Skip to main content
The NCI Community Hub will be retiring in May 2024. For more information please visit the NCIHub Retirement Page:https://ncihub.cancer.gov/groups/ncihubshutdown/overview
close

High-Throughput Truthing Year 3

Year 3: High-throughput truthing of microscope slides to validate artificial intelligence algorithms analyzing digital scans of pathology slides: data collection to create the medical device development tool (MDDT)

In Year 3 of the High Throughput Truthing project, the team will collect data at various collaborating sites and conferences.
Check out what was completed in Year 2 and Year 1

Pitch: We are launching a project to crowdsource pathologists and collect data (images + pathologist annotations) that can be qualified by the FDA/CDRH medical device development tool program (MDDT). The MDDT qualified data along with a statistical software package for data analysis would be available to any algorithm developer to be used to validate their algorithm performance in a submission to the FDA/CDRH.

Researchers from the U.S. Food and Drug Administration, alongside academic collaborators, are collecting pathologist annotations as data for AI/ML algorithm validation for tumor infiltrating lymphocyte (TIL) detection and quantitation. We are asking board-certified anatomic pathologists and anatomic pathology residents to score ROIs as part of a research study. We anticipate that this task will take participants a total of 60 minutes. The data will be used to inform the agency’s approach to novel algorithm validation, ensuring high quality commercial products with a faster FDA-pipeline to approval.

HTT Data collection before the USCAP meeting on February 28 at the JW Marriott Los Angeles LA LIVE


Sign up to collect data at the pre-USCAP event here

Recruiting Advertisement:

  • HTTatUSCAP.pdf (222 KB, uploaded by Brandon D. Gallas 4 years 1 month ago)

Data Collection by Pathologists:
You will be presented pre-selected fields of view (FOV) digitally or on a microscope (Figure below, left). For each FOV, you will be prompted for the tumor-associated stromal TIL density, which is a number from 0-100. Once you have determined stromal TIL density for each FOV, you will type or click the value using the software’s guided user interface (Figure below, right). Study training materials on TIL density evaluations can be found in the following section and may be completed ahead of time.

(Image(eeDAPsystemComponents.png, 800px, text here) failed - File not found) eeDAPinterfaceHTT.png (Left: eeDAP system components, labeled. Read more here. Right: Screenshot of eeDAP user interface for stromal TIL density data collection)

Training documents:
TILs evaluation manuscript, TILs training slides – Participating pathologists can review the training slides (required to participate), and read the related manuscript if possible. These materials were created by the TILs in breast cancer working group.

eeDAP – Annotations will be collected in digital and microscope modes. For the microscope mode, we will use eeDAP, an Evaluation Environment for Digital and Analog Pathology (eeDAP), a registration system between the microscope and digital whole slide images.

Consent Form:
HTT_IRBinformedConsent.pdf (47 KB, uploaded by Brandon D. Gallas 4 years 1 month ago)

Created on , Last modified on