In a significant breakthrough for cancer care, researchers at the Indian Institute of Technology Hyderabad (IITH) have unveiled a novel web-based tool capable of estimating the survival risk of brain cancer patients by analysing their MRI scans. Named RadGLO (Radiology of Glioma), this innovative platform leverages advanced computer analysis to uncover hidden patterns within medical images that are typically invisible to the human eye.
A Non-Invasive Support for Complex Diagnoses
RadGLO presents a promising, non-invasive alternative in situations where traditional biopsy or molecular testing is challenging or poses a risk to the patient. Rahul Kumar, the project's principal investigator and an assistant professor at IITH, explained that while biopsy remains the definitive diagnostic standard, imaging-based tools like RadGLO can offer rapid insights. These insights empower doctors to devise early and personalised treatment strategies. The tool calculates a patient's probability of survival, providing crucial data to help clinicians determine the most appropriate therapeutic path.
How RadGLO Transforms MRI Data into Predictions
Although MRI scans are routinely used to detect brain tumours, the IITH team's work demonstrates that medical imaging can go further, predicting a tumour's aggressiveness and a patient's likely response over time. RadGLO operates on the principles of radiomics, an emerging field that extracts vast amounts of quantifiable data from medical images. Doctors and researchers can upload MRI scans of glioma patients to the platform, which then uses a prediction model to deliver an estimated survival risk.
The models powering RadGLO were rigorously validated using two large, independent datasets. They demonstrated strong accuracy in forecasting survival risk and classifying tumours across diverse patient populations. The research detailing these findings has been published in the prestigious journal, npj Precision Oncology.
Testing and Future Development with Indian Data
Kavitha Kundal, a research scholar on the project, noted a key next step. The initial validation datasets were sourced from the United States. To ensure the tool's efficacy for the Indian population, the team has partnered with Hyderabad's MNJ Cancer Hospital. They have already begun collecting Indian patient data, with 50 MRI scans received and a target of at least 100 more in the coming months. The project also enjoys collaboration with AIIMS Delhi.
Looking ahead, the IITH researchers plan to enhance RadGLO's capabilities by integrating genetic, molecular, and tissue-based data. The ultimate goal is to create a more comprehensive tool that advances the field of precision oncology, paving the way for truly individualised cancer treatment plans.
