Madison Reporter

Researchers at the University of Wisconsin–Madison are employing artificial intelligence to identify sex-specific risk factors associated with glioblastoma, a deadly form of brain cancer. This initiative aims to uncover characteristics that could help doctors predict tumor growth rates and improve patient care.

Professor Pallavi Tiwari, from the departments of radiology and biomedical engineering, along with her colleagues, has published their findings in Science Advances. "There’s a ton of data collected in a cancer patient’s journey," Tiwari stated. She emphasized the potential of AI in studying this data comprehensively rather than in isolation.

Tiwari joined UW–Madison in 2022 to lead its AI initiative in medical imaging and co-directs the Imaging and Radiation Sciences Program at the Carbone Cancer Center. Her research uses AI models to analyze large volumes of medical images for patterns that could assist oncologists.

"We want to address the entire spectrum of challenges in a cancer patient’s journey," Tiwari said, highlighting their focus on diagnosis, prognosis, and treatment response assessment.

In collaboration with former graduate student Ruchika Verma, Tiwari analyzed digital images of pathology slides from over 250 studies on glioblastoma patients. They trained an AI model to detect unique tumor characteristics and correlate them with survival times while considering sex differences.

The study found that higher-risk tumors in females were often infiltrating healthy tissue, whereas males showed more aggressive tumors associated with pseudopalisading cells surrounding dying tissue. The model also identified tumor traits linked to poor prognoses for both sexes.

"By uncovering these unique patterns, we hope to inspire new avenues for personalized treatment," Verma remarked.

Tiwari's team is expanding their work using MRI data and applying AI analysis to pancreatic and breast cancers. Additionally, she contributes to UW–Madison's RISE-AI and RISE-THRIVE initiatives aimed at advancing cross-disciplinary research on artificial intelligence.

"UW has a rich and diverse expertise across our engineering and medical campuses," said Tiwari. She noted that these initiatives position the university as a leader in translating AI research into clinical care improvements.