NIH Challenge Seeks Early Alzheimer’s Indicators with A.I.

(B_Me, Pixabay. https://pixabay.com/photos/pedestrians-people-busy-movement-400811/)

7 Sept. 2023. A new crowdsourced competition seeks more robust early indicators of Alzheimer’s disease and related dementias using analytics powered by artificial intelligence. The challenge with a total purse of $650,000 is sponsored by National Institute on Aging or NIA, part of National Institutes of Health, and hosted by the data science crowdsourcing company DrivenData, with an infrastructure provided by challenge technology company HeroX.

NIA says early detection of Alzheimer’s disease and related dementias is hampered by a lack of sensitive tools to identify these disorders in their earliest stages, when more treatment options are available. The agency says it created the challenge to find solutions for all populations affected by dementias like Alzheimer’s disease, particularly those individuals often left behind by current detection techniques. Challenge participants are expected to specifically address health disparities among older segments of the population spelled out in NIA’s health disparities research framework that identifies environmental, sociocultural, behavioral, and biological factors related to research on aging.

The competition has three parts, beginning with identification of more inclusive and shareable data sets on variables contributing to cognitive decline and aging that offer a more representative foundation than available today for building algorithms to detect Alzheimer’s and related diseases. Participants are expected to identify a wide range of sources, including research findings funded by NIH or others, real-world data such as electronic medical records and insurance claims, social media posts, and data from personal devices.

First submissions in Jan. 2024

The competition’s first stage expects to award a total of $200,000 that includes bonuses of $25,000 for proposals addressing populations disproportionately impacted by Alzheimer’s and related dementias, and data sets addressing biases  in current sources. Summary drafts for this part of the challenge are due on 17 January 2024, with full submissions due on 31 January. Judging of entries is done in two stages, first for finalists and then winners, with the winners announced in Sept. 2024.

The second part of the challenge asks participants to write algorithms and analytics using the data sources generated in part 1, judged on accuracy, openness, addressing biases, and ability to explain their mechanisms, scheduled to begin in Sept. 2024. In the competition’s third stage, participants are asked to put together their work on data source identification and algorithm building into a single package for early detection of Alzheimer’s disease and related dementias, with top teams presenting their solutions at an “innovation event.” The third part of the challenge is scheduled for Mar. to Sept. 2025.

While NIA is sponsoring the competition, the challenge is offered through DrivenData, a social enterprise organization staging competitions with outcomes affecting public policies and civil society, based on artificial intelligence and data science. DrivenData, in Denver, Colorado, hosts the competitions and provides a process for evaluating machine learning and other algorithms and analytics submitted by participants. The challenge technology is provided by HeroX in Vancouver, British Columbia and several sites in the U.S. HeroX offers its technology as a turnkey service with templates for competition design, crowdsourcing promotion, evaluation, awards, and project management.

“Being able to detect dementia diseases early can transform outcomes for impacted patients,” says Kal Sahota, CEO of HeroX in a statement released through Cision. “This challenge,” adds DrivenData co-founder Greg Lipstein, “is a powerful opportunity to leverage advances in data and A.I. to equip care teams with greater capabilities for early prediction.”

More from Science & Enterprise:

• Industry, Univ. Labs Partner on Parkinson’s Digital Biomarkers
• A.I. Analytics Detect Heart Failure in Single Ultrasound View
• NIH Funds Depression Game Therapy Clinical Trial
• Digital Biomarkers Shown to Highlight Parkinson’s Disease
• Neuro Biomarkers, Home EEG Assessed in Upcoming Trial

We designed Science & Enterprise for busy readers including investors, researchers, entrepreneurs, and students. Except for a narrow cookies and privacy strip for first-time visitors, we have no pop-ups blocking the entire page, nor distracting animated GIF graphics. If you want to subscribe for daily email alerts, you can do that here, or find the link in the upper left-hand corner of the desktop page. The site is free, with no paywall. But, of course, donations are gratefully accepted.

*     *     *