Mr
George Batchkala
Doctoral Student
Email:
College: Linacre College
Location: Institute of Biomedical Engineering

George Batchkala is a DPhil candidate in Health Data Science at Oxford, funded by Professor Fergus Gleeson. He focuses on applications of computer vision and machine learning to medical imaging, particularly to early lung cancer diagnostics. George works on digital pathology and radiomics modelling under the supervision of Prof Jens Rittscher and Dr Tapabrata (Rohan) Chakraborty.

Before joining the CDT, George received a Bachelors (undergraduate) degree in Mathematics and Statistics from the University of Warwick and his Masters in Statistical Science degree from the University of Oxford. At the end of his BSc, George conducted a funded summer research project via participating in the Ultrasound Nerve Segmentation competition on Kaggle. He developed a customisable model with 160 configurations, achieved a score in the top 10% of participants and received Kaggle-notebooks and discussions bronze medals for his contributions to the Kaggle community.

George completed his MSc dissertation in an industry collaboration with Exscientia – a leading Oxford drug-discovery startup. He worked on machine learning models for small-molecule property prediction and methods of estimating the uncertainty of these models. This research showed that predictions and uncertainty estimates could be used together to select candidate molecules with desired properties.

George Batchkala currently works on Digital pathology AI & radiomics model development (DART’s Work Package 4). During his DPhil, he will participate in the AI model validation on LCS data (WP3) and potentially in the Integration of blood Biomarkers (WP7).

  • Applications of computer vision and NLP to lung-cancer diagnostics
  • Improving patient pathway

Biomedical Image Analysis

  • DART, Digital pathology AI & radiomics model development (work package 4)
    This work package improves digital pathology Artificial Intelligence (AI) interpretation of nodule and cancer histology and develops a radiomics model.