bia_mifa_models.yaml

---
id: https://w3id.org/BioImage-Archive/bia-mifa-models
name: bia-mifa-models
title: bia-mifa-models
description: |-
  Schemas and models for working with the BioImage Archive's MIFA (Metadata, Incentives, Formats and Accessibility) metadata implementation
license: CC0
see_also:
  - https://BioImage-Archive.github.io/bia-mifa-models

prefixes:
  bia_mifa_models: https://w3id.org/BioImage-Archive/bia-mifa-models/
  linkml: https://w3id.org/linkml/
  schema: http://schema.org/
  ORCID: https://orcid.org/
  ROR: https://ror.org/
  DOI: https://doi.org/
  PMID: https://pubmed.ncbi.nlm.nih.gov/
  WIKIDATA_PROPERTY: https://www.wikidata.org/prop/
  pav: http://purl.org/pav/
  example: https://example.org/
default_prefix: bia_mifa_models
default_range: string

imports:
  - linkml:types

classes:

  Study:
    description: >-
      General information about the study
    mixins:
      - PublicationsCollection
      - AuthorCollection
      - Links
      - GrantReferenceCollection
    slots:
      - title
      - description
      - keywords
      - license
      - ai_models_trained
      - acknowledgements
      - funding_statement

  Publications:
    description: >-
      Information about any publications associated with the dataset
    slots:
      - publication_title
      - publication_authors
      - publication_doi
      - publication_year
      - pubmed_id
    id_prefixes:
      - DOI
  
  PublicationsCollection:
    mixin: true
    description: >-
      A holder for Publications objects
    attributes:
      publications:
        description: a collection of publications associated with a study
        range: Publications
        multivalued: true
        inlined: true
  
  Author:
    description: >-
      Information about the authors
    mixins: 
      - OrganisationInfoCollection
    slots:
      - author_first_name
      - author_last_name
      - email
      - orcid_id
      - role

  AuthorCollection:
    mixin: true
    description: >-
      A holder for Author objects
    attributes:
      authors:
        description: a collection of the authors of a study
        range: Author
        multivalued: true
        inlined_as_list: true
  
  OrganisationInfo: 
    description: Information about the organisation the author is affiliated with
    slots:
      - organisation_name
      - address
      - ror_id

  OrganisationInfoCollection:
    mixin: true
    description: >-
      A holder for OrganisationInfo objects
    attributes:
      organisation:
        description: a collection of the name and address of organisations authors are affiliated with
        range: OrganisationInfo
        multivalued: true
        inlined_as_list: true

  GrantReference:
    description: Information about grant ID and funding body that funded the study
    slots:
      - grant_id
      - funder

  GrantReferenceCollection:
    mixin: true
    description: >-
      A holder for GrantReference objects
    attributes:
      grants:
        description: a collection of grant ids and funder names associated with the study
        range: GrantReference
        multivalued: true
        inlined_as_list: true

  Links:
    description: Links related to the study
    mixin: true
    slots:
      - link_url
      - link_description
  
  FileLevelMetadata:
    description: metadata atributes that must be detailed at the file level
    slots:
      - annotation_id
      - annotation_type
      - source_image_id
      - transformations
      - spatial_information
      - annotation_creation_time

  FileLevelMetadataCollection:
    mixin: true
    description: >-
      A holder for FileLevelMetadata objects
    attributes:
      file_metadata:
        description: a collection of the file level metadata for each annotation
        range: FileLevelMetadata
        multivalued: true
        inlined_as_list: true

  Annotations:
    tree_root: true
    description: >-
      A set of annotations for an AI-ready dataset
    mixins:
      - AuthorCollection
      - FileLevelMetadataCollection
    slots:
      - annotation_overview
      - annotation_type
      - annotation_method
      - annotation_criteria
      - annotation_coverage
      - annotation_confidence_level
      
  Version:
    description: >-
      Information about the dataset version
    slots:
      - version
      - timestamp
      - changes
      - previous_version

slots:

  title:
    slot_uri: schema:title
    description: The title for your dataset. This will be displayed when search results including your data are shown. Often this will be the same as an associated publication.
    required: true
    examples:
      - value: An annotated fluorescence image dataset for training nuclear segmentation methods
      - value: Embryonic mice ultrasound volumes with body and brain volume segmentation masks
  
  description:
    slot_uri: schema:description
    description: Brief description of the detaset. Must contain the biological application of the dataset 
    required: true
    examples:
      - value: >-
          Accurate quantification of nuclear features is critical for the characterization of tumours. 
          This dataset contains annotated fluorescent images of cell nuclei from a normal human skin cell line, and from different tumour samples from cancer patients. 
          This dataset can be used to train nuclear image segmentation algorithms.
      - value: >-
          The number of cells in an embryo at early stages of development can be used to understand its viability, however counting cells on DIC images is a challenging task. 
          This dataset contains annotated DIC images of mouse embryos to train machine learning models.
  
  keywords:
    slot_uri: schema:keywords
    description: Keywords or tags used to describe the dataset
    required: true
    multivalued: true
    examples:
      - value:
        - segmentation
        - nuclei
        - cancer
      - value:
        - image segmentation 
        - high-frequency ultrasound 
        - mouse embryo 
        - volumetric deep learning
  
  license:
    slot_uri: schema:license
    description: The license under which the data are available
    required: true
    range: LicenseType
    examples:
      - value: CC0
      - value: CC_BY
 
  ai_models_trained:
    description: Link to the models that have been trained with this dataset (if any). This could be links to GitHub or repositories like the BioImage Zoo.
    range: uriorcurie
    multivalued: true
    examples:
      - value: https://github.com/BIIFSweden/MariaUlvmar2020-1
      - value: 
        - https://github.com/perlfloccri/NuclearSegmentationPipeline
        - https://bioimage.io/#/?id=10.5281%2Fzenodo.6028097
  
  funding_statement:
    description: Description of how the study was funded
    required: true
    examples:
      - value: >-
          T.G. was supported by an EMBO Long-Term Fellowship (ALTF 1234-5678), and the European Union Horizon 2020 research and innovation programme 
          under the Marie Sklodowska-Curie grant agreement No 12345. This work was supported by a Cancer Research UK Programme Grant (DRCNPG-123\123456) to J.C.
      - value: >-
          This work was facilitated by an EraSME grant (project TisQuant) under the grant no. 844198 and by a COIN grant (project VISIOMICS) 
          under the grant no. 861750, both grants kindly provided by the Austrian Research Promotion Agency (FFG), and the St. Anna Kinderkrebsforschung. 
          Partial funding was further provided by BMK, BMDW, and the Province of Upper Austria in the frame of the COMET Programme managed by FFG.
  
  grant_id:
    description: The identifier for the grant
    identifier: true
    required: true
    slot_uri: schema:identifier
    examples: 
      - value: ALTF 1234-567
      - value: "840767"

  funder:
    slot_uri: schema:funder
    description: The funding body provididing support
    required: true
    examples:
      - value: EMBO Long-Term Fellowship
      - value: Marie Sklodowska-Curie Individual Fellowship
  
  publication_title:
    slot_uri: schema:title
    description: The title for a publication associated to the dataset
    required: true
    examples:
      - value: An annotated fluorescence image dataset for training nuclear segmentation methods
      - value: "Deep Mouse: An End-to-End Auto-Context Refinement Framework for Brain Ventricle & Body Segmentation in Embryonic Mice Ultrasound Volumes"

  publication_authors:
    slot_uri: schema:author
    description: Authors of associated publication
    required: true
    examples:
      - value: Truman Grayson, Josiah Carberry
      - value: J. Perez, F. Fulan, J. Doe, W. Wu and J. Novak 

  publication_doi:
    identifier: true
    slot_uri: WIKIDATA_PROPERTY:P356
    range: uriorcurie
    description: Digital Object Identifier (DOI)
    examples: 
      - value: DOI:10.1242/jcs.033340
      - value: DOI:10.1242/jcs.174946

  publication_year:
    description: Year of publication
    examples:
      - value: "2021"
      - value: "1986"
  
  pubmed_id:
    slot_uri: WIKIDATA_PROPERTY:P698
    range: uriorcurie
    description: Identifier for journal articles/abstracts in PubMed
    examples: 
      - value: PMID:18492790
      - value: PMID:26136366
  
  link_url:
    description: URL of relevant link
    range: uriorcurie
    multivalued: true
    required: true
    examples:
      - value:
        - https://github.com/BioImage-Archive
        - https://www.ebi.ac.uk/empiar/EMPIAR-10459/
  
  link_description:
    description: The description of the linked content
    multivalued: true
    examples:
      - value:
        - Image analysis code
        - FIB SEM dataset matching the light microscopy data 
  
  acknowledgements:
    description: Any people or groups that should be acknowledged as part of the dataset
    examples:
      - value: >-
          We thank the Institute microscopy facility for support with image acquisition. 
          Flybase provided important information for this study.
      - value: >-
          We are grateful to Y. Papadopoulos for providing reagents and access to her confocal microscope. 
          We thank J. Dupont and E. Mustermann for comments on the manuscript, and G. Hong for useful discussions. 
        
  orcid_id:
    slot_uri:  WIKIDATA_PROPERTY:P496
    range: uriorcurie
    description: A unique identifier for an author
    examples: 
      - value: ORCID:0000-0002-1825-0097
      - value: ORCID:0000-0002-9079-593X
  
  author_first_name:
    slot_uri: schema:name
    description: First name for an author
    required: true
    examples:
      - value: Josiah
      - value: Marie
  
  author_last_name:
    slot_uri: schema:name
    description: Last name for an author
    required: true
    examples:
      - value: Carberry
      - value: "Skłodowska–Curie"
  
  email:
    slot_uri: schema:email
    description: Email address of a person
    pattern: "^\\S+@[\\S+\\.]+\\S+"
    examples: 
      - value: jcarberry@example.com
      - value: mariesc@example.com
  
  organisation_name:
    slot_uri: schema:name
    description: Organisation name
    required: true
    examples:
      - value: EMBL-EBI
      - value: University of Toronto

  ror_id:
    range: uriorcurie
    slot_uri:  WIKIDATA_PROPERTY:P6782
    description: Identifier for the Research Organization Registry
    examples: 
      - value: ROR:02catss52
      - value: ROR:03dbr7087

  address:
    slot_uri: schema:name
    description: Organisation address
    examples:
      - value: Wellcome Genome Campus, Hinxton, Cambridgeshire, CB10 1SD, UK.
      - value: 27 King's College Cir, Toronto, ON M5S, Canada.

  role:
    slot_uri: schema:roleName
    description: Role of the author when creating the dataset
    multivalued: true
    examples: 
      - value: 
        - data annotation
        - data acquisition
        - conceptualization
      - value: data curator
  
  annotation_overview:
    description: Short descriptive summary indicating the type of annotation and how it was generated
    required: true
    examples: 
      - value: Segmentation masks of human cell nuclei curated by experts from a model prediction 
      - value: Manual quantification of the number of cells in mouse embryos performed by experts
  
  annotation_type:
    required: true
    range: AnnotationType
    multivalued: true
    description: Annotation type, for example class labels, segmentation masks, counts...
    examples: 
      - value: 
        - segmentation_mask 
        - class_labels
      - value: counts 
  
  annotation_method:
    required: true
    description: >-
      Description of how the annotations were created. For example, were the annotations crowdsourced or expertly annotated, 
      produced by  a human or software, what software was used, when were the annotations created, 
      protocols used for consensus and quality assurance
    examples: 
      - value: >-
          A machine learning-based framework was utilized to produce a first coarse annotation of nuclear contours. 
          Then, a disease expert curated all annotations by editing, adding, or removing polygons. 
          Finally, an expert pathologist revised all image annotations and a final version of the annotations was curated.
      - value: >-
          Cell counts were performed manually by expert biologists. For cases difficult to quantify, 
          a vote including all experts was taken and annotations were corrected according to the majority view.
  
  annotation_criteria:
    description: Rules used to generate annotations
    examples:
      - value: Only nuclei in focus were annotated, if parts of a nucleus were out of focus, only the part of the nucleus being in focus was annotated
      - value: >-
          During cell counting, cell division was tracked using anillin localization. Mitosis was considered completed when anillin stop localising 
          at the contractile ring and re-entered the cell nucleus, at which point the dividing cell was considered two different cells.
  
  annotation_coverage:
    description: Which images from the dataset were annotated, and what percentage of the data has been annotated from what is available
    examples:
      - value: All images were annotated
      - value: "Images: image_1, image_2 and image_3 were not annotated due to the abnormal cell death rates compared to other samples."
  
  source_image_id:
    required: true
    description: identifier of the source image from which the annotation originated
    examples:
      - value: IM1
      - value: "12532806"
  
  annotation_id:
    description: The identifier for the annotation
    identifier: true
    required: true
    slot_uri: schema:identifier
    examples: 
      - value: AN1
      - value: "12345678"
  
  annotation_confidence_level:
    description: Confidence on annotation accuracy
    examples: 
      - value: Curators had 15 and 20 years of experience working in cancer pathology
      - value: >- 
          To characterize the quality of gold standard segmentation annotations, 
          we calculated the average and standard deviation performance of the three independent annotators 
          relative to the gold truth that was established by merging the triplets of manual annotations. 
          For the whole dataset this metric was 0.904±0.081
  
  spatial_information:
    description: Spatial information for non-pixel annotations
    examples:
      - value: The pixels belonging to the mouse embryo related to this cell count are provided in the binary mask with annotation id AN1237946
      - value: Each annotated object contains a category_id indicating the annotation class label, and an associated bounding box delineating the corresponding blod vessel
  
  transformations:
    description: Any transformations required to link the images to the annotations
    examples:
      - value: Rotation_angle = 37°
      - value: Translation_horizontal_pixel_displacement = 150, Translation_vertical_pixel_displacement = 92
  
  annotation_creation_time:
    description: Date and time when the annotation was created
    range: datetime
    slot_uri: pav:authoredOn
    examples:
      - value: "2021-04-11T11:00:15-05:00"
      - value: "2017-01-23T18:22:04-00:00"
  
  version:
    required: true
    range: string
    slot_uri: pav:version
    description: Unique version number
    examples:
      - value: v1.1
      - value: "3.0"
  
  timestamp:
    required: true
    range: datetime
    slot_uri: pav:authoredOn
    description: Date and time when the version was created
    examples:
      - value: "2023-12-10T12:20:00-05:00"
      - value: "2021-07-03T13:26:00-00:00"
  
  changes:
    range: string
    description: Textual description of changes compared to previous version
    examples:
      - value: The segmentation mask for image 10 (segmentation_10.tif) has been updated with the annotation of two nuclei that were missing in the previous version
      - value: New annotations for images 12047635 and 12063916 have been included in the dataset with annotation ids AN12047635 and AN12063916 respectively
  
  previous_version:
    range: uriorcurie
    slot_uri: pav:previousVersion
    description: Pointer to previous version
    examples:
      - value: example:version1.0
      - value: example:version2.9

enums:

  AnnotationType:
    permissible_values:

      class_labels:
        description: tags that identify specific features, patterns or classes in images
        todos:
          - map this to an ontology
      
      bounding_boxes:
        description: rectangles completely enclosing a structure of interest within an image
        todos:
          - map this to an ontology
      
      counts:
        description: number of objects, such as cells, found in an image
        todos:
          - map this to an ontology
      
      derived_annotations:
        description: additional analytical data extracted from the images. For example, the image point spread function,the signal to noise ratio, focus information…
        todos:
          - map this to an ontology
      
      geometrical_annotations:
        description: polygons and shapes that outline a region of interest in the image. These can be geometrical primitives, 2D polygons, 3D meshes…
        todos:
          - map this to an ontology
      
      graphs:
        description: graphical representations of the morphology, connectivity, or spatial arrangement of biological structures in an image. Graphs, such as skeletons or connectivity diagrams, typically consist of nodes and edges, where nodes represent individual elements or regions and edges represent the connections or interactions between them
        todos:
          - map this to an ontology
      
      point_annotations:
        description:  X, Y, and Z coordinates of a point of interest in an image (for example an object's centroid  or landmarks).
        todos:
          - map this to an ontology
      
      segmentation_mask:
        description: an image, the same size as the source image, with the value of each pixel representing some biological identity or background region
        todos:
          - map this to an ontology
      
      tracks:
        description: annotations marking the movement or trajectory of objects within a sequence of bioimages
        todos:
          - map this to an ontology
      
      weak_annotations:
        description: rough imprecise annotations that are fast to generate. These annotations are used, for example,  to detect an object without providing accurate boundaries
        todos:
          - map this to an ontology
      
      other:
        description: other types of annotations, please specify in the annotation overview section
        todos:
          - map this to an ontology
  
  LicenseType:
    permissible_values:

      CC0:
        description: No Copyright. You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission.
        meaning: https://creativecommons.org/publicdomain/zero/1.0/
      
      CC_BY:
        description: >-
         You are free to: Share — copy and redistribute the material in any medium or format. Adapt — remix, transform, and build upon the material 
         for any purpose, even commercially. You must give appropriate credit, provide a link to the license, and indicate if changes were made. 
         You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
        meaning: https://creativecommons.org/licenses/by/4.0/