Accelerating Clinical Adoption of Machine Intelligence (MI) Applications in Medical Imaging

Institute or Center: National Institute of Biomedical Imaging and Bioengineering (NIBIB)

Project: Accelerating Clinical Adoption of Machine Intelligence (MI) Applications in Medical Imaging

Skills sought:

  • cloud architecture/engineering expertise (AWS, Google)
  • demonstrated leadership, management and communication skills
  • comfortable with data science and software engineering principles
  • knowledge of advanced data management:
    • FHIR® standards and FAIR principles
    • challenges around provenance and quality of dynamic datasets to support open science
    • ethical issues on use and reuse of data
  • experience with novel concepts of data privacy/data security –cryptography, provenance (blockchain, etc.)
  • Proven ability to build trust and form alliances with diverse, nationwide stakeholders

About the project: NIBIB seeks a DATA scholar to revolutionize the practice of medicine by accelerating the clinical adoption of MI applications in medical imaging through development of the blueprint for a proposed one-of-a-kind central national resource.

The Scholar will:

  • create a cloud-based infrastructure for the most comprehensive collection of diverse and interoperable medical image datasets in a large repository with a low barrier to access.
  • share this aggregated, cleaned, and curated data with researchers across the country who can collaborate to create MI methods/tools to accelerate clinical applications.
  • lay the foundation for this platform to allow integration and analysis of disparate data from many sources.
  • expose potential data security issues – protect privacy, develop and test de-identification algorithms.
  • prepare to measure data quality and normalize data coming from heterogeneous sources.
  • help develop and validate methods and tools to enable discovery and dissemination.

The Scholar will establish a full suite of requirements to create this resource in close collaboration with both internal and external stakeholders and create a plan for the project’s implementation and sustainability.

About the work: Although the U.S. generates over 1 billion medical images annually, only about 0.01% make it into useful datasets. These are often small, secluded in specific medical centers, and reflective of local populations only. This untenable state renders it difficult to develop universally applicable and transferable MI algorithms that accurately diagnose actionable pathology. To address this NIBIB is establishing a public-private collaboration to accelerate validation of machine learning applications in medical imaging to improve patient outcomes and widen clinical adoption. The Scholar has an opportunity to address an unmet need identified by the medical community by developing centralized and sustainable datasets, methods, and tools.

Datasets involved: Digital Imaging and Communications in Medicine (DICOM ®) medical images; these datasets are of importance to the entire medical imaging community (practitioners, industry, government, etc.)

Why this project matters: Most medical specialties will benefit from the proposed collaboration, which will create a national resource and provide opportunities for the assimilation of advances in image-based diagnoses with clinical metadata to improve patient outcomes, thereby revolutionizing U.S. healthcare.

Work Location: Bethesda, MD

Work environment: The Scholar will report to the NIBIB Director, or a senior designee, as a senior special project leader. The Scholar will interact closely with a team of leaders and content-expert staff within the Division of Health Information Technology, who will serve as the NIBIB technical contacts and provide guidance relevant to navigating the NIH. The Scholar will also collaborate with several stakeholders from the medical imaging community – professional societies of imaging physicists and physicians, various hospitals and healthcare systems, industry (imaging vendors, healthcare-related software companies), and the government.

To apply to this or other DATA Scholar positions, please see instructions here: datascience.nih.gov/data-scholars

This page last reviewed on January 29, 2020