This project aims to enhance pre-operative planning for complex pancreatic cancer surgeries by leveraging machine learning (ML) techniques. Pancreatic cancer is highly aggressive, with significant risks of post-operative complications. Sarcopenia, the loss of muscle mass, is a critical risk factor that adversely affects recovery outcomes.
The project employs ML algorithms to analyze pre-operative CT imaging using radiomics. Detailed features related to pancreatic structure and patient body composition, including muscle mass, are extracted and integrated with clinical data to develop predictive models. These models assess the risk of complications, such as pancreatic fistulas, and identify sarcopenia with greater accuracy than traditional methods.
Preliminary results suggest these predictive tools enable more informed and personalized surgical planning. For instance, identifying sarcopenia could lead to pre-operative interventions like nutritional and physical rehabilitation programs to improve muscle mass and mitigate risks.
While these AI-based tools provide invaluable support for surgical risk assessment, they are designed to complement, not replace, the clinical judgment of surgeons. The ultimate goal is to improve surgical safety and outcomes by advancing precision medicine tailored to individual patient conditions.