Hello, World!
My name is Agam Tomar and I am Research Scientist at Johnson & Johnson working on improving the future of digital surgery. My current focus is on improving the navigation of Monarch platform by generating insights on procedure phases using raw telemetry and object detection models.
I graduated with a double major (M.S. in Electrical and Computer Engineering and Ph.D. in Structural and Earthquake Engineering with minor in Statistics) from University of California - Los Angeles. I specialize in Engineering Applications of Artificial Intelligence. My research focused on modeling risk and resilience of water distribution systems under the supervision of Prof. Henry Burton. Thesis
I spent the summers of 2019 working at Genentech on developing a prognostic deep learner for Chronic Obstructive Pulmonary Disease (COPD) Endpoint. I worked with High-Resolution Computed Tomography (HRCT) scans of COPD patients collected in an internal study (TESRA) conducted by Roche and used Transfer Learning approach to build the prognostic deep learner. The model performed with an accuracy of ~95% in classifying emphysema (COPD endpoint) progression in two years from baseline. This deep learner will be used to enrich future clinical trials. I also worked on creating saliency maps to visualize and understand the working of the deep learning model. Internship Exit Presentation
My research interests include:
- Applications of Artificial Intelligence: Decision Systems, Deep Learning, Computer Vision, Bayesian Inference
- Modeling and Learning from complex systems: Discrete Event Simulation, Bayesian Networks, and Network Analysis
- Big Data Analytics, Machine Learning, and Optimization: Algorithms, Data Analysis and Inference, Convex Optimization