Publications by authors named "Gil Shamai"
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.
View Article and Find Full Text PDFArticle Synopsis
- PD-L1 is now recognized as a crucial biomarker for predicting responses to immunotherapy in breast cancer, but current methods for measuring it through immunohistochemistry are expensive and inconsistent.
- Researchers have developed a deep learning approach to predict PD-L1 expression by analyzing standard H&E-stained images, which are commonly used in cancer diagnosis.
- In a study of 3,376 patients, the new system demonstrated high predictive accuracy, validated on multiple datasets, and can also identify cases at risk of misinterpretation by pathologists, enhancing clinical decision-making and quality assurance.
Importance: Immunohistochemistry (IHC) is the most widely used assay for identification of molecular biomarkers. However, IHC is time consuming and costly, depends on tissue-handling protocols, and relies on pathologists' subjective interpretation. Image analysis by machine learning is gaining ground for various applications in pathology but has not been proposed to replace chemical-based assays for molecular detection.
View Article and Find Full Text PDFEfficient Inter-Geodesic Distance Computation and Fast Classical Scaling.
IEEE Trans Pattern Anal Mach Intell
January 2020
Article Synopsis
- Multidimensional scaling (MDS) is a tool that reduces data dimensions for analysis and visualization, aiming to keep distances between points close to their original differences.
- A new efficient solver for classical scaling is introduced, which utilizes distance information from a subset of points to optimize the calculations for the full dataset.
- This method significantly lowers the computational complexity from quadratic to quasi-linear, while also improving the calculation of geodesic distances in the process.