Full Format Reference

Items List and `tasks.json`

Most RedBrick flows incorporate two key JSON files:

Items List - a file which points RedBrick AI to visual assets within a third-party storage solution;
tasks.json - a file generated upon export that contains a record of the annotation work completed within a Project. Upon export, the tasks.json file will contain a single entry for each Task;

If you'd like to upload annotations along with your data using either the CLI or the SDK, please see the corresponding documentation.

Object Reference

Please see the definition (in TypeScript) of RedBrick's various objects below:

type Tasks = Task[];

// Single task on RedBrick can be single/multi-series
type Task = {
  // Required on upload and export
  name: string;
  series: Series[];

  // Task level annotation information
  classification?: Classification;

  // Not required on upload, present in tasks.json
  taskId?: string;
  currentStageName?: string;
  createdBy?: string;
  createdAt?: string;
  updatedBy?: string;
  updatedAt?: string;
  
  // assign metadata to a Task on upload, present in tasks.json
  metaData?: { [key: string]: string }
  
  // assign priorities to a Task on upload
  priority?: number;
  
  // Prescribe task assignent upon upload
  preAssign?: {
    [stageName: string]: string
  }
};

// How to correctly format a Series object for upload
// A single series can be 2D, 3D, video etc.
// Also present in tasks.json
type Series = {
  items: string | string[];
  name?: string;

  segmentations?: string | string[];
  segmentMap?: {
    [instanceId: string]: number | string | string[] | {
      category: number | string | string[];
      attributes?: Attributes;
      mask?: string;
    };
  };
  binaryMask?: boolean;
  semanticMask?: boolean;
  pngMask?: boolean;
  
  landmarks?: Landmark[];
  landmarks3d?: Landmark3D[];
  measurements?: (MeasureLength | MeasureAngle)[];
  boundingBoxes?: BoundingBox[];
  cuboids?: Cuboid[];
  ellipses?: Ellipse[];
  polygons?: Polygon[];
  polylines?: Polyline[];
  
  classifications?: Classification[];
  instanceClassifications?: InstanceClassification[];
   metaData?: { [ key: string ]: string };
};

// Label Types

type Landmark = {
  point: Point2D;
  category: number | string | string[];
  attributes?: Attributes;

  // video meta-data
  video?: VideoMetaData;
};

type Landmark3D = {
  point: VoxelPoint;
  category: number | string | string[];
  attributes?: Attributes;
};

type MeasureLength = {
  type: 'length';
  point1: VoxelPoint;
  point2: VoxelPoint;
  absolutePoint1: WorldPoint;
  absolutePoint2: WorldPoint;
  normal: [number, number, number];
  length: number;
  category: number | string | string[];
  attributes?: Attributes;
};

type MeasureAngle = {
  type: 'angle';
  point1: VoxelPoint;
  point2: VoxelPoint;
  vertex: VoxelPoint;
  absolutePoint1: WorldPoint;
  absolutePoint2: WorldPoint;
  absoluteVertex: WorldPoint;
  normal: [number, number, number];
  angle: number;
  category: number | string | string[];
  attributes?: Attributes;
};

type BoundingBox = {
  pointTopLeft: Point2D;
  wNorm: number;
  hNorm: number;
  category: number | string | string[];
  attributes?: Attributes;
  stats?: MeasurementStats;

  // video meta-data
  video?: VideoMetaData;
};

type Cuboid = {
  point1: VoxelPoint;
  point2: VoxelPoint;
  absolutePoint1: WorldPoint;
  absolutePoint2: WorldPoint;
  category: number | string | string[];
  attributes?: Attributes;
  stats?: MeasurementStats;
}

type Ellipse = {
  pointCenter: Point2D;
  xRadiusNorm: number;
  yRadiusNorm: number;
  rotationRad: number;
  category: number | string | string[];
  attributes?: Attributes;
  stats?: MeasurementStats;
  
  // video meta-data
  video?: VideoMetaData;
}

type Polygon = {
  points: Point2D[];

  category: number | string | string[];
  attributes?: Attributes;
  stats?: MeasurementStats;

  // video meta-data
  video?: VideoMetaData;
};

type Polyline = {
  points: Point2D[];
  category: number | string | string[];
  attributes?: Attributes;

  // video meta-data
  video?: VideoMetaData;
};

type Classification = {
  category: number | string | string[];

  // video meta-data
  video?: VideoMetaData;
};

type InstanceClassification = {
  fileIndex: number;
  fileName?: string;
  values: { [attributeName: string] : boolean};
}

type Attributes =
  | { // Taxonomy V2 attributes
      attrId?: number;
      name?: string;
      optionId?: number | number[];
      value?: string | boolean | string[];
    }[]
  | { // Taxonomy V1 attributes
      [attributeName: string]: string | boolean | string[];
    };

type VideoMetaData = {
  frameIndex: number;
  trackId?: string;
  keyFrame?: number;
  endTrack?: Boolean;
};

// i is rows, j is columns, k is slice
type VoxelPoint = {
  i: number;
  j: number;
  k: number;
};
// The position of VoxelPoint in physical space (world coordinate) computed using the Image Plane Module
type WorldPoint = {
  x: number;
  y: number;
  z: number;
};
type Point2D = {
  xNorm: number;
  yNorm: number;
};

type MeasurementStats = {
  average: number;
  area?: number;
  volume?: number;
  minimum: number;
  maximum: number;
};

Object Glossary

Task

The Task object represents a single task on RedBrick AI. It contains task-level meta-data information about all the series within the task. A task can contain a single series or multiple series (ex. a full MRI study).

`name: string`

A user-defined string is defined on upload, it is required to be unique across all tasks in your project. The name is meant to be a human-readable string that can help identify tasks ex. you can set name of a task to patient/study01.

`taskId?: string`

A unique identifier generated for each Task by RedBrick AI. This value is provided on export.

`currentStageName?: string`

The Stage a Task is in is when it is exported.

`createdBy?: string`

The email address of the user who uploaded the Task.

`createdAt?: string`

The datetime this Task was created (i.e. uploaded).

`updatedBy?: string`

The email of the last user to make edits to this Task.

`updatedAt?: string`

The datetime that this Task was last edited.

`preAssign?: { [ stageName : string ] : string }`

When uploading a Task, prescribe which users will have the Task assigned to them by Stage.

You can define the assignment for each Stage of the workflow, for example, Label and Review, {"Label": "name1@redbrickai.com", "Review": "name2@redbrickai.com"}.

`classification: { attributes : [ string : boolean ] }`

A list of attributes assigned to an entire Task (or study, if the Task encapsulates an entire study).

metaData?: { [ key: string ]: string }

A list of key value pairs that can be affixed to a Task. This information is visible in the Annotation Tool.

priority?: number

Assign a priority value to a specific Task, which will influence the order in which the Task displays on the Data Page. The Automatic Assignment protocol will also auto-assign Tasks with a priority value to a user’s Labeling / Review Queue before moving on to Tasks without priority values.

Series

The Series object has metadata and annotations for a single Series within a Task. A Series can represent anything from a single MRI/CT series, a video, or a single 2D image.

If a Series contains annotations, you can expect one or more of the label entries to be present (e.g. segmentations, polygons etc.).

`items: string | string[]`

The items entry is a list of file paths that point to your data. Please have a look at the documentation for a fuller explanation of how to format for various modalities and series/study uploads.

`name: string`

An optional user-defined string, needs to be unique across all series. Individual series will be named after this value on the labeling tool. Exported segmentation files will also be named using this value. Using the Series Instance UID here is good practice.

`classifications: { attributes: [ string : boolean ] }`

A list of attributes assigned to a specific Series.

`instanceClassifications: fileIndex | fileName | [ values: { string : boolean } ]`

The instanceClassifications object defines a series of boolean values that can be assigned to individual instances (e.g. frames in a video).

metaData?: { [ key: string ]: string }

A list of key value pairs that can be affixed to a Series. This information is visible in the Annotation Tool.

binaryMask?: boolean

Reflects the user’s choice of optionally exporting annotations as a binary mask.

semanticMask?: boolean

Reflects the user’s choice of optionally using semantic export.

pngMask?: boolean

Reflects the user’s choice of optionally exporting annotations as a PNG mask.

Common Label Keys

Here are the definition for some common entries present in some/all label entries.

`category: string | string[]`

The class of your annotations. This value is part of your Project Taxonomy. If the class is nested, category will be string[].

`attributes: { [ attributeName: string ]: string | boolean }`

Each annotation can have accompanying attributes, that are also defined in your Project Taxonomy. attributeName is defined when creating your Taxonomy.

`voxelPoint: { i: number, j: number, k: number }`

VoxelPoint represents a three-dimensional point in image space, where i and j are columns and rows, and k is the slice number.

`worldPoint: { x: number, y: number, j: number }`

WorldPoint represents a three-dimensional point in physical space/world coordinates. The world coordinates are calculated using VoxelPoint and the Image Plane Module.

`point2D: { xnorm: number, ynorm: number }`

Point2D represents a two dimensional point. This is used to define annotation types on 2D data. xnorm has been normalized by image width, hnorm has been normalized by image height.

`fileIndex: number`

fileIndex is an integer that corresponds to a specific frame in a video series.

`fileName: string`

fileName represents the name given to an image or specific frame in a video series.

`measurementStats: Dict`

A dictionary containing a variety of geometric information about certain Object Labels.

Measurement Stats

A dictionary (measurementStats) containing geometric information about certain Object Labels.

average: number

The average pixel intensity value inside of a structure.

area?: number

The area of a 2D Object Label (e.g. Bounding Box, Ellipse), measured in square millimeters.

volume?: number

The volume of a 3D structure (e.g. Cuboid), measured in cubic millimeters.

minimum: number

The lowest pixel intensity value present in the structure.

maximum: number

The highest pixel intensity value present in the structure.

Video Meta Data

`videoMetaData: Dict`

A dictionary containing frameIndex, trackId, keyFrame, and endFrame.

`frameIndex: number` (video)

This specifies which frame in a video sequence the annotation was created.

`trackId: string` (video)

A unique string that identifies distinct object tracks in a video sequence.

`keyFrame: boolean` (video)

If true, this annotation was manually added on a particular video sequence. If false, this annotation is a result of linear interpolation.

`endFrame: boolean` (video)

If true, the annotation is the last annotation for a particular video track segment.

Segmentations and `segmentMap`

`segmentations?: string | string[]`

A list of file paths of segmentation files for this series. Either a single .nii file, or multiple .nii files containing different instances.

segmentMap?: { [ instanceId: number ]: { category: string | string[]; attributes?: Attributes } };

A mapping between a segmentation's instance ID, your Taxonomy category name, and any accompanying attributes. The mapping will apply only to the current series, and instance IDs must be unique across all series in a task (this is useful for instance segmentation).

Please note that the segmentMap's instanceId is generated incrementally based on the order in which annotations were created by the labeler. You can find an example JSON output below.

"items": ["image-file.ima",],
"segmentations": "./path/to/segmentation/file.nii.gz",
"segmentMap": {
  "1": {
    // This is the first annotation the labeler created
    "category": "Vertebral Body"
  },
  "2": {
    // This is the second annotation the labeler created
    "category": "Vertebral Body"
  },
  "3": {
    // This is the third annotation the labeler created
    "category": "Vertebral Body"
  },
  "4": {
    // This is the fourth annotation the labeler created
    "category": "Spinal Canal Mass"
  },
  "5": {
    // This is the fifth annotation the labeler created
    "category": "Disc Pathology"
  }
},

`mask?: string`

The path for the annotation file associated with a specific instanceId.

BoundingBox

Contains information about the Bounding Box Object Label.

`pointTopLeft:` `Point2D`

The location of the top-left point of the bounding box.

`wNorm, hNorm: number`

The width and height of the bounding box, normalized by the width and height of the image.

Polygon

Contains information about the Polygon Object Label.

`points:` `Point2D[]`

A list of 2D points that are connected to form a polygon. This list is ordered such that, $point_i$ is connected to $point_{i+1}$ . The last point is also connected to the first point to close the polygon.

MeasureLength

Contains information about the Length Measurement Object Label.

`point1, point2 :` `VoxelPoint`

A length measurement is defined by two points, and the length measurement is the distance between the two points.

`absolutePoint1, absolutePoint2 :` `WorldPoint`

Corresponding to point1, point2 these are points in physical space.

`normal: [number, number, number]`

Measurements can be made on oblique planes. normal defines the normal unit vector to the slice on which this annotation was made. For annotations made on non-oblique planes, the normal will be [0,0,1].

`length: number`

The value of the measurement in mm.

MeasureAngle

Contains information about the Angle Object Label.

`point1, point2, vertex :` `VoxelPoint`

Angle measurement is defined by three points, where the vertex is the middle point. The angle between the two vectors (vertex -> point1 and vertex -> point2) defines the angle measurement. These points are all represented in IJK image coordinate space.

`absolutePoint1, absolutePoint2 :` `WorldPoint`

Corresponding to point1, point2, vertex, these values are coordinates in the DICOM world coordinate system i.e. physical space.

`normal: [number, number, number]`

`angle: number`

The value of the angle in degrees.

Ellipse

Contains information about the Ellipse Object Label.

pointCenter: point2D

Information regarding the exact center of the Ellipse Object Label.

xRadiusNorm: number

A numeric value equivalent to half the length of the Ellipse Object Label’s major axis.

yRadiusNorm: number

A numeric value equivalent to half the length of the Ellipse Object Label’s minor axis.

rotationRad: number

The rotation angle of the Ellipse Object Label, expressed in radians.

Landmark

Contains information about the Landmark Object Label on 2D images.

point: point2D

The point in physical space on a 2D image where the Landmark is located.

Landmark 3D

Contains information about the Landmark Object Label on 3D volumes.

point: voxelPoint

The point in physical space on a 3D volume where the Landmark is located.

Cuboid

Contains information about the Cuboid Object Label.

point1, point2: voxelPoint

Information about the initial point of the Cuboid (point1) and the final point (point2, opposite diagonal corner).

absolutePoint1, absolutePoint2: worldPoint

The position of VoxelPoints point1 and point2 in physical space (world coordinate) computed using the Image Plane Module.

Consensus Formats

Consensus `tasks.json`

// Single task on RedBrick can be single/multi-series
type Task = {
  // Required on upload and export
  name: string;
  consensus: true; 
  consensusScore: number; // overall consensus score
  consensusTasks: ConsensusTask[]

  // Only required on export
  taskId?: string;
  currentStageName?: string;
  createdBy?: string;
  createdAt?: string;
  updatedBy?: string;
  updatedAt?: string;
};

type ConsensusTask = {
  updatedBy: string;
  updatedAt: string; 
  scores: {secondaryUserEmail: string, score: number}[]
  series: Series[]
}

The `ConsensusTask` Object

The consensus task object contains information about the consensus annotations for this task. There will be a single entry for every annotator who annotated this task. For example, if 3 users annotated each task in your project, the length of the consensusTasks array will be 3.

`updatedBy: string`

The e-mail of the user who annotated the task.

`updatedAt: string`

The datetime when the user last updated the task.

`scores: {secondaryUserEmail: string, score: number}[]`

The scores entry compares the current users' annotations with every other user. The scores array will be of length n-1, where n is the number of users who annotated this task. score is the similarity score between the current user, and secondaryUserEmail.

`series: Series[]`

The series entry for the current user only.

Taxonomy Object

type Taxonomy = {
    orgId: string;
    name: string;
    createdAt: datetime;
    archived: boolean;
    isNew: true;
    taxId: string;
    studyClassify: Attribute[];
    seriesClassify: Attribute[];
    instanceClassify: Attribute[];
    objectTypes: ObjectType[];
}

type ObjectType = {
    category: string;
    classId: number; // [0, n)
    labelType: BBOX | POINT | POLYLINE | POLYGON | ELLIPSE | SEGMENTATION | LENGTH | ANGLE | CUBOID;
    attributes?: Attribute[];
    color?: string;
    archived?: boolean;
    parents?: string[];
    hint?: string;
}

type Attribute = {
    name: string;
    attrType: BOOL | TEXT | SELECT | MULTISELECT;
    attrId: number;
    options?: AttributeOption[];
    archived?: boolean;
    parents?: string[];
    hint?: string;
    defaultValue?: number | number[]; // pre-populated optionId(s) for SELECT/MULTISELECT
}

type AttributeOption = {
    name: string;
    optionId: number;
    color?: string;
    archived?: boolean;
}

PreviousFormats NextExport Structure

Last updated 2 months ago

Items List and tasks.json

Object Reference

Object Glossary

Task

name: string

taskId?: string

currentStageName?: string

createdBy?: string

createdAt?: string

updatedBy?: string

updatedAt?: string

preAssign?: { [ stageName : string ] : string }

classification: { attributes : [ string : boolean ] }

Series

items: string | string[]

name: string

classifications: { attributes: [ string : boolean ] }

instanceClassifications: fileIndex | fileName | [ values: { string : boolean } ]

Common Label Keys

category: string | string[]

attributes: { [ attributeName: string ]: string | boolean }

voxelPoint: { i: number, j: number, k: number }

worldPoint: { x: number, y: number, j: number }

point2D: { xnorm: number, ynorm: number }

fileIndex: number

fileName: string

measurementStats: Dict

Measurement Stats

Video Meta Data

videoMetaData: Dict

frameIndex: number (video)

trackId: string (video)

keyFrame: boolean (video)

endFrame: boolean (video)

Segmentations and segmentMap

segmentations?: string | string[]

mask?: string

BoundingBox

pointTopLeft: Point2D

wNorm, hNorm: number

Polygon

points: Point2D[]

MeasureLength

point1, point2 : VoxelPoint

absolutePoint1, absolutePoint2 : WorldPoint

normal: [number, number, number]

length: number

MeasureAngle

point1, point2, vertex : VoxelPoint

absolutePoint1, absolutePoint2 : WorldPoint

normal: [number, number, number]

angle: number

Ellipse

Landmark

Landmark 3D

Cuboid

Consensus Formats

Consensus tasks.json

The ConsensusTask Object

updatedBy: string

updatedAt: string

scores: {secondaryUserEmail: string, score: number}[]

series: Series[]

Taxonomy Object

Items List and `tasks.json`

`name: string`

`taskId?: string`

`currentStageName?: string`

`createdBy?: string`

`createdAt?: string`

`updatedBy?: string`

`updatedAt?: string`

`preAssign?: { [ stageName : string ] : string }`

`classification: { attributes : [ string : boolean ] }`

`items: string | string[]`

`name: string`

`classifications: { attributes: [ string : boolean ] }`

`instanceClassifications: fileIndex | fileName | [ values: { string : boolean } ]`

`category: string | string[]`

`attributes: { [ attributeName: string ]: string | boolean }`

`voxelPoint: { i: number, j: number, k: number }`

`worldPoint: { x: number, y: number, j: number }`

`point2D: { xnorm: number, ynorm: number }`

`fileIndex: number`

`fileName: string`

`measurementStats: Dict`

`videoMetaData: Dict`

`frameIndex: number` (video)

`trackId: string` (video)

`keyFrame: boolean` (video)

`endFrame: boolean` (video)

Segmentations and `segmentMap`

`segmentations?: string | string[]`

`mask?: string`

`pointTopLeft:` `Point2D`

`wNorm, hNorm: number`

`points:` `Point2D[]`

`point1, point2 :` `VoxelPoint`

`absolutePoint1, absolutePoint2 :` `WorldPoint`

`normal: [number, number, number]`

`length: number`

`point1, point2, vertex :` `VoxelPoint`

`absolutePoint1, absolutePoint2 :` `WorldPoint`

`normal: [number, number, number]`

`angle: number`

Consensus `tasks.json`

The `ConsensusTask` Object

`updatedBy: string`

`updatedAt: string`

`scores: {secondaryUserEmail: string, score: number}[]`

`series: Series[]`