CRASH Lab · Ashoka University
Active research projects, benchmarks, and accepted papers — with evaluation grounded in clinical reality, not presentation-friendly metrics alone.
The flagship RadLE benchmark comparing frontier multimodal AI systems with human radiology expertise and defining a taxonomy of visual reasoning errors in radiology.
Authors: Suvrankar Datta, Divya Buchireddygari, Lakshmi Vennela Chowdary Kaza, Mrudula Bhalke, Kautik Singh, Ayush Pandey, Sonit Sai Vasipalli, Upasana Karnwal, +19 more
A global multimodal dataset of clinical reasoning traces and visual attention annotations for chest X-ray interpretation, developed by the Global Radiology Consortium with CRASH Lab contributing to the Indian consortium.
Authors: Sonali Sharma, Jin Long, George Shih, Sarah Eid, Christian Bluethgen, Francine L. Jacobson, Emily B. Tsai, Global Radiology Consortium, +2 more
Multidimensional evaluation framework for autonomous optimization pipelines in personalized head CT report generation.
Comprehensive evaluation framework for multimodal foundation models in chest radiograph analysis with radiologist-blinded validation.
Scalable autonomous optimization pipeline focused on style-aware improvements in head CT report generation.
Autonomous optimization pipeline for hyper-personalized chest X-ray report generation with quality improvements.
Novel assessment scale for evaluating reliability and referencing integrity in AI-generated radiology reports.
Analysis of inter-reader agreement and draft variability using RADAR and TRUST metrics in agentic radiology reporting.
We began securing support from leading national bodies, including ANRF, DST, and the IndiaAI Mission, to advance responsible healthcare AI research.
Supported by Koita Foundation
Funding for the lab is provided by the Koita Foundation and Google, alongside contributions from institutional partners and private philanthropy
