An segformer-mix vision transformer-based GIST segmentation in fused CT-PET images
Venu Allapakam, Yepuganti Karuna · 2026
Read An segformer-mix vision transformer-based GIST segmentation in fused CT-PET images by Venu Allapakam, Yepuganti Karuna (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/plos-pone-0355109.txt.
Description
Venu Allapakam and Yepuganti Karuna build an automated system for outlining a rare digestive tract tumor in medical scans, in this 2026 paper. These tumors, called GISTs, blur into surrounding intestine on CT images, and doctors currently trace them by hand, a slow and subjective task. The authors' two-stage model first fuses CT and PET scans into a single richer image. It then segments the tumor using a vision transformer architecture designed to capture both fine detail and broad context.
Across a battery of standard scores for both fusion quality and segmentation accuracy, the combined system beat existing models. The authors argue that better fused images make the transformer's job easier, and that the pairing could give clinicians faster, more consistent tumor maps for diagnosis and treatment planning. The work is a methods demonstration on scan data rather than a clinical trial.
Catalog record
| CREATOR | Venu Allapakam, Yepuganti Karuna |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56042 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |