Fusion of Clinical and Deep Learning Features for Predicting Pembrolizumab Monotherapy Response in Advanced Non-Small Cell Lung Cancer
Liton Devnath, Ian Janzen, Cheryl Ho, Barbara Melosky, Stephen Lam, Calum MacAulay, Ren Yuan · 2026
Read Fusion of Clinical and Deep Learning Features for Predicting Pembrolizumab Monotherapy Response in Advanced Non-Small Cell Lung Cancer by Liton Devnath, Ian Janzen, Cheryl Ho, Barbara Melosky, Stephen Lam, Calum MacAulay, Ren Yuan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13300846.txt.
Description
Liton Devnath and co-authors combine deep learning image features with clinical data to predict who will not respond to pembrolizumab alone in advanced lung cancer, in a 2026 study. Over half of patients with high PD-L1 expression still fail this first-line immunotherapy, so early warning would speed a switch to other treatment. The retrospective cohort held ninety-seven patients for training and seventeen for a preliminary test, with responses judged from follow-up scans.
From 128 image features, the team ranked the ten most discriminative and paired them with five clinical descriptors in a support vector machine. That blend reached the best balance on the small test set, catching most non-responders while keeping false alarms moderate. Performance looked stronger inside the training data, a gap the authors openly note. They present the fused model as a promising decision aid that needs bigger validation cohorts.
Catalog record
| CREATOR | Liton Devnath, Ian Janzen, Cheryl Ho, Barbara Melosky, Stephen Lam, Calum MacAulay, Ren Yuan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |