Presentation Title: An AI-enabled Study of Osteosarcopenia Progression in Prostate Cancer
Abstract: Osteosarcopenia, a condition that combines osteopenia and sarcopenia, poses significant challenges to the musculoskeletal health of prostate cancer patients, especially due to the prolonged impact of disease and treatment. Despite its prevalence, the combined effects of osteopenia and sarcopenia have been insufficiently explored, with existing studies often relying on single time-point evaluations. This study aims to quantify the progression of osteosarcopenia in patients with advanced prostate cancer by utilizing artificial intelligence (AI) to analyze electronic medical records (EMR) and imaging biomarkers from serial 3D CT scans. The hypothesis is that AI-enhanced analysis can predict osteosarcopenia progression and evaluate the impact of cancer treatments on both bone and muscle health. The methodology involves a retrospective analysis of medical imaging and EMR from over 1,500 patients across multiple international cancer centers. A biomarker extraction pipeline, employing advanced deep learning techniques, is developed to assess osteosarcopenia in CT scans. Predictive modeling utilizes time-series analysis and deep neural networks to understand disease progression and treatment effects. This research seeks to transform osteosarcopenia management by providing robust progression quantification and insights into the interplay between sarcopenia and osteopenia, ultimately improving diagnosis and treatment strategies.
Supervisor Name: Cari Whyne, Michael Hardisty
Year of Study: 3
Program of Study: PhD