Aurobrata Ghosh

I currently hold the position of Director of Analytics (R&D AI/ML/Computer-Vision) at Claims Insight & Analytics, Verisk Analytics.

I lead technical and research projects in Computer-Vision, AI and ML to enhance business requirements.
My watchword is: "Creative solutions founded in rigorous mathematics", which I try and apply daily in my research work.

I used to work at Verisk, Munich, Germany. But now I am attached to Verisk, Hyderabad, India.

I used to work in the Microstructure Imaging Group (MIG), Centre for Medical Image Computing (CMIC), UCL, UK, as a post-doctoral research associate.

I worked on machine learning (ML) algorithms for Image Quality Transfer (IQT), to enhance clinical MRI acquisitions from high quality datasets acquired on rare, bespoke scanners, such as the Connectome scanner of the Human Connectome Project (HCP). The aim was to connect the two worlds of high-end and expensive research scanners (3T - 7T, 300mT/m) and the modest clinical workhorse (1.5T - 3T, 40mT/m), therefore, bridging the gap and delivering the advantages of these state-of-the-art technical advances to day-to-day applications in hospitals worldwide.

In 2017, my supervisor Prof D. C. Alexander and I, were the recipients of the prestigious Global Challenges Research Fund (GCRF) of the Engineering & Physical Sciences Council (EPSRC), UK. The award led to a research project of 3 years to improve clinical MRI in developing nations using IQT (details).

During my PhD, at Athena, Inria Sophia Antipolis - Méditerranée, France,
I worked on Diffusion MRI (dMRI) signal modelling and data analysis for inferring the microstructure and connectivity of the human brain and the central nervous system.
My PhD supervisor was Prof Rachid Deriche.

I had the privilege to bundle the algorithms I developed for Canon MRI scanners via an industrial collaboration with Oléa Medical (Canon Group).

contact: aurobrata.ghosh<at>gmail.com