Ido Ben Shaul

Title: Ido Ben Shaul and His Patent for Tissue Image Classification

Introduction

Ido Ben Shaul is an inventor based in Raanana, Israel. His patent work includes a method for classifying tissue images using multiple-instance learning.

Latest Patents

Ben Shaul is associated with 1 patent, titled “Multiple Instance Learner for Tissue Image Classification.” The method extracts a feature vector from each tile of an image and uses a multiple-instance-learning model to classify the image into one of at least two classes. It calculates a certainty value for each tile to indicate how much that tile’s features contribute to the classification. The model then uses certainty-based pooling to combine the image’s feature vectors or their predictive values into an aggregated predictive value, which supports the final classification.

Career Highlights

Ben Shaul works at Hoffmann-La Roche Inc. His patent describes an approach to tissue image analysis that accounts for the varying contributions of individual image tiles.

Collaborations

His coworker Jacob Gildenblat is also associated with his professional work.

Conclusion

Ido Ben Shaul’s patent presents a certainty-based multiple-instance-learning method for classifying tissue images, reflecting his work in image analysis at Hoffmann-La Roche Inc.