Joachim A. Behar

Jonathan Fhima, Jan Van Eijgen, Ingeborg Stalmans et al.

Introduction: Blood vessels can be non-invasively visualized from a digital fundus image (DFI). Several studies have shown an association between cardiovascular risk and vascular features obtained from DFI. Recent advances in computer vision and image segmentation enable automatising DFI blood vessel segmentation. There is a need for a resource that can automatically compute digital vasculature biomarkers (VBM) from these segmented DFI. Methods: In this paper, we introduce a Python Vasculature BioMarker toolbox, denoted PVBM. A total of 11 VBMs were implemented. In particular, we introduce new algorithmic methods to estimate tortuosity and branching angles. Using PVBM, and as a proof of usability, we analyze geometric vascular differences between glaucomatous patients and healthy controls. Results: We built a fully automated vasculature biomarker toolbox based on DFI segmentations and provided a proof of usability to characterize the vascular changes in glaucoma. For arterioles and venules, all biomarkers were significant and lower in glaucoma patients compared to healthy controls except for tortuosity, venular singularity length and venular branching angles. Conclusion: We have automated the computation of 11 VBMs from retinal blood vessel segmentation. The PVBM toolbox is made open source under a GNU GPL 3 license and is available on physiozoo.com (following publication).

Or Abramovich, Hadas Pizem, Jan Van Eijgen et al.

Objective: Ophthalmological pathologies such as glaucoma, diabetic retinopathy and age-related macular degeneration are major causes of blindness and vision impairment. There is a need for novel decision support tools that can simplify and speed up the diagnosis of these pathologies. A key step in this process is to automatically estimate the quality of the fundus images to make sure these are interpretable by a human operator or a machine learning model. We present a novel fundus image quality scale and deep learning (DL) model that can estimate fundus image quality relative to this new scale. Methods: A total of 1,245 images were graded for quality by two ophthalmologists within the range 1-10, with a resolution of 0.5. A DL regression model was trained for fundus image quality assessment. The architecture used was Inception-V3. The model was developed using a total of 89,947 images from 6 databases, of which 1,245 were labeled by the specialists and the remaining 88,702 images were used for pre-training and semi-supervised learning. The final DL model was evaluated on an internal test set (n=209) as well as an external test set (n=194). Results: The final DL model, denoted FundusQ-Net, achieved a mean absolute error of 0.61 (0.54-0.68) on the internal test set. When evaluated as a binary classification model on the public DRIMDB database as an external test set the model obtained an accuracy of 99%. Significance: the proposed algorithm provides a new robust tool for automated quality grading of fundus images.

Eran Zvuloni, Sheina Gendelman, Sanghamitra Mohanty et al.

Introduction: 12-lead electrocardiogram (ECG) is recorded during atrial fibrillation (AF) catheter ablation procedure (CAP). It is not easy to determine if CAP was successful without a long follow-up assessing for AF recurrence (AFR). Therefore, an AFR risk prediction algorithm could enable a better management of CAP patients. In this research, we extracted features from 12-lead ECG recorded before and after CAP and train an AFR risk prediction machine learning model. Methods: Pre- and post-CAP segments were extracted from 112 patients. The analysis included a signal quality criterion, heart rate variability and morphological biomarkers engineered from the 12-lead ECG (804 features overall). 43 out of the 112 patients (n) had AFR clinical endpoint available. These were utilized to assess the feasibility of AFR risk prediction, using either pre or post CAP features. A random forest classifier was trained within a nested cross validation framework. Results: 36 features were found statistically significant for distinguishing between the pre and post surgery states (n=112). For the classification, an area under the receiver operating characteristic (AUROC) curve was reported with AUROC_pre=0.64 and AUROC_post=0.74 (n=43). Discussion and conclusions: This preliminary analysis showed the feasibility of AFR risk prediction. Such a model could be used to improve CAP management.

Eran Zvuloni, Jesse Read, Antônio H. Ribeiro et al.

Objective: Machine learning techniques have been used extensively for 12-lead electrocardiogram (ECG) analysis. For physiological time series, deep learning (DL) superiority to feature engineering (FE) approaches based on domain knowledge is still an open question. Moreover, it remains unclear whether combining DL with FE may improve performance. Methods: We considered three tasks intending to address these research gaps: cardiac arrhythmia diagnosis (multiclass-multilabel classification), atrial fibrillation risk prediction (binary classification), and age estimation (regression). We used an overall dataset of 2.3M 12-lead ECG recordings to train the following models for each task: i) a random forest taking the FE as input was trained as a classical machine learning approach; ii) an end-to-end DL model; and iii) a merged model of FE+DL. Results: FE yielded comparable results to DL while necessitating significantly less data for the two classification tasks and it was outperformed by DL for the regression task. For all tasks, merging FE with DL did not improve performance over DL alone. Conclusion: We found that for traditional 12-lead ECG based diagnosis tasks DL did not yield a meaningful improvement over FE, while it improved significantly the nontraditional regression task. We also found that combining FE with DL did not improve over DL alone which suggests that the FE were redundant with the features learned by DL. Significance: Our findings provides important recommendations on what machine learning strategy and data regime to chose with respect to the task at hand for the development of new machine learning models based on the 12-lead ECG.

Raphael Azriel, Cecil D. Hahn, Thomas De Cooman et al.

Objective: Epileptic seizures are relatively common in critically-ill children admitted to the pediatric intensive care unit (PICU) and thus serve as an important target for identification and treatment. Most of these seizures have no discernible clinical manifestation but still have a significant impact on morbidity and mortality. Children that are deemed at risk for seizures within the PICU are monitored using continuous-electroencephalogram (cEEG). cEEG monitoring cost is considerable and as the number of available machines is always limited, clinicians need to resort to triaging patients according to perceived risk in order to allocate resources. This research aims to develop a computer aided tool to improve seizures risk assessment in critically-ill children, using an ubiquitously recorded signal in the PICU, namely the electrocardiogram (ECG). Approach: A novel data-driven model was developed at a patient-level approach, based on features extracted from the first hour of ECG recording and the clinical data of the patient. Main results: The most predictive features were the age of the patient, the brain injury as coma etiology and the QRS area. For patients without any prior clinical data, using one hour of ECG recording, the classification performance of the random forest classifier reached an area under the receiver operating characteristic curve (AUROC) score of 0.84. When combining ECG features with the patients clinical history, the AUROC reached 0.87. Significance: Taking a real clinical scenario, we estimated that our clinical decision support triage tool can improve the positive predictive value by more than 59% over the clinical standard.

Jonathan Fhima, Jan Van Eijgen, Moti Freiman et al.

Introduction: For supervised deep learning (DL) tasks, researchers need a large annotated dataset. In medical data science, one of the major limitations to develop DL models is the lack of annotated examples in large quantity. This is most often due to the time and expertise required to annotate. We introduce Lirot. ai, a novel platform for facilitating and crowd-sourcing image segmentations. Methods: Lirot. ai is composed of three components; an iPadOS client application named Lirot. ai-app, a backend server named Lirot. ai-server and a python API name Lirot. ai-API. Lirot. ai-app was developed in Swift 5.6 and Lirot. ai-server is a firebase backend. Lirot. ai-API allows the management of the database. Lirot. ai-app can be installed on as many iPadOS devices as needed so that annotators may be able to perform their segmentation simultaneously and remotely. We incorporate Apple Pencil compatibility, making the segmentation faster, more accurate, and more intuitive for the expert than any other computer-based alternative. Results: We demonstrate the usage of Lirot. ai for the creation of a retinal fundus dataset with reference vasculature segmentations. Discussion and future work: We will use active learning strategies to continue enlarging our retinal fundus dataset by including a more efficient process to select the images to be annotated and distribute them to annotators.

Jonathan A. Sobel, Ronit Almog, Leo Anthony Celi et al.

A datathon is a time-constrained competition involving data science applied to a specific problem. In the past decade, datathons have been shown to be a valuable bridge between fields and expertise . Biomedical data analysis represents a challenging area requiring collaboration between engineers, biologists and physicians to gain a better understanding of patient physiology and of guide decision processes for diagnosis, prognosis and therapeutic interventions to improve care practice. Here, we reflect on the outcomes of an event that we organized in Israel at the end of March 2022 between the MIT Critical Data group, Rambam Health Care Campus (Rambam) and the Technion Israel Institute of Technology (Technion) in Haifa. Participants were asked to complete a survey about their skills and interests, which enabled us to identify current needs in machine learning training for medical problem applications. This work describes opportunities and limitations in medical data science in the Israeli context.

Shany Biton, Mohsin Aldhafeeri, Erez Marcusohn et al.

To drive health innovation that meets the needs of all and democratize healthcare, there is a need to assess the generalization performance of deep learning (DL) algorithms across various distribution shifts to ensure that these algorithms are robust. This retrospective study is, to the best of our knowledge, the first to develop and assess the generalization performance of a deep learning (DL) model for AF events detection from long term beat-to-beat intervals across ethnicities, ages and sexes. The new recurrent DL model, denoted ArNet2, was developed on a large retrospective dataset of 2,147 patients totaling 51,386 hours of continuous electrocardiogram (ECG). The models generalization was evaluated on manually annotated test sets from four centers (USA, Israel, Japan and China) totaling 402 patients. The model was further validated on a retrospective dataset of 1,730 consecutives Holter recordings from the Rambam Hospital Holter clinic, Haifa, Israel. The model outperformed benchmark state-of-the-art models and generalized well across ethnicities, ages and sexes. Performance was higher for female than male and young adults (less than 60 years old) and showed some differences across ethnicities. The main finding explaining these variations was an impairment in performance in groups with a higher prevalence of atrial flutter (AFL). Our findings on the relative performance of ArNet2 across groups may have clinical implications on the choice of the preferred AF examination method to use relative to the group of interest.

Eran Zvuloni, Leo Anthony Celi, Joachim A. Behar

The scientific community is increasingly recognizing the importance of generalization in medical AI for translating research into practical clinical applications. A three-level scale is introduced to characterize out-of-distribution generalization performance of medical AI models. This scale addresses the diversity of real-world medical scenarios as well as whether target domain data and labels are available for model recalibration. It serves as a tool to help researchers characterize their development settings and determine the best approach to tackling the challenge of out-of-distribution generalization.

Theogene Habineza, Antônio H. Ribeiro, Daniel Gedon et al.

Background: Atrial fibrillation (AF) is one of the most common cardiac arrhythmias that affects millions of people each year worldwide and it is closely linked to increased risk of cardiovascular diseases such as stroke and heart failure. Machine learning methods have shown promising results in evaluating the risk of developing atrial fibrillation from the electrocardiogram. We aim to develop and evaluate one such algorithm on a large CODE dataset collected in Brazil. Results: The deep neural network model identified patients without indication of AF in the presented ECG but who will develop AF in the future with an AUC score of 0.845. From our survival model, we obtain that patients in the high-risk group (i.e. with the probability of a future AF case being greater than 0.7) are 50% more likely to develop AF within 40 weeks, while patients belonging to the minimal-risk group (i.e. with the probability of a future AF case being less than or equal to 0.1) have more than 85% chance of remaining AF free up until after seven years. Conclusion: We developed and validated a model for AF risk prediction. If applied in clinical practice, the model possesses the potential of providing valuable and useful information in decision-making and patient management processes.

Jonathan Fhima, Jan Van Eijgen, Hana Kulenovic et al.

The retina is the only part of the human body in which blood vessels can be accessed non-invasively using imaging techniques such as digital fundus images (DFI). The spatial distribution of the retinal microvasculature may change with cardiovascular diseases and thus the eyes may be regarded as a window to our hearts. Computerized segmentation of the retinal arterioles and venules (A/V) is essential for automated microvasculature analysis. Using active learning, we created a new DFI dataset containing 240 crowd-sourced manual A/V segmentations performed by fifteen medical students and reviewed by an ophthalmologist, and developed LUNet, a novel deep learning architecture for high resolution A/V segmentation. LUNet architecture includes a double dilated convolutional block that aims to enhance the receptive field of the model and reduce its parameter count. Furthermore, LUNet has a long tail that operates at high resolution to refine the segmentation. The custom loss function emphasizes the continuity of the blood vessels. LUNet is shown to significantly outperform two state-of-the-art segmentation algorithms on the local test set as well as on four external test sets simulating distribution shifts across ethnicity, comorbidities, and annotators. We make the newly created dataset open access (upon publication).

Kenta Tsutsui, Shany Biton Brimer, Noam Ben-Moshe et al.

Atrial fibrillation (AF) is a common atrial arrhythmia that impairs quality of life and causes embolic stroke, heart failure and other complications. Recent advancements in machine learning (ML) and deep learning (DL) have shown potential for enhancing diagnostic accuracy. It is essential for DL models to be robust and generalizable across variations in ethnicity, age, sex, and other factors. Although a number of ECG database have been made available to the research community, none includes a Japanese population sample. Saitama Heart Database Atrial Fibrillation (SHDB-AF) is a novel open-sourced Holter ECG database from Japan, containing data from 100 unique patients with paroxysmal AF. Each record in SHDB-AF is 24 hours long and sampled at 200 Hz, totaling 24 million seconds of ECG data.

Jonathan Fhima, Jan Van Eijgen, Lennert Beeckmans et al.

Generalization in medical segmentation models is challenging due to limited annotated datasets and imaging variability. To address this, we propose Retinal Layout-Aware Diffusion (RLAD), a novel diffusion-based framework for generating controllable layout-aware images. RLAD conditions image generation on multiple key layout components extracted from real images, ensuring high structural fidelity while enabling diversity in other components. Applied to retinal fundus imaging, we augmented the training datasets by synthesizing paired retinal images and vessel segmentations conditioned on extracted blood vessels from real images, while varying other layout components such as lesions and the optic disc. Experiments demonstrated that RLAD-generated data improved generalization in retinal vessel segmentation by up to 8.1%. Furthermore, we present REYIA, a comprehensive dataset comprising 586 manually segmented retinal images. To foster reproducibility and drive innovation, both our code and dataset will be made publicly accessible.

Noam Ben-Moshe, Kenta Tsutsui, Shany Biton et al.

Introduction: Deep learning models for detecting episodes of atrial fibrillation (AF) using rhythm information in long-term, ambulatory ECG recordings have shown high performance. However, the rhythm-based approach does not take advantage of the morphological information conveyed by the different ECG waveforms, particularly the f-waves. As a result, the performance of such models may be inherently limited. Methods: To address this limitation, we have developed a deep learning model, named RawECGNet, to detect episodes of AF and atrial flutter (AFl) using the raw, single-lead ECG. We compare the generalization performance of RawECGNet on two external data sets that account for distribution shifts in geography, ethnicity, and lead position. RawECGNet is further benchmarked against a state-of-the-art deep learning model, named ArNet2, which utilizes rhythm information as input. Results: Using RawECGNet, the results for the different leads in the external test sets in terms of the F1 score were 0.91--0.94 in RBDB and 0.93 in SHDB, compared to 0.89--0.91 in RBDB and 0.91 in SHDB for ArNet2. The results highlight RawECGNet as a high-performance, generalizable algorithm for detection of AF and AFl episodes, exploiting information on both rhythm and morphology.

Yevgeniy Men, Jonathan Fhima, Leo Anthony Celi et al.

Diabetic retinopathy (DR) is a prevalent complication of diabetes associated with a significant risk of vision loss. Timely identification is critical to curb vision impairment. Algorithms for DR staging from digital fundus images (DFIs) have been recently proposed. However, models often fail to generalize due to distribution shifts between the source domain on which the model was trained and the target domain where it is deployed. A common and particularly challenging shift is often encountered when the source- and target-domain supports do not fully overlap. In this research, we introduce DRStageNet, a deep learning model designed to mitigate this challenge. We used seven publicly available datasets, comprising a total of 93,534 DFIs that cover a variety of patient demographics, ethnicities, geographic origins and comorbidities. We fine-tune DINOv2, a pretrained model of self-supervised vision transformer, and implement a multi-source domain fine-tuning strategy to enhance generalization performance. We benchmark and demonstrate the superiority of our method to two state-of-the-art benchmarks, including a recently published foundation model. We adapted the grad-rollout method to our regression task in order to provide high-resolution explainability heatmaps. The error analysis showed that 59\% of the main errors had incorrect reference labels. DRStageNet is accessible at URL [upon acceptance of the manuscript].

Shirel Attia, Revital Shani Hershkovich, Alissa Tabakhov et al.

Background: Sleep staging is a fundamental component in the diagnosis of sleep disorders and the management of sleep health. Traditionally, this analysis is conducted in clinical settings and involves a time-consuming scoring procedure. Recent data-driven algorithms for sleep staging, using the photoplethysmogram (PPG) time series, have shown high performance on local test sets but lower performance on external datasets due to data drift. Methods: This study aimed to develop a generalizable deep learning model for the task of four class (wake, light, deep, and rapid eye movement (REM)) sleep staging from raw PPG physiological time-series. Six sleep datasets, totaling 2,574 patients recordings, were used. In order to create a more generalizable representation, we developed and evaluated a deep learning model called SleepPPG-Net2, which employs a multi-source domain training approach.SleepPPG-Net2 was benchmarked against two state-of-the-art models. Results: SleepPPG-Net2 showed consistently higher performance over benchmark approaches, with generalization performance (Cohen's kappa) improving by up to 19%. Performance disparities were observed in relation to age, sex, and sleep apnea severity. Conclusion: SleepPPG-Net2 sets a new standard for staging sleep from raw PPG time-series.

Or Abramovich, Hadas Pizem, Jonathan Fhima et al.

Glaucomatous optic neuropathy (GON) is a prevalent ocular disease that can lead to irreversible vision loss if not detected early and treated. The traditional diagnostic approach for GON involves a set of ophthalmic examinations, which are time-consuming and require a visit to an ophthalmologist. Recent deep learning models for automating GON detection from digital fundus images (DFI) have shown promise but often suffer from limited generalizability across different ethnicities, disease groups and examination settings. To address these limitations, we introduce GONet, a robust deep learning model developed using seven independent datasets, including over 119,000 DFIs with gold-standard annotations and from patients of diverse geographic backgrounds. GONet consists of a DINOv2 pre-trained self-supervised vision transformers fine-tuned using a multisource domain strategy. GONet demonstrated high out-of-distribution generalizability, with an AUC of 0.85-0.99 in target domains. GONet performance was similar or superior to state-of-the-art works and was significantly superior to the cup-to-disc ratio, by up to 21.6%. GONet is available at [URL provided on publication]. We also contribute a new dataset consisting of 768 DFI with GON labels as open access.

Hagai Hamami, Yosef Solewicz, Daniel Zur et al.

Introduction: 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. Methods: 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. Results: 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%. Conclusion: uPVC-Net exhibits strong generalization across diverse lead configurations and populations, highlighting its potential for robust, real-world clinical deployment.

Benjamin A. Cohen, Jonathan Fhima, Meishar Meisel et al.

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. 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. Our results show that iBOT pretrained on natural images achieves the highest out-of-distribution generalization, with AUROCs of 0.80-0.97, outperforming domain-specific models, which achieved AUROCs of 0.78-0.96 and a baseline ViT-L with no pretraining, which achieved AUROCs of 0.68-0.91. These findings highlight the value of foundation models in improving AMD identification and challenge the assumption that in-domain pretraining is necessary. Furthermore, we release BRAMD, an open-access dataset (n=587) of DFIs with AMD labels from Brazil.

Márton Á. Goda, Helen Badge, Jasmeen Khan et al.

Objective. Large vessel occlusion (LVO) stroke presents a major challenge in clinical practice due to the potential for poor outcomes with delayed treatment. Treatment for LVO involves highly specialized care, in particular endovascular thrombectomy, and is available only at certain hospitals. Therefore, prehospital identification of LVO by emergency ambulance services, can be critical for triaging LVO stroke patients directly to a hospital with access to endovascular therapy. Clinical scores exist to help distinguish LVO from less severe strokes, but they are based on a series of examinations that can take minutes and may be impractical for patients with dementia or those who cannot follow commands due to their stroke. There is a need for a fast and reliable method to aid in the early identification of LVO. In this study, our objective was to assess the feasibility of using 30-second photoplethysmography (PPG) recording to assist in recognizing LVO stroke. Method. A total of 88 patients, including 25 with LVO, 27 with stroke mimic (SM), and 36 non-LVO stroke patients (NL), were recorded at the Liverpool Hospital emergency department in Sydney, Australia. Demographics (age, sex), as well as morphological features and beating rate variability measures, were extracted from the PPG. A binary classification approach was employed to differentiate between LVO stroke and NL+SM (NL.SM). A 2:1 train-test split was stratified and repeated randomly across 100 iterations. Results. The best model achieved a median test set area under the receiver operating characteristic curve (AUROC) of 0.77 (0.71--0.82). \textit{Conclusion.} Our study demonstrates the potential of utilizing a 30-second PPG recording for identifying LVO stroke.

Kevin Kotzen, Peter H. Charlton, Sharon Salabi et al.

Introduction: Sleep staging is an essential component in the diagnosis of sleep disorders and management of sleep health. It is traditionally measured in a clinical setting and requires a labor-intensive labeling process. We hypothesize that it is possible to perform robust 4-class sleep staging using the raw photoplethysmography (PPG) time series and modern advances in deep learning (DL). Methods: We used two publicly available sleep databases that included raw PPG recordings, totalling 2,374 patients and 23,055 hours. We developed SleepPPG-Net, a DL model for 4-class sleep staging from the raw PPG time series. SleepPPG-Net was trained end-to-end and consists of a residual convolutional network for automatic feature extraction and a temporal convolutional network to capture long-range contextual information. We benchmarked the performance of SleepPPG-Net against models based on the best-reported state-of-the-art (SOTA) algorithms. Results: When benchmarked on a held-out test set, SleepPPG-Net obtained a median Cohen's Kappa ($κ$) score of 0.75 against 0.69 for the best SOTA approach. SleepPPG-Net showed good generalization performance to an external database, obtaining a $κ$ score of 0.74 after transfer learning. Perspective: Overall, SleepPPG-Net provides new SOTA performance. In addition, performance is high enough to open the path to the development of wearables that meet the requirements for usage in clinical applications such as the diagnosis and monitoring of obstructive sleep apnea.

Jeremy Levy, Daniel Alvarez, Felix del Campo et al.

Objective: Chronic obstructive pulmonary disease (COPD) is a highly prevalent chronic condition. COPD is a major source of morbidity, mortality and healthcare costs. Spirometry is the gold standard test for a definitive diagnosis and severity grading of COPD. However, a large proportion of individuals with COPD are undiagnosed and untreated. Given the high prevalence of COPD and its clinical importance, it is critical to develop new algorithms to identify undiagnosed COPD, especially in specific groups at risk, such as those with sleep disorder breathing. To our knowledge, no research has looked at the feasibility of COPD diagnosis from the nocturnal oximetry time series. Approach: We hypothesize that patients with COPD will exert certain patterns and/or dynamics of their overnight oximetry time series that are unique to this condition. We introduce a novel approach to nocturnal COPD diagnosis using 44 oximetry digital biomarkers and 5 demographic features and assess its performance in a population sample at risk of sleep-disordered breathing. A total of n=350 unique patients polysomnography (PSG) recordings. A random forest (RF) classifier is trained using these features and evaluated using the nested cross-validation procedure. Significance: Our research makes a number of novel scientific contributions. First, we demonstrated for the first time, the feasibility of COPD diagnosis from nocturnal oximetry time series in a population sample at risk of sleep disordered breathing. We highlighted what digital oximetry biomarkers best reflect how COPD manifests overnight. The results motivate that overnight single channel oximetry is a valuable pathway for COPD diagnosis.