Dive Brief:
- Aidoc has received the Food and Drug Administration’s breakthrough device designation for an artificial intelligence feature designed to read chest X-rays and generate draft report text, the company said Thursday.
- The system, called First Read, is expected to generate preliminary reports on more than 100 pre-specified findings, a spokesperson for Aidoc wrote in an email. A radiologist still must sign every report made by First Read. Aidoc declined to provide the full list of findings as the product is not yet on the market.
- Aidoc CEO Elad Walach said in a statement the feature is intended to help reduce the time radiologists spend on reporting and interpretation.
Dive Insight:
The FDA’s breakthrough designation prioritizes devices for agency review and provides more opportunities for feedback. To be eligible, a device must provide more effective treatment or diagnosis of a life threatening or debilitating condition.
Aidoc’s technology is based on the same architecture as another application made by the company to triage multiple findings in a single CT scan. First Read is intended to generate reports and is separate from Aidoc’s other products that help triage and prioritize worklists for radiologists, the company said in an email.
Aidoc raised $150 million in financing in April to advance its AI models and is now used in more than 2,000 hospitals worldwide.
Aidoc’s breakthrough designation comes as medical device firms test using generative AI, a type of technology that generates text and images.
On Thursday, UpDoc received FDA clearance for a patient-facing device that uses large language models, a type of generative AI. UpDoc received the clearance for a platform that can integrate with a provider’s electronic health record and reach out to patients in between visits. For example, it could adjust insulin dosing for a person with Type 2 diabetes whose blood glucose is drifting out of range, within parameters approved by a physician.
While companies are experimenting with using generative AI, the FDA has not yet laid out a framework for how it plans to regulate the technology. The agency has hosted two advisory panels on the technology and issued draft guidance in early 2025 on AI transparency, mitigating bias and post-market monitoring. The agency has not yet finalized the draft.
