Skip to content

Task Management

Overview

Task management in Flywheel enables structured human review workflows for clinical trials, reader studies, and quality control processes. The task system provides a framework for assigning image review tasks to radiologists and other readers with customizable viewer configurations, validation rules, and data capture forms.

Feature Availability

Tasks are currently behind a feature flag and are being released to a limited number of customers for testing. If you are interested in hearing more about Tasks, please contact your Flywheel Representative or contact Flywheel Support.

Key Concepts

Read Tasks

Read tasks are structured workflows that guide readers through specific image interpretation and data collection activities. Each task includes a defined set of actions, validation schemas, and data capture points to ensure consistent and complete reviews.

Components of a Read Task:

  • Project context and data container
  • Assigned reader and due date
  • Viewer protocol defining the workflow
  • Task status and completion tracking
  • Electronic signature support for regulatory compliance

Use Cases:

  • Quality control reviews of incoming imaging data
  • Structured reader studies with annotation requirements
  • Clinical trial endpoint assessments
  • Multi-reader consensus workflows
  • Blinded and double-blinded image reviews

Learn about read tasks

Viewer Protocols

Viewer protocols define the specific image interpretation workflow and viewer configuration for read tasks. They control what readers see, how they interact with images, and what data they must collect during task completion.

Protocol Configuration Options:

  • Annotation Labels: Custom labels, colors, and validation rules for image annotations
  • Hanging Protocol: Viewport layouts, modality mappings, and series display rules
  • Hotkeys and Mouse Actions: Custom keyboard shortcuts and mouse behaviors for efficient workflows
  • Toolbar Customization: Show or hide specific viewer tools based on workflow needs
  • General Settings: Draft saving, timing capture, annotation overlap rules, and more

Form Definitions:

  • Custom forms for structured data capture
  • Conditional logic and validation schemas
  • Required fields and dropdown options
  • Integration with annotation data

Learn how to create viewer protocols

Task Workflow

Creating and Assigning Tasks

  1. Define Viewer Protocols - Create protocols in project settings that define viewer configuration and form requirements
  2. Create Tasks - Assign tasks to specific data containers (sessions, acquisitions) with a selected protocol
  3. Assign Readers - Designate specific users to complete tasks
  4. Set Due Dates - Establish timelines for task completion
  5. Track Progress - Monitor task status and completion through the task list

Completing Tasks

  1. Access Assigned Tasks - Readers see their assigned tasks in the task list
  2. Open Task in Viewer - Launch the configured viewer with protocol-specific settings
  3. Perform Review - Create annotations and complete form questions as defined by the protocol

Task Management

  • View Task Lists - See all tasks for a project with status, assignee, and due date information
  • Filter and Search - Find tasks by status, assignee, protocol, or container
  • Reassign Tasks - Transfer tasks between readers when needed
  • Monitor Progress - Track completion rates and time spent on tasks

Common Workflows

Quality Control Pipeline

  1. Configure a viewer protocol with required quality checks and pass/fail criteria
  2. Create tasks automatically for new data acquisitions using the SDK
  3. Assign tasks to QC readers for review
  4. Track completion and flag issues for follow-up

Multi-Reader Study

  1. Define protocols with specific annotation requirements and data collection forms
  2. Create identical tasks for multiple readers on the same dataset
  3. Compare annotations and form responses across readers for consensus analysis

Best Practices

  • Test protocols thoroughly on sample data before deploying to production studies
  • Provide clear instructions in protocol descriptions and form labels
  • Use validation rules to ensure data quality and completeness
  • Train readers on protocols and viewer tools before assigning critical tasks
  • Monitor task progress regularly to identify bottlenecks or issues
  • Version protocol changes by duplicating and modifying rather than editing active protocols
  • Document protocol details for reproducibility