Marilyn M. Bui, MD, PhD; Moffitt Cancer Center and Stephen D. Jett, PhD; National Cancer Institute

 

One of the advantages of digital pathology and whole slide imaging is to enable sharing of pathology images to eliminate the limitation inherited by microscope, location, and time using conventional glass slides. The second advantage is within the power of image analysis which has shown it boundless opportunities. Molecular pathology has enriched the tool box for pathologists and scientists to investigate the diagnostic, prognostic and predictive factors of diseases, especially cancer.  With the advancement of digital pathology and computer assisted artificial intelligent technology, it is inevitable that digital health will play an important role in the strategic planning of cancer research and precision medicine delivery.

 

At my home institution Moffitt Cancer Center (https://www.moffitt.org), which is a NIH/NCI designated comprehensive cancer center in Florida, digital pathology is used every day in cancer research, education and patient care. Digital images connect radiologists, pathologists, other subspecialty physicians and scientists in the discovery of novel cancer biomarkers and treatment modalities.  There are collaborative initiatives established among cancer institutions/hospitals to build information sharing network nationally. It is timely that these initiatives begin to take a closer look of digital pathology, whole slide imaging and artificial intelligence to lavage this powerful tool to augment cancer research and precision medicine delivery.

 

As a member of Digital Pathology Association, you are welcome to response to the NCI Requests Input on the Development of an Imaging Data Commons, see below.

 

In 2016, a Blue Ribbon Panel (BRP) was established, as part of the Beau Biden Cancer Moonshot, to make key recommendations that would support the Moonshot goals of accelerating progress in cancer research and breaking down barriers to developing new treatments.  The Enhanced Data Sharing Working group of the BRP recommended the development of a National Cancer Data Ecosystem to address the “lack of searchable and interconnected data repositories with associated tools and services.”  This Ecosystem would “collect, share, and interconnect a broad array of large datasets,” and provide the necessary tools to “contribute and analyze data.”

 

NCI will play an important role in supporting development of such an Ecosystem, providing components that allow for access to and sharing of consistent and harmonized cancer research data. To that end, the NCI has launched a series of initiatives to create an NCI Cancer Research Data Commons (CRDC), which will become a key contribution to this Ecosystem. The Genomic Data Commons and the Cancer Genomics Cloud Resources define some of the core elements required for development of a CRDC.  The vision for the CRDC is a virtual, expandable infrastructure that provides secure access to diverse data types, allowing users to analyze, share, and store results, leveraging the storage and elastic compute of the cloud.  One of those data types is imaging.

 

As highlighted in a previous blog post, images and the data contained within and associated with them play significant roles in research, diagnosis, and treatment of cancer.  To support the integration of imaging data in cancer research, the NCI is initiating the development of an Imaging Data Commons (IDC), a data node of the CRDC.

 

Input from the imaging community will be crucial to the development of an IDC that meets the community’s needs.  To gather that input and feedback, the NIH has issued a Request for Information (RFI) NOT-CA-18-060 Input on Development of the NCI Imaging Data Commons.

 

We are interested in learning more about:

 

  • Who are the potential users of an Imaging Data Commons?
  • What are the primary uses of imaging data for the cancer research community (generation, analysis, training sets, etc.)?
  • What computing resources are necessary for analysis, visualization, sharing, and feature mining of imaging data?
  • What are the imaging community’s views on cloud and scalable computing resources?

 

We invite input from all who foresee that the IDC could facilitate their work.  Expected contributors include, but are not limited to, basic researchers, computer scientists, tool developers, data scientists, bioinformaticians, and clinicians.

 

Responses are due May 4, 2018.  Please send responses, as well as any questions pertaining to the RFI, to NCIIDCRFI@mail.nih.gov.