Skip to main content

Raneem Toman

Research Topic: Deep learning-based multi-modal data correlation for stratifying patients with IBD into low-risk and high-risk categories of developing colorectal cancer.

Supervisors: Dr Sharib Ali, Prof Animesh Jha, Dr Venkataraman Subramanian.

About Raneem: I did my BSc in Physics at Al-Quds University in Palestine. I then trained as a junior researcher in computational neuroscience and electrophysiology, which led me to do an MSc in Biomedical Engineering at the University of Warwick. Upon graduation, I gained industrial experience working at a medical supplies and services company.

Project Description: Early detection and diagnosis of Colorectal Cancer (CRC) can significantly reduce its related deaths globally. Inflammatory Bowel Disease (IBD) patients are at 6 times higher risk for developing CRC, and would also have a worse prognosis compared to non-IBD CRC. These ramifications impair IBD patients’ quality of life and increase the economic burden on healthcare systems. However, there is still high operator-dependency in identifying precancerous lesions, so AI models have been developed to address this issue using imaging data from colonoscopies. However, this task cannot be achieved solely relying on one specific data type, as the molecular structure of colonic lesions can hold significant information about the stage and severity of IBD.
This project will aim to use deep learning to correlate the phenotypic features in imaging data and molecular data with clinical outcomes for minimised subjectivity in polyp detection and characterisation. The idea of leveraging different modalities (spectroscopy, imaging) and building dictionaries around these modalities will be key in understanding what mucosal changes have what likelihood of developing into cancer. This project will therefore open novel understanding of cancer progression and its prediction.