GSGil Shamai

Gil Shamai

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Gil Shamai has a total of 33 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Prediction of OncotypeDX recurrence score using hematoxylin and eosin-stained whole slide images

    by Shachar Cohen, Gil Shamai, Edmond Sabo, Alexandra Cretu, Iris Barshack, Tal Goldman, Gil Bar-Sela, Alexander T. Pearson, Dezheng Huo, Frederick M. Howard, et al.

    The OncotypeDX 21-gene assay guides adjuvant chemotherapy decisions in early-stage, hormone receptor–positive, HER2-negative breast cancer, but cost and turnaround time limit access. This study presents a deep learning-based approach for predicting OncotypeDX recurrence scores directly from hematoxylin and eosin-stained whole slide images. Our approach leverages a deep learning foundation model pre-trained on 171,189 slides via self-supervised learning, which is fine-tuned for our task. The model was developed and validated using five independent cohorts, out of which three are external. On the two external cohorts that include OncotypeDX scores, the model achieved an AUC of 0.836 and 0.817, and identified 22% and 16.3% of the patients as low-risk with sensitivity of 0.97 and 0.97 and negative predictive value of 0.97 and 0.96, showing strong generalizability despite variations in staining protocols and imaging devices. Kaplan-Meier analysis demonstrated that patients classified as low-risk by the model had a significantly better prognosis than those classified as high-risk, with a hazard ratio of 4.1 ( P < 0.001) and 2.0 ( P < 0.01) on the two external cohorts that include patient outcomes. This artificial intelligence-driven solution offers a rapid, cost-effective, and scalable alternative to genomic testing, with the potential to enhance personalized treatment planning, especially in resource-constrained settings.


  • Open Access

    Prediction of OncotypeDX recurrence score using H&E stained WSI images

    by Shachar Cohen, Gil Shamai, Edmond Sabo, Alexandra Cretu, Iris Barshack, Tal Goldman, Gil Bar-Sela, Alexander T. Pearson, Dezheng Huo, Frederick M. Howard, et al.

    The OncotypeDX 21-gene assay is a widely adopted tool for estimating recurrence risk and informing chemotherapy decisions in early-stage, hormone receptor-positive, HER2-negative breast cancer. Although informative, its high cost and long turnaround time limit accessibility and delay treatment in low- and middle-income countries, creating a need for alternative solutions. This study presents a deep learning-based approach for predicting OncotypeDX recurrence scores directly from hematoxylin and eosin-stained whole slide images. Our approach leverages a deep learning foundation model pre-trained on 171,189 slides via self-supervised learning, which is fine-tuned for our task. The model was developed and validated using five independent cohorts, out of which three are external. On the two external cohorts that include OncotypeDX scores, the model achieved an AUC of 0.825 and 0.817, and identified 21.9% and 25.1% of the patients as low-risk with sensitivity of 0.97 and 0.95 and negative predictive value of 0.97 and 0.96, showing strong generalizability despite variations in staining protocols and imaging devices. Kaplan-Meier analysis demonstrated that patients classified as low-risk by the model had a significantly better prognosis than those classified as high-risk, with a hazard ratio of 4.1 (P<0.001) and 2.0 (P<0.01) on the two external cohorts that include patient outcomes. This artificial intelligence-driven solution offers a rapid, cost-effective, and scalable alternative to genomic testing, with the potential to enhance personalized treatment planning, especially in resource-constrained settings.


  • Open Access

    Prediction of B/T Subtype and ETV6–RUNX1 Translocation in Pediatric Acute Lymphoblastic Leukemia by Deep Learning Analysis of Giemsa‐Stained Whole Slide Images of Bone Marrow Aspirates

    by Arkadi Piven, Gil Shamai, Sarah Elitzur, Galit Pinto Berger, Yoav Binenbaum, Ron Kimmel, Ronit Elhasid

    Background: Accurate determination of B/T‐cell lineage and the presence of the ETV6–RUNX1 translocation is critical for diagnosing acute lymphoblastic leukemia (ALL), as these factors influence treatment decisions and outcomes. However, these diagnostic processes often rely on advanced tools unavailable in low‐resource settings, creating a need for alternative solutions.Procedure: We developed a deep learning pipeline to analyze Giemsa‐stained bone marrow (BM) aspirate smears. The models were trained to distinguish between ALL, acute myeloid leukemia (AML), and non‐leukemic BM samples, predict B‐ and T‐cell lineage in ALL, and detect the presence of the ETV6–RUNX1 translocation. The performance was evaluated using cross‐validation (CV) and an external validation cohort.Results: The models achieved a statistically significant area under the curve (AUC) of 0.99 in distinguishing ALL from AML and control samples. In cross‐validation (CV), the models achieved a cross‐validation AUC of 0.74 for predicting B/T subtypes. For predicting ETV6–RUNX1 translocation, the models achieved an AUC of 0.80. External cohort validation confirmed significant AUCs of 0.72 for B/T subtype classification and 0.69 for ETV6–RUNX1 translocation prediction.Conclusions: Convolutional neural networks (CNNs) demonstrate potential as a diagnostic tool for pediatric ALL, enabling the identification of B/T lineage and ETV6–RUNX1 translocation from Giemsa‐stained smears. These results pave the way for future utilization of CNNs as a diagnostic modality for pediatric leukemia in low‐resource settings, where access to advanced diagnostic techniques is limited.


  • Open Access

    Deep Learning on Histopathological Images to Predict Breast Cancer Recurrence Risk and Chemotherapy Benefit

    by Gil Shamai, Shachar Cohen, Yoav Binenbaum, Edmond Sabo, Alexandra Cretu, Chen Mayer, Iris Barshack, Tal Goldman, Gil Bar-Sela, António Polónia, et al.

    Genomic testing has transformed treatment decisions for hormone receptor-positive, HER2-negative (HR+/HER2-) early breast cancer; however, it remains inaccessible to many patients worldwide due to high costs and logistical barriers. Here, we developed an artificial intelligence (AI) model using a multimodal deep learning approach that estimates Oncotype DX 21-gene recurrence scores (RS) from routine histopathology images and clinicopathologic variables, including age at diagnosis, tumor size, and receptor status. Using a foundation model pre-trained on 171,189 histopathological slides, we fine-tuned and validated our AI model on the TAILORx randomized trial (n=8,284). Among 2,407 patients in the TAILORx validation, the model classifies 45.6% of patients as low-risk, 42.4% as intermediate risk, and 12.0% as high-risk. For predicting high genomic risk disease (RS≥26), occurring in 15.9% in the TAILORx validation set, the model achieves AUC=0.898. Patient stratification by our model shows strong prognostic value across multiple clinical endpoints, including recurrence-free interval, distant recurrence-free interval, and disease-free survival. Importantly, chemotherapy benefit is demonstrated for premenopausal patients classified by our model as high AI risk and chemotherapy benefit is ruled out for postmenopausal patients classified as low AI risk. External validation across six independent cohorts (n=5,497 patients) demonstrates robust generalization of the AI model for prognostication and prediction of RS. Notably, in postmenopausal patients, the AI model reclassifies approximately 30% of clinically high-risk cases, defined by the MINDACT criteria, as low-risk. These findings demonstrate that artificial intelligence applied to standard histopathology can be a valuable tool for chemotherapy decision-making in HR+/HER2- early breast cancer. This approach can help reduce unnecessary chemotherapy and extend precision medicine, particularly in resource-limited settings, where genomic testing is not widely accessible.


  • Open Access

    Clinical utility of receptor status prediction in breast cancer and misdiagnosis identification using deep learning on hematoxylin and eosin-stained slides

    by Gil Shamai, Ran Schley, Alexandra Cretu, Tal Neoran, Edmond Sabo, Yoav Binenbaum, Shachar Cohen, Tal Goldman, António Polónia, Keren Drumea, et al.

    Background Molecular profiling of estrogen receptor (ER), progesterone receptor (PR), and ERBB2 (also known as Her2) is essential for breast cancer diagnosis and treatment planning. Nevertheless, current methods rely on the qualitative interpretation of immunohistochemistry and fluorescence in situ hybridization (FISH), which can be costly, time-consuming, and inconsistent. Here we explore the clinical utility of predicting receptor status from digitized hematoxylin and eosin-stained (H&E) slides using machine learning trained and evaluated on a multi-institutional dataset. Methods We developed a deep learning system to predict ER, PR, and ERBB2 statuses from digitized H&E slides and evaluated its utility in three clinical applications: identifying hormone receptor-positive patients, serving as a second-read tool for quality assurance, and addressing intratumor heterogeneity. For development and validation, we collected 19,845 slides from 7,950 patients across six independent cohorts representative of diverse clinical settings. Results Here we show that the system identifies 30.5% of patients as hormone receptor-positive, achieving a specificity of 0.9982 and a positive predictive value of 0.9992, demonstrating its ability to determine eligibility for hormone therapy without immunohistochemistry. By restaining and reassessing samples flagged as potential false negatives, we discover 31 cases of misdiagnosed ER, PR, and ERBB2 statuses. Conclusions These findings demonstrate the utility of the system in diverse clinical settings and its potential to improve breast cancer diagnosis. Given the substantial focus of current guidelines on reducing false negative diagnoses, this study supports the integration of H&E-based machine learning tools into workflows for quality assurance.


  • Top Cited Papers
    Open Access

    Deep learning-based image analysis predicts PD-L1 status from H&E-stained histopathology images in breast cancer

    by Gil Shamai, Amir Livne, António Polónia, Edmond Sabo, Alexandra Cretu, Gil Bar-Sela, Ron Kimmel

    Programmed death ligand-1 (PD-L1) has been recently adopted for breast cancer as a predictive biomarker for immunotherapies. The cost, time, and variability of PD-L1 quantification by immunohistochemistry (IHC) are a challenge. In contrast, hematoxylin and eosin (H&E) is a robust staining used routinely for cancer diagnosis. Here, we show that PD-L1 expression can be predicted from H&E-stained images by employing state-of-the-art deep learning techniques. With the help of two expert pathologists and a designed annotation software, we construct a dataset to assess the feasibility of PD-L1 prediction from H&E in breast cancer. In a cohort of 3,376 patients, our system predicts the PD-L1 status in a high area under the curve (AUC) of 0.91 – 0.93. Our system is validated on two external datasets, including an independent clinical trial cohort, showing consistent prediction performance. Furthermore, the proposed system predicts which cases are prone to pathologists miss-interpretation, showing it can serve as a decision support and quality assurance system in clinical practice.


  • Open Access

    Synthesizing Facial Photometries and Corresponding Geometries Using Generative Adversarial Networks

    by Gil Shamai, Ron Slossberg, Ron Kimmel

    Artificial data synthesis is currently a well-studied topic with useful applications in data science, computer vision, graphics, and many other fields. Generating realistic data is especially challenging, since human perception is highly sensitive to non-realistic appearance. In recent times, new levels of realism have been achieved by advances in GAN training procedures and architectures. These successful models, however, are tuned mostly for use with regularly sampled data such as images, audio, and video. Despite the successful application of the architecture on these types of media, applying the same tools to geometric data poses a far greater challenge. The study of geometric deep learning is still a debated issue within the academic community, as the lack of intrinsic parametrization inherent to geometric objects prohibits the direct use of convolutional filters, a main building block of today’s machine learning systems. In this article, we propose a new method for generating realistic human facial geometries coupled with overlayed textures. We circumvent the parametrization issue by utilizing a specialized non-rigid alignment procedure, and imposing a global mapping from our data to the unit rectangle. This mapping enables the representation of our geometric data as regularly sampled 2D images. We further discuss how to design such a mapping to control the distortion and conserve area within the target image. By representing geometric textures and geometries as images, we are able to use advanced GAN methodologies to generate new plausible textures and geometries. We address the often-neglected topic of relationship between texture and geometry and propose different methods for fitting generated geometries to generated textures. In addition, we widen the scope of our discussion and offer a new method for training GAN models on partially corrupted data. Finally, we provide empirical evidence demonstrating our generative model’s ability to produce examples of new facial identities, independent from the training data, while maintaining a high level of realism—two traits that are often at odds.


  • Open Access

    Synthesizing facial photometries and corresponding geometries using generative adversarial networks

    by Gil Shamai, Ron Slossberg, Ron Kimmel

    Artificial data synthesis is currently a well studied topic with useful applications in data science, computer vision, graphics and many other fields. Generating realistic data is especially challenging since human perception is highly sensitive to non realistic appearance. In recent times, new levels of realism have been achieved by advances in GAN training procedures and architectures. These successful models, however, are tuned mostly for use with regularly sampled data such as images, audio and video. Despite the successful application of the architecture on these types of media, applying the same tools to geometric data poses a far greater challenge. The study of geometric deep learning is still a debated issue within the academic community as the lack of intrinsic parametrization inherent to geometric objects prohibits the direct use of convolutional filters, a main building block of today's machine learning systems. In this paper we propose a new method for generating realistic human facial geometries coupled with overlayed textures. We circumvent the parametrization issue by imposing a global mapping from our data to the unit rectangle. We further discuss how to design such a mapping to control the mapping distortion and conserve area within the mapped image. By representing geometric textures and geometries as images, we are able to use advanced GAN methodologies to generate new geometries. We address the often neglected topic of relation between texture and geometry and propose to use this correlation to match between generated textures and their corresponding geometries. We offer a new method for training GAN models on partially corrupted data. Finally, we provide empirical evidence demonstrating our generative model's ability to produce examples of new identities independent from the training data while maintaining a high level of realism, two traits that are often at odds.

  • Efficient Inter-Geodesic Distance Computation and Fast Classical Scaling

    by Gil Shamai, Michael Zibulevsky, Ron Kimmel

    Multidimensional scaling (MDS) is a dimensionality reduction tool used for information analysis, data visualization and manifold learning. Most MDS procedures embed data points in low-dimensional Euclidean (flat) domains, such that distances between the points are as close as possible to given inter-point dissimilarities. We present an efficient solver for classical scaling, a specific MDS model, by extrapolating the information provided by distances measured from a subset of the points to the remainder. The computational and space complexities of the new MDS methods are thereby reduced from quadratic to quasi-linear in the number of data points. Incorporating both local and global information about the data allows us to construct a low-rank approximation of the inter-geodesic distances between the data points. As a by-product, the proposed method allows for efficient computation of geodesic distances.