Clinical AI technology company Aidoc and Amazon Web Services (AWS) announced a landmark partnership earlier this year: a multiyear investment on behalf of the cloud services giant to help develop and optimize Aidoc’s CARE framework and respective foundation model.
CARE entails “a groundbreaking clinical-grade foundation model” trained with “millions of exams and [leverages] advanced multimodal capabilities” as a means to empower “a new benchmark for precision and speed in developing clinical AI solutions.”
AWS’ investment is already paying for itself. Last week, Aidoc announced that it has secured FDA clearance for an AI solution powered by its foundation model, marking a huge step forward for the company’s work in clinical medicine. Specifically, the approval is for Aidoc’s Rib Fractures triage solution, built on the company’s CARE1 framework. Leveraging the company’s proprietary aiOS platform, the newly approved tech will further empower the analyses and insights generated from medical data to better inform radiology diagnostics and interventions in the clinical setting.
Elad Walach, co-founder and CEO of Aidoc, explains that clinical AI has uncovered a massive field and has incredible potential to improve patient care: “in the last year alone, we have made incredible progress…but we are just getting started.” Walach explains that the numerous developments the company has rapidly progressed on and its partnership with AWS are “ways to move the needle forward on innovation” and ultimately, improve patient outcomes.
Dan Sheeran, General Manager of Healthcare and Life Sciences at AWS comments: “We are particularly excited to partner with Aidoc because most companies do not even move past the pilot stage; most companies do not take into account the real word environments faced by physicians and patients; we need to start with the reality faced on the ground.” Only then, Sheeran explains, can enterprises come up with actual, usable solutions for bedside clinicians.
Lately, it seems that the entire AI ecosystem is increasingly shifting its focus to the clinical realm. Google, for example, continues to invest heavily in its Med-PaLM 2 large language model specifically developed for healthcare and clinical settings; in fact, one of the main goals of its work is to build and test “AI models with the goal of helping alleviate the global shortages of physicians, as well as the low access to modern imaging and diagnostic tools.” Other smaller companies are increasingly focusing on specific and niche fields in which AI can be used for clinical applications. For example, significant investment is going into dermatology and pathology applications to leverage AI as a means to develop better diagnostic tools. Radiology itself is a field that is exploding with a variety of AI use-cases and potential applications, especially with the significant growth of image detection and computer vision technology.
Undoubtedly, the work in this arena is still in its first innings, and hence, companies will have to remain patient to understand how these applications will truly provide value to patients. Assuredly, more developments will emerge in the year ahead.
