Loading dataset details…
harbor run -d viratzzs/petct-pilotharbor run -d viratzzs/petct-pilotPrivate two-task PET/CT benchmark pilot; separately scored lesion mapping and result repair tasks.
harbor run -d viratzzs/petct-pilotPrivate two-task PET/CT benchmark pilot; separately scored lesion mapping and result repair tasks.
harbor run -d viratzzs/petct-pilotThis private Harbor dataset contains two independent, separately scored tasks:
fdg-lesion-mapping-001: map FDG-avid lesions from PET/CT images.fdg-result-repair-002: adjudicate and repair a whole-study FDG result.Each task directory includes its own instructions, scan data, verifier, and reference files. The saved pilot rollouts are not part of this dataset. Clinical image-only fairness and difficulty have not yet been established; see each task's README for provenance and limits.
This private Harbor dataset contains two independent, separately scored tasks:
fdg-lesion-mapping-001: map FDG-avid lesions from PET/CT images.fdg-result-repair-002: adjudicate and repair a whole-study FDG result.Each task directory includes its own instructions, scan data, verifier, and reference files. The saved pilot rollouts are not part of this dataset. Clinical image-only fairness and difficulty have not yet been established; see each task's README for provenance and limits.