{"id": "package:4738b0af-2090-4045-9055-3d3cfe2534b4", "name": "Prediction3D.tif", "self_uri": "https://services.scicrunch.io/sparc/drs/v1/objects/4738b0af-2090-4045-9055-3d3cfe2534b4", "size": 571348, "created_time": "2025-12-12T04:17:23,990660Z", "updated_time": "2025-12-12T04:20:07,728774Z", "version": "1", "mime_type": "image/tiff", "checksums": [{"checksum": "2f344f4c7b50a65e22700f44baa7fb1e", "type": "sha256"}], "access_methods": [{"type": "s3", "access_url": {"url": "s3://sparc-prod-aod-discover-publish50-use1/521/files/derivative/sub-SR009/sam-SR009-14/Prediction3D.tif"}, "region": "us-east-1"}], "dataset": {"id": "521", "doi": "DOI:10.26275/wfud-am1l", "title": "Automated 3D segmentation of human vagus nerve fascicles and epineurium from micro-computed tomography images using anatomy-aware neural networks", "description": "Data associated with our publication on automated 3D segmentation of microCT images of human vagus nerves.", "abstract": "Microcomputed tomography (microCT) of embalmed cadaveric human vagus nerves stained with phosphotungstic acid. Fascicle and nerve boundaries were manually segmented in a subset of images to train 2D and 3D U-nets. The remaining images were automatically segmented with the trained U-nets. We quantified the performance of each U-net using established image analysis metrics and anatomically relevant metrics."}}