Named Models
MedSAMBo Wang Lab, TorontoReleased2024Runs onnot in the notesClassvisionMedical image segmentation model adapted from SAM, published in Nature Communications 15:654 (2024). Developer location (Toronto) unverified.Commercial use: allowed; Jakarta hosting: unknown; Indonesian support: not statedMedSAM2Bo Wang Lab, TorontoReleasedApr 2025Runs onnot in the notesClassvisionSuccessor to MedSAM for 3D and video segmentation, released 7 Apr 2025.Commercial use: allowed; Jakarta hosting: unknown; Indonesian support: not stated
nnU-Net v2DKFZ (German Cancer Research Center), Helmholtz ImagingReleasednot in the notesRuns onnot in the notesClassvision
Medical image segmentation framework from DKFZ and Helmholtz Imaging, Apache-2.0.Commercial use: allowed; Jakarta hosting: unknown; Indonesian support: not stated
MedImageParse (BiomedParse)Microsoft AIReleasedNov 2024Runs onnot in the notesClassvision
Microsoft biomedical model for joint segmentation, detection and recognition across nine imaging modalities from text prompts.Commercial use: non-commercial; Jakarta hosting: unknown; Indonesian support: not stated
MONAI model bundlesNVIDIAReleasednot in the notesRuns onnot in the notesClassvision
MONAI consortium medical imaging segmentation bundles built and published by NVIDIA in the NGC catalog.Commercial use: not stated; Jakarta hosting: unknown; Indonesian support: not statedVISTA-2D and VISTA-3DNVIDIAReleasednot in the notesRuns onnot in the notesClassvision
NVIDIA foundation models for cell segmentation (VISTA-2D) and interactive 3D CT segmentation (VISTA-3D).Commercial use: not stated; Jakarta hosting: unknown; Indonesian support: not stated
U-NetUniversity of FreiburgReleasedMay 2015Runs onmobile or NPU, CPU serverClassvision
Encoder-decoder segmentation network with a contracting path and a symmetric expanding path, first shown on biomedical images.Commercial use: not stated; Jakarta hosting: unknown; Indonesian support: not stated
UNet++Arizona State UniversityReleasedJul 2018Runs onnot in the notesClassvision
Nested U-Net with dense skip pathways for medical image segmentation.Commercial use: not stated; Jakarta hosting: unknown; Indonesian support: not statedWhere this comes from
Models the research names for this use case. Research, with sources on each model record.
Who Runs It
No deployment is recorded for this use case.
Where this comes from
Documented uses by named organisations. The evidence column says who made the claim.
Candidate Models
COCO val
A general benchmark for this model class, not a test on this use case. Object detection and segmentation AP (mAP50-95). Higher is better. Showing 10 of 11.
RF-DETR 2XLRoboflow vendor-reported60.1
YOLO26xUltralytics vendor-reported57.5YOLO26lUltralytics vendor-reported55
D-FINEUniversity of Science and Technology of China (USTC), Institute of Artificial Intelligence, Hefei Comprehensive National Science Center independent55RF-DETR MRoboflow vendor-reported54.7YOLO26mUltralytics vendor-reported53.1RF-DETR SRoboflow vendor-reported52.9YOLO26sUltralytics vendor-reported48.6YOLO11Ultralytics independent48.3RF-DETR NRoboflow vendor-reported48050100
YOLO26lUltralytics, visionReleasedJan 2026open weightsruns on edgehas scoreslast 18 monthsYOLO26mUltralytics, visionReleasedJan 2026open weightsruns on edgehas scoreslast 18 monthsYOLO26nUltralytics, visionReleasedJan 2026open weightsruns on edgehas scoreslast 18 monthsYOLO26sUltralytics, visionReleasedJan 2026open weightsruns on edgehas scoreslast 18 monthsYOLO26xUltralytics, visionReleasedJan 2026open weightshas scoreslast 18 monthsRF-DETR 2XLRoboflow, visionReleasedNov 2025runs on edgehas scoreslast 18 monthsRF-DETR MRoboflow, visionReleasedNov 2025runs on edgehas scoreslast 18 monthsRF-DETR NRoboflow, visionReleasedNov 2025runs on edgehas scoreslast 18 monthsRF-DETR SRoboflow, visionReleasedNov 2025runs on edgehas scoreslast 18 months
SAM 3Meta, visionReleasedNov 2025open weightshas scoreslast 18 monthsShowing 10 of 653. All models of this class
Where this comes from
Derived, not researched: models in the directory whose class matches a class this use case uses, excluding family records and models shut down or deprecated. Ranked by how many fit facts they meet, then by release date. Each fit fact is read from the model record.
Links
ModelsnnU-Net v2DKFZ (German Cancer Research Center), visionMedSAMBo Wang Lab, Toronto, visionMedSAM2Bo Wang Lab, Toronto, visionMedImageParse (BiomedParse)Microsoft AI, vision+4 more
nnU-Net v2DKFZ (German Cancer Research Center), visionMedSAMBo Wang Lab, Toronto, visionMedSAM2Bo Wang Lab, Toronto, visionMedImageParse (BiomedParse)Microsoft AI, vision+4 moreWho Runs Itnone recorded
Models 8
U-NetUniversity of Freiburg, visionMedSAMBo Wang Lab, Toronto, visionMedSAM2Bo Wang Lab, Toronto, visionUNet++Arizona State University, visionnnU-Net v2DKFZ (German Cancer Research Center), Helmholtz Imaging, visionVISTA-2D and VISTA-3DNVIDIA, visionMedImageParse (BiomedParse)Microsoft AI, visionMONAI model bundlesNVIDIA, visionSources
- GitHub - MIC-DKFZ/nnUNet github.com https://github.com/MIC-DKFZ/nnUNet
- GitHub - bowang-lab/MedSAM: Segment Anything in Medical Images github.com https://github.com/bowang-lab/MedSAM
- GitHub - Project-MONAI/MONAI: AI Toolkit for Healthcare Imaging github.com https://github.com/Project-MONAI/MONAI
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Medical Image Segmentation", https://aurion.semai.id/usecases/hc-medical-image-segmentation/, records of October 2026.