detected GON with an area under the receiver operating curve (AUROC) of 0.860 (sensitivity and specificity of 88.2% and 74.3%, respectively), rAMD with an AUROC of 0.923 (sensitivity and specificity of 92.9% and 89.6%, respectively), and rDR with an AUROC of 0.954 (sensitivity and specificity of 91.7% and 94.0%, respectively) where sensitivities and specificities maximize Youden’s Index. Lirot.ai met or exceeded readers’ disease detection across GON, rDR, and rAMD as evident by readers’ sensitivity-specificity performance along or below the ROC curve. Lirot.ai demonstrated expert-level performance across three referable eye diseases and even surpassed human readers in GON detection, demonstrating the potential of AI to provide consistent, high-quality interpretation of fundus images. These results highlight how AI systems can enhance clinical decision-making and expand access to eye care by supporting large-scale, efficient screening pathways.
- 24 June 2026
- Published by IOP Publishing in Machine Learning: Health
- Vol. 2 (2) , 025011
Open AccessBenchmarking AI Against Expert Ophthalmologists for Clinical Detection of Referable Eye Diseases
The past few years have witnessed a rapid proliferation of AI systems for the automated interpretation of digital fundus images (DFIs). Although these systems have achieved strong performance on retrospective datasets and several have recently obtained FDA approval, their comparative performance against human expert interpretation remains insufficiently characterized. The aim of this study was to compare the diagnostic performance of an AI system, Lirot.ai, with that of board-certified ophthalmologists in detecting referable eye diseases from single, color non-mydriatic DFIs. This study evaluated the detection of three vision threatening ophthalmic diseases: (1) referable diabetic retinopathy (rDR), (2) referable age-related macular degeneration (rAMD), and (3) glaucomatous optic neuropathy (GON). Primary endpoints included the sensitivity and specificity for detection of each referable eye disease by Lirot.ai versus human readers. Secondary endpoints include inter-reader agreement (κ), human-hours required to annotate, and subset analysis of detection performance on images marked ungradable by human readers. A total of 10 ophthalmologists were recruited for the study: 80% of whom were specialists and 80% of whom had ⩾6 years of clinical experience. Fleiss’ k among readers was 0.401 for GON detection, 0.663 for rAMD detection, and 0.719 for rDR detection. Lirot.ai’s sensitivities across GON, rAMD, and rDR detection were the maximum among readers while maintaining specificities within the observed range of reader performance. Lirot.ai detected GON with an area under the receiver operating curve (AUROC) of 0.860 (sensitivity and specificity of 88.2% and 74.3%, respectively), rAMD with an AUROC of 0.923 (sensitivity and specificity of 92.9% and 89.6%, respectively), and rDR with an AUROC of 0.954 (sensitivity and specificity of 91.7% and 94.0%, respectively) where sensitivities and specificities maximize Youden’s Index. Lirot.ai met or exceeded readers’ disease detection across GON, rDR, and rAMD as evident by readers’ sensitivity-specificity performance along or below the ROC curve. Lirot.ai demonstrated expert-level performance across three referable eye diseases and even surpassed human readers in GON detection, demonstrating the potential of AI to provide consistent, high-quality interpretation of fundus images. These results highlight how AI systems can enhance clinical decision-making and expand access to eye care by supporting large-scale, efficient screening pathways.
- Open Access
- 19 June 2026
- Published by Springer Nature in npj Digital Medicine
- Vol. 9 (1)
Modeling day-long ECG signals to predict heart failure risk with explainable AI
Heart failure (HF) affects 11.8% of adults aged 65 and older, reducing quality of life and longevity. Preventing HF can reduce morbidity and mortality. We hypothesized that artificial intelligence (AI) applied to 24-hour single-lead electrocardiogram (ECG) data could predict the risk of HF within five years. To research this, the Technion-Leumit Holter ECG (TLHE) dataset, including 69,663 recordings from 47,729 patients, collected over 20 years, was used. Our deep learning model, DeepHHF, trained on 24-hour ECG recordings, achieved an area under the receiver operating characteristic curve of 0.80 that outperformed a model using 30-second segments and a clinical score. High-risk individuals identified by DeepHHF had a two-fold chance of hospitalization or death incidents. Explainability analysis showed DeepHHF focused on arrhythmias and heart abnormalities. This study highlights the feasibility of deep learning to model 24-hour continuous ECG data, capturing paroxysmal events essential for reliable risk prediction. Artificial intelligence applied to single-lead Holter ECG is non-invasive, inexpensive, and widely accessible, making it a promising tool for HF risk prediction.
- 19 June 2026
- Published by Springer Nature in npj Digital Medicine
- Vol. 9 (1)
Open AccessModeling day-long ECG signals to predict heart failure risk with explainable AI
Heart failure (HF) affects 11.8% of adults aged 65 and older, reducing quality of life and longevity. Preventing HF can reduce morbidity and mortality. We hypothesized that artificial intelligence (AI) applied to 24-hour single-lead electrocardiogram (ECG) data could predict the risk of HF within five years. To research this, the Technion-Leumit Holter ECG (TLHE) dataset, including 69,663 recordings from 47,729 patients, collected over 20 years, was used. Our deep learning model, DeepHHF, trained on 24-hour ECG recordings, achieved an area under the receiver operating characteristic curve of 0.80 that outperformed a model using 30-second segments and a clinical score. High-risk individuals identified by DeepHHF had a two-fold chance of hospitalization or death incidents. Explainability analysis showed DeepHHF focused on arrhythmias and heart abnormalities. This study highlights the feasibility of deep learning to model 24-hour continuous ECG data, capturing paroxysmal events essential for reliable risk prediction. Artificial intelligence applied to single-lead Holter ECG is non-invasive, inexpensive, and widely accessible, making it a promising tool for HF risk prediction.
- Open Access
- 14 May 2026
- Published by IOP Publishing in Machine Learning: Health
- Vol. 2 (1) , 015013
Machine learning for carotid stenosis identification from infrared fundus images
Carotid artery stenosis (CAS) is a significant risk factor for ischemic stroke and a marker of systemic vascular burden. Hemodynamically significant CAS may induce microvascular changes in the retinal circulation, detectable through near-infrared (NIR) reflectance imaging. This study explored the potential of machine learning to identify CAS from NIR fundus images. We used a dataset of 1483 NIR images from 469 patients, including 121 with confirmed CAS and 348 controls. The data were split into training and testing sets (70:30). Two models were developed: (1) a logistic regression model using quantitative vascular biomarkers computed by a specialized deep learning segmentation model, and (2) a deep learning model trained on raw NIR images. Both models achieved an area under the receiver operating characteristic curve of 0.70. These results highlight the potential of machine learning for detecting CAS in NIR fundus imaging, though further validation in larger cohorts is needed.
- 14 May 2026
- Published by IOP Publishing in Machine Learning: Health
- Vol. 2 (1) , 015013
Open AccessMachine learning for carotid stenosis identification from infrared fundus images
Carotid artery stenosis (CAS) is a significant risk factor for ischemic stroke and a marker of systemic vascular burden. Hemodynamically significant CAS may induce microvascular changes in the retinal circulation, detectable through near-infrared (NIR) reflectance imaging. This study explored the potential of machine learning to identify CAS from NIR fundus images. We used a dataset of 1483 NIR images from 469 patients, including 121 with confirmed CAS and 348 controls. The data were split into training and testing sets (70:30). Two models were developed: (1) a logistic regression model using quantitative vascular biomarkers computed by a specialized deep learning segmentation model, and (2) a deep learning model trained on raw NIR images. Both models achieved an area under the receiver operating characteristic curve of 0.70. These results highlight the potential of machine learning for detecting CAS in NIR fundus imaging, though further validation in larger cohorts is needed.
- Open Access
- 6 March 2026
- Published by IOP Publishing in Physiological Measurement
- Vol. 47 (3) , 035004
Ophthalmology foundation models for clinically significant age macular degeneration detection
Objective . Self-supervised learning (SSL) has enabled vision transformers (ViTs) to learn robust representations from large-scale natural image datasets, enhancing their generalization across domains. In retinal imaging, foundation models pretrained on either natural or ophthalmic data have shown promise, but the benefits of in-domain pretraining remain uncertain. Approach . To investigate this, we benchmark six SSL-pretrained ViTs on seven digital fundus image (DFI) datasets totaling 70 000 expert-annotated images for the task of moderate-to-late age-related macular degeneration (AMD) identification. Main results . Our results show that DINOv2, pretrained on natural images, shows similar performance than domain-specific models. These findings highlight the value of foundation models in improving AMD identification, and challenge the assumption that in-domain pretraining is necessary. Significance . We present our model AMDNet, which performs state-of-the-art out-of-domain AUROCs on six public datasets. Furthermore, we release BRAMD, an open-access dataset ( n = 587) of DFIs with AMD labels from Brazil. Project page: www.aimlab-technion.com/lirot-ai .
- 6 March 2026
- Published by IOP Publishing in Physiological Measurement
- Vol. 47 (3) , 035004
Open AccessOphthalmology foundation models for clinically significant age macular degeneration detection
Objective . Self-supervised learning (SSL) has enabled vision transformers (ViTs) to learn robust representations from large-scale natural image datasets, enhancing their generalization across domains. In retinal imaging, foundation models pretrained on either natural or ophthalmic data have shown promise, but the benefits of in-domain pretraining remain uncertain. Approach . To investigate this, we benchmark six SSL-pretrained ViTs on seven digital fundus image (DFI) datasets totaling 70 000 expert-annotated images for the task of moderate-to-late age-related macular degeneration (AMD) identification. Main results . Our results show that DINOv2, pretrained on natural images, shows similar performance than domain-specific models. These findings highlight the value of foundation models in improving AMD identification, and challenge the assumption that in-domain pretraining is necessary. Significance . We present our model AMDNet, which performs state-of-the-art out-of-domain AUROCs on six public datasets. Furthermore, we release BRAMD, an open-access dataset ( n = 587) of DFIs with AMD labels from Brazil. Project page: www.aimlab-technion.com/lirot-ai .
- Open Access
- 5 March 2026
- Published by Springer Nature in npj Digital Medicine
- Vol. 9 (1)
Shifting the retinal foundation models paradigm from slices to volumes for optical coherence tomography
Optical Coherence Tomography (OCT) is essential in ophthalmology for cross-sectional imaging of the retina. Pretrained foundation models facilitate task-specific model development by enabling fine-tuning with limited labeled data. However, current foundation models rely on a single B-scan (usually the central slice), overlooking volumetric context. This research investigates video foundation models to capture full 3D retinal structure and improve diagnostic performance. V-JEPA, a state-of-the-art video foundation model, was benchmarked against retinal foundation models (RETFound, VisionFM) and a natural image foundation model (DINOv2). All were fine-tuned to detect Age-related Macular Degeneration or Glaucomatous Optic Neuropathy using five OCT datasets. V-JEPA consistently equaled or outperformed image-based models, achieving an average AUROC of 0.94 (0.80–0.99), versus 0.90 (0.76–0.98) for the best image model, a statistically significant improvement (p < 0.001). To our knowledge, this is the first application of transformer-based video models to volumetric OCT, highlighting their promise in 3D medical imaging.
- 5 March 2026
- Published by Springer Nature in npj Digital Medicine
- Vol. 9 (1)
Open AccessShifting the retinal foundation models paradigm from slices to volumes for optical coherence tomography
Optical Coherence Tomography (OCT) is essential in ophthalmology for cross-sectional imaging of the retina. Pretrained foundation models facilitate task-specific model development by enabling fine-tuning with limited labeled data. However, current foundation models rely on a single B-scan (usually the central slice), overlooking volumetric context. This research investigates video foundation models to capture full 3D retinal structure and improve diagnostic performance. V-JEPA, a state-of-the-art video foundation model, was benchmarked against retinal foundation models (RETFound, VisionFM) and a natural image foundation model (DINOv2). All were fine-tuned to detect Age-related Macular Degeneration or Glaucomatous Optic Neuropathy using five OCT datasets. V-JEPA consistently equaled or outperformed image-based models, achieving an average AUROC of 0.94 (0.80–0.99), versus 0.90 (0.76–0.98) for the best image model, a statistically significant improvement (p < 0.001). To our knowledge, this is the first application of transformer-based video models to volumetric OCT, highlighting their promise in 3D medical imaging.
- Open Access
- 24 February 2026
- Published by IOP Publishing in Machine Learning: Health
- Vol. 2 (1)
Multisource domain training with retinal expert-in-the-loop for accurate diabetic retinopathy staging from fundus images
Open-source datasets of digital fundus images annotated for diabetic retinopathy (DR) staging often suffer from inconsistent or inaccurate labeling but manually relabeling them at scale is impractical. A key need is a semi-supervised methodology which automatically flags images most likely to be mislabeled for expert review. We present DRStageNet2, our model improvement to DRStageNet, which results exclusively from our data refinement strategy with an expert-in-the-loop who reannotates images selectively in a data-driven manner across multiple rounds of model retraining and inference. Through this process, the expert reviewed a total of 3984 images among the 91 984 images used (2.8%). The expert modified the labels a total of 2592 (65%) of the reviewed images. DRStageNet2 improved out-of-domain generalization performance across five external validation sets. DRStageNet2’s multiclass accuracy of International Clinical Diabetic Retinopathy Scale scores 0–4 improved across all datasets with a mean increase of 3.6%. Its quadratic-weighted Cohen kappa score (Q-kappa) increased from an initial average of 0.89 (range: 0.86–0.91) to 0.91 (range: 0.89–0.93) with improvements of up to 6.2% in one dataset. DRStageNet2 then, achieves a Q-kappa score at the upper end of previously reported inter-rater agreement between retinal specialists.
- 24 February 2026
- Published by IOP Publishing in Machine Learning: Health
- Vol. 2 (1)
Open AccessMultisource domain training with retinal expert-in-the-loop for accurate diabetic retinopathy staging from fundus images
Open-source datasets of digital fundus images annotated for diabetic retinopathy (DR) staging often suffer from inconsistent or inaccurate labeling but manually relabeling them at scale is impractical. A key need is a semi-supervised methodology which automatically flags images most likely to be mislabeled for expert review. We present DRStageNet2, our model improvement to DRStageNet, which results exclusively from our data refinement strategy with an expert-in-the-loop who reannotates images selectively in a data-driven manner across multiple rounds of model retraining and inference. Through this process, the expert reviewed a total of 3984 images among the 91 984 images used (2.8%). The expert modified the labels a total of 2592 (65%) of the reviewed images. DRStageNet2 improved out-of-domain generalization performance across five external validation sets. DRStageNet2’s multiclass accuracy of International Clinical Diabetic Retinopathy Scale scores 0–4 improved across all datasets with a mean increase of 3.6%. Its quadratic-weighted Cohen kappa score (Q-kappa) increased from an initial average of 0.89 (range: 0.86–0.91) to 0.91 (range: 0.89–0.93) with improvements of up to 6.2% in one dataset. DRStageNet2 then, achieves a Q-kappa score at the upper end of previously reported inter-rater agreement between retinal specialists.