What happened
Varian Medical Systems built Ethos therapy to adapt radiotherapy plans to each day's anatomy. The Malaysian Health Technology Assessment Section (MaHTAS) of the Ministry of Health reviewed it in 2024. MaHTAS writes that convolutional neural networks outline anatomy on the daily cone-beam CT scan and that an optimization engine generates an adapted plan. Its search, last run on 15 July 2024, found 10 observational studies of 244 patients in six countries. Most studies compared the adapted plans with the original scheduled plans. MaHTAS concluded from moderate-quality evidence that adaptive radiotherapy improves short-term target coverage and organ sparing. In one Dutch study of 11 prostate cancer patients, target coverage was 99.4% with the adapted plans and 86.7% with the scheduled plans. No study reported survival, tumor control or long-term toxicity, and safety evidence was limited. MaHTAS found that adaptation adds 13 to 34 minutes to each session and found no evidence on cost-effectiveness. It warned that AI contouring may be biased when training data do not represent local patients. The review reports no use of Ethos in Malaysia.
Sources
-
Assessment report, by authors with no tie to the system.
There was moderate quality evidence showing that online adaptive radiotherapy (oART) may improve target coverage while reducing normal tissue exposure, palpable by superior dosimetric benefits in many observational studies.
There is moderate quality of evidence from observational studies showing that online adaptive radiotherapy (oART) improves short-term dosimetric outcomes, such as target coverage and organ sparing, in patients with head and neck, lung, bone metastases, and pelvic cancer.
Details
AI system
- Workflow
-
Convolutional neural networks outlined the main anatomical structures on each day's cone-beam CT scan. An optimization engine generated an adapted plan, and clinicians chose between it and the scheduled plan while the patient stayed on the couch. Clinicians reviewed and edited the AI contours before treatment.
Follow-up
- Conditions for replication
-
A radiation oncologist, physicist, dosimetrist and therapist are available during each session. Staff are trained to edit and validate AI contours. Patients are selected with care, since the gain may be larger in anatomically complex sites such as the pelvis and bladder. The AI contouring is trained on data that represent the local population.
Entry details
- Sector
-
Healthcare.
- Purpose
-
Varian Medical Systems built Ethos therapy to adapt radiotherapy plans to each day's anatomy, using AI contouring and automated planning while the patient lies on the treatment couch.
31 more questions not yet checked
model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.
Cite this record
Peter H. Vartanian. (2026). Ethos daily replanning improves tumor coverage in Malaysia's review (AIBD-1095). Artificial Intelligence Benefits Database, edition of 1 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1095.html (archived at https://doi.org/10.7910/DVN/AADSII, version 7.0)