Your Donations in Action: Pankaj Gupta, MD

Unresectable gallbladder cancer (GBC) is often diagnosed at an advanced stage, leaving patients with limited treatments and poor survival outcomes, particularly in resource-limited regions where the disease is most prevalent. While recent research has identified HER2 status as a frequent molecular biomarker for GBC, accessing this information typically requires invasive and costly tissue biopsies. Anti-HER2 therapy has demonstrated improved response rates compared to conventional chemotherapy.

For his 2022 RSNA Research Seed Grant, “Deep Learning Radiogenomics for Individualized Therapy in Unresectable Gallbladder Cancer,” Pankaj Gupta, MD, associate professor in the Department of Radiodiagnosis at the Postgraduate Institute of Medical Education and Research in Chandigarh, India, and colleagues explored whether a fully automated, CT-based deep learning pipeline could predict HER2 status in unresectable GBC.

The team conducted a combined retrospective and prospective study involving 214 patients with advanced GBC, with HER2 status identified using standard pathology techniques. The researchers then trained and validated machine learning, radiomics and deep learning models using portal venous-phase CT images.

After testing multiple approaches, a tumor detection and segmentation framework paired with deep learning classifiers achieved high predictive accuracy, with sensitivity reaching up to 75% and specificity up to 86% across tested models. Automated tumor localization enabled an efficient pipeline requiring minimal human input.

The noninvasive radiogenomic approach outperformed traditional clinical and radiomics models, demonstrating that AI can effectively decode molecular characteristics from routine CT scans and support molecular subtyping when tissue biopsies are limited.

“In resource-limited settings or cases where tissue sampling is insufficient or risky, this AI model can rapidly identify patients with a high probability of being HER2-positive, thereby expediting their access to targeted therapies,” Dr. Gupta said.

The R&E grant allowed the work to transition from a concept to a funded, multifaceted research program with a potential for global impact.

“The preliminary data generated from this project provided the proof-of-concept necessary to secure a substantial grant from the Department of Biotechnology, Government of India, to expand our work into chemotherapy response prediction,” Dr. Gupta said. “The success of this project has opened doors for international collaborations in hepatobiliary cancer research.”

For More Information

Learn more about R&E Foundation funding opportunities.

Read our previous Your Donations in Action story.