IBM Research - Israel

Accelerated Discovery

We are developing causal inference technologies and multimodal learning models that leverage the most advanced deep learning and generative models, aimed at accelerating molecules and biomarker discovery.

 

Open Source


Causal Inference: CausalLib

Multimodal Deep Learning Over Patient Data: FuseMedML

News and Blogs

BICauseTree: Bias-balancing Interpretable Causal Tree

In this blog post, ISRL intern Lucile Ter-Minassian outlines a binary decision tree model for causal inference that interprets Average Treatment Effects (ATEs) without bias from high-dimensional datasets.

Read More

Impact of COVID-19 on Clinical Findings in Medical Imaging Exams

This study, coauthored by ISRL researchers, examines the increase in the rate of abnormal findings in MR-MSK and CT-brain exams after the onset of the COVID-19 pandemic.

Read More

IBM and Cleveland Clinic unveil the first quantum computer dedicated to healthcare research

Research projects by ISRL researchers Yishai Shimoni, Michael Danziger, and Liran Szlak were profiled in this blog post about IBM's partnership with Cleveland Clinic.

Read More

Causal Inference is a Mindset

Researcher Ehud Karavani blogs about the philosophical challenges and a myriad of practical tools as we strive for causal inference from observational data.

Read More

Research Projects

Causal Machine Learning for Healthcare & Life Sciences

The Causal Machine-Learning for Healthcare and Life-Sciences team includes researchers from the fields of physics, computer science, statistics, and epidemiology. We conduct analyses for many types of healthcare-related data, including electronic health records and insurance claims data holding the medical history of over 150 million people.
More recently, our efforts are focused on causal inference, which answers the question "what is the effect of doing something?" We used this technology to create an efficient engine to identify drug-repurposing candidates and have released the base causal inference technology as open-source code. In general, we develop technology that allows the acceleration of meaningful clinical discoveries based on real-world data.

Multimodal AI for Healthcare & Life Sciences

Our research includes advanced computer vision techniques that enable the automatic extraction of diagnostically relevant features in multimodal healthcare images. We are developing machine learning tools that combine multimodal semantic image descriptions (for mammography, ultrasound and MRI) with clinical data to help estimate a correct differential diagnosis and support patient management recommendations.

Publications

IBM researchers in Israel publish a wide variety of work every year as part of their work on research projects in the lab, in collaboration with other researchers and scientists in IBM, and together with academic and industrial partners from around the world.

Researchers in our group publish works at conferences and in scientific journals such as the SPIE Medical Imaging conferences, Frontiers in Pharmacology, Radiology, and others.

Explore

Tools & Code

Causal Inference 360 Open Source Toolkit

This open source Python toolkit is designed to bring long-standing machine-learning methodologies to the field of causal inference.

View project

fuse-med-ml

FuseMedML is an open-source Python-based framework designed to enhance collaboration and accelerate discoveries in fused medical data through advanced machine learning technologies. This initial version is PyTorch-based and focuses on deep learning for medical imaging and digital pathology.

View project

Academic Collaboration

Collaborate with our researchers in accelerating scientific discoveries in the medical domain - from drug discovery to biomarker discovery - using AI-based technology applied to health, chemical, and molecular data, embedded with medical knowledge.

Explore

Let's talk

We're always happy to talk. Feel free to get in touch.

 


Manager, Accelerated Discovery, IBM Research - Israel