Ashish Manne, Associate Professor at Ohio State University Comprehensive Cancer Center-James Cancer Hospital and Solove Research Institute, shared a post on LinkedIn:
“AI in Pancreatic Cancer: Excited to share the preprint of our work identifying Moffitt classical vs. basal-like pancreatic cancer subtypes using a clinically deployable AI approach.
Let’s be honest: most AI models being published today are not clinically useful. They are complex, opaque, and optimized for benchmarks, and not for patients, physicians, or real clinical workflows.
In this work, we deliberately push back against that trend. Rather than another ‘fancy’ black-box model, we focus on a clinically deployable, interpretable starting point. The one that respects how oncology is actually practiced and how decisions are made at the bedside.
If AI is going to matter in pancreatic cancer, it has to move beyond publications and start earning its place in clinical care. This study is one small but concrete step in that direction.
A great collaboration with Khalid Niazi.
Fantastic job by Abdul Rehman Akbar and Alejandro Leyva.”
Read the full preprint.
More posts on Pancreatic Cancer.