Diagnostics Pipeline: More AI-Reliant Platforms Being Approved by the FDA

Artificial intelligence (AI) seems to have appeared in the healthcare space virtually overnight. In the space of perhaps a couple of years, it went from a theoretical concept to an application being used in a wide range of services, including claims adjudication and customer service.

And now, AI appears to not only be making its presence known in the laboratory space, but the U.S. Food and Drug Administration (FDA) is taking notice.

In recent weeks, the FDA issued three marketing approvals for tests that do not rely on tissue samples per se, but instead leverage both imaging and AI to make diagnoses.

David Latshaw, a biomolecular engineer by training and co-founder of Philadelphia-based pharmaceutical development firm BioPhy, observed that “the shift from traditional tissue testing to AI and imaging technologies is driven by the demand for increased precision, speed, and minimally invasive diagnostic methods.” 

The first significant approval is for Massachusetts-based Hologic for its Genius™ Digital Diagnostics System, which combines volumetric imaging technology and deep-learning based AI to identify precancerous lesions and even cervical cancer cells.

The platform uses digitally scanned images from laboratory Pap smear tests. AI is applied to give cytologists and pathologists a better guide of what they should be looking for in their exams. According to Hologic’s own studies, the platform reduced false negatives of high-grade squamous intraepithelial and more severe lesions by 28 percent compared to standard reviews under a microscope.¹

“The system delivers more actionable and accurate insights for laboratories and healthcare professionals to enhance patient care,” said Jennifer Schneiders, Hologic’s president of diagnostic solutions, in a press release.¹

Hologic had previously released the Aptima^® HPV Assay, which was the first FDA-approved mRNA-based test for detecting HPV. It also released the ThinPrep^® Pap test, which was the first FDA-approved liquid cytology test.

The company said the Genius platform should be available in the US sometime during the first part of 2024. It is already commercially available in Europe, Australia, and New Zealand.¹

Avoiding costly, more invasive tests

Another FDA de novo approval was granted to San Francisco-based brain health company Darmiyan for BrainSee, a clinical test that combines brain imaging, cognitive tests, and AI to determine the progression of amnestic mild cognitive impairment into Alzheimer’s disease within a five-year timeframe. The intent of the BrainSee platform is to help keep patients who are at low risk of developing Alzheimer’s from having to undergo batteries of unnecessary procedures and testing, which can include lumbar punctures, or amyloid PET scans that require an injection of a radiotracer substance into the body. For those patients diagnosed as a high risk of progressing to Alzheimer’s dementia, it buys time for the patient and their family to prepare for the disease’s progression. The platform had previously received a breakthrough designation from the FDA in 2021.²

“Our vision is to redefine brain health screening and monitoring standards and impact the lives of patients and their family members in a meaningful way,” said Darmiyan CEO and co-founder Padideh Kamali-Zare, in a press release.³ The company’s other co-founder and chief medical officer, Kaveh Vejdani, MD, didn’t respond to a request seeking comment.

The third test that received FDA de novo approval was for IMVARIA’s Fibresolve platform, which uses AI to detect lung fibrosis, particularly idiopathic pulmonary fibrosis (IPF).⁴ There are about 100,000 Americans with IPF, and another 30,000 to 40,000 cases are diagnosed every year. IPF in its mildest form might lead to fatigue, but it can also cause blood clots in the lungs and even lung cancer.⁵

As with the BrainSee platform, Fibreresolve is intended to avoid costlier and more invasive diagnostic tests, particularly for patients with suspected interstitial lung disease.

“Fibresolve serves as an adjunct to clinicians in assessing patients with suspected lung fibrosis to provide a diagnostic subtype classification, potentially facilitating proper treatments at an earlier stage of the disease process,” said IMVARIA CEO Joshua Reicher, MD, in a press release.⁵

Although there will likely be more such approvals in the near-term, Latshaw cautioned that AI-driven testing still has various obstacles to overcome.

“Integrating AI into diagnostics presents challenges such as regulatory hurdles, ensuring data security, and validating the reliability of AI algorithms,” Latshaw said in an email. “Overcoming these obstacles is crucial for the successful adoption of these technologies.”

Takeaway: AI is leading to a new generation of diagnostic platforms that could wind up supplementing or even supplanting many more traditional forms of testing.

References:

1. 1. https://www.businesswire.com/news/home/20240131786693/en/Hologic-Announces-First-and-Only-FDA-Cleared-Digital-Cytology-System-%E2%80%93-Genius%E2%84%A2-Digital-Diagnostics-System

1. 1. https://brainsee.ai/

1. 1. https://www.prnewswire.com/news-releases/darmiyan-receives-fda-approval-for-brainsee-the-first-prognostic-test-for-predicting-likelihood-of-progression-to-alzheimers-dementia-302033616.html

1. 1. https://www.businesswire.com/news/home/20240116509267/en/IMVARIA-Announces-FDA-De-Novo-Marketing-Authorization-of-Fibresolve-an-AI-Biomarker-in-Lung-Fibrosis-and-the-Adoption-of-Novel-CPT-Billing-Codes-by-the-American-Medical-Association

1. 1. https://www.healthline.com/health/managing-idiopathic-pulmonary-fibrosis/ipf-facts#prevalence

****************

Key recent FDA approvals related to automated and AI-related platforms, as well as to over-the-counter tests and products include: