{"EMPIAR-11037":{"imagesets":[{"segmentations":[],"name":"Images and masks correpsonding to grayscale 2D EM image patches and paired mitochondrial label maps","directory":"data","category":"micrographs - single frame","header_format":"TIFF","data_format":"TIFF","num_images_or_tilt_series":43720,"frames_per_image":1,"frame_range_min":null,"frame_range_max":null,"voxel_type":"UNSIGNED BYTE","pixel_width":null,"pixel_height":null,"micrographs_file_pattern":"","picked_particles_file_pattern":"","picked_particles_directory":"","details":"PLEASE READ!\nThere are 21,860 annotated images in CEM-MitoLab. They contain 135,285 mitochondrial instances in total. The images are a subset of CEM1.5M (also on EMPIAR; EMPIAR-11035). The image patches are mostly 224 x 224 pixels, however some are 512 x 512, and some are smaller. The directory names are either randomized for in-house or unpublished data, or kept as-is for published data. \nEach directory has \"image\" and \"mask\" subdirectory. These directories are populated by grayscale images and \"paired\" mitochondrial instance label map files, respectively. Both are unsigned byte tiff files. For the mito labels, background is 0, individual instances per image are 1,2,3...255. \n2D or 3D = dimensionality of the original EM data \nLOC-0,1,2 = is the index along z, y, x axis (ie. the numbers following are the location in the original image). Extents of 5, m, n, when present, is because images were present as a flipbook of 5. Only the middle one is annotated and uploaded here with the paired grayscale images. For practical purposes these can be ignored. \nPLEASE SEE .XLS FILE <</metadata/cem_mitolab_metadata.xlsx>> FOR METADATA USING REMBI PRINCIPLES (Sarkans et al Nature Methods 2021)","image_width":"224","image_height":"224"}],"workflow_file":null,"grant_references":[],"version_history":[],"title":"CEM-MitoLab: a dataset of ~22K cellular EM 2D images with label maps of ~135K mitochondrial instances, for deep learning","principal_investigator":[{"author_orcid":"0000-0001-7982-6494","middle_name":null,"organization":"NCI/FNLCR","street":null,"town_or_city":"Frederick","state_or_province":"MD","post_or_zip":"21701","telephone":null,"fax":null,"first_name":"Kedar","last_name":"Narayan","email":"narayank [at] mail.nih.gov","country":"United States","entry":"EMPIAR-11037"}],"status":"REL","deposition_date":"2022-04-28","release_date":"2022-05-10","obsolete_date":null,"update_date":"2022-05-10","corresponding_author":{"author":{"author_orcid":"0000-0001-7982-6494","middle_name":null,"organization":"NCI/FNLCR","street":null,"town_or_city":"Frederick","state_or_province":"MD","post_or_zip":"21701","first_name":"Kedar","last_name":"Narayan","country":"United States"}},"authors":[{"author":{"name":"Narayan K","author_orcid":"0000-0001-7982-6494"}},{"author":{"name":"Conrad RW","author_orcid":null}}],"cross_references":[],"biostudies_references":[],"idr_references":[],"empiar_references":[{"name":"EMPIAR-11035"}],"citation":[{"authors":[{"name":"Narayan K","author_orcid":"0000-0001-7982-6494"}],"editors":[],"published":false,"j_or_nj_citation":true,"title":"Instance segmentation of mitochondria in electron microscopy images with a generalist deep learning model","volume":null,"country":"","first_page":null,"last_page":null,"year":null,"language":null,"doi":null,"pubmedid":null,"details":null,"book_chapter_title":null,"publisher":null,"publication_location":null,"journal":"","journal_abbreviation":"","issue":null,"preprint":false}],"dataset_size":"2.8 GB","experiment_type":"FIB-SEM","scale":null,"entry_doi":"10.6019/EMPIAR-11037"}}