Geometry-encoded deep learning (GeoDL) framework for real-time 3D dose verification for online adaptive radiotherapy.
Source check: Source checked. Not independent scientific confirmation.
Finding: A geometry-encoded 3D U-Net estimates dose from integrated fluence maps and CT volumes; 281 prostate cases were used for training/validation and 24 separate cases for testing against treatment-planning dose. Reported agreement and speed support a retrospective dosimetric feasibility result; use remains limited to prostate VMAT/single LINAC configuration, with clinical implementation still future work.
Technology: A geometry-encoded CT/fluence-map input is processed by 3D U-Net to estimate full-volume dose.
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- A geometry-encoded CT/fluence-map input is processed by 3D U-Net to estimate full-volume dose.
- Test and result
- 281 prostate cases used for training/validation and 24 separate test cases compared against TPS calculations; single site/VMAT/LINAC limits remain.
- Limitation / still unknown
- External replication and real-world or clinical benefit were not established by the reviewed study; author affiliation is not stated in the reviewed material.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC12753030/ · https://doi.org/10.1088/3049-477x/adcb4b
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.