With AI becoming a prominent topic of conversation in many fields, it is not surprising to see it permeating healthcare so heavily. Though many teams are in the midst of extensive research regarding potential machine learning solutions, there are only a handful that have received FDA-approval. This number is rising rapidly, however, increasing from 14 to 26 since this January according to Dr. Eric Topol. Read on for a summary of the most recent approvals to see the versatility of AI in the medical space.
1. Apple Watch ECG
Approved by the FDA in September of last year, the ECG technology on the Apple Watch Series 4 uses electrodes to capture heart rhythm irregularities. This technology is the first consumer-available product that allows users to take an ECG from their own wrist, and can provide critical data to physicians. It can detect atrial fibrillation, a dangerous arrhythmia that can result in stroke if left untreated. AFib is the most common heart arrhythmia experienced in patients and is estimated to affect up to nine percent of those over the age of 65, two percent of younger populations in the U.S. By providing a convenient and rapid means of measuring heart rhythm, the Apple Watch ECG has the potential to lower this statistic.
Aidoc is an app that allows radiologists to identify acute intracranial hemorrhages in head CT scans. Approved last summer, the system was the first deep learning technology designed to assist radiologists in assigning urgency to patients’ injuries. This solution reviews images immediately after the patient is scanned, and informs the radiologists of potentially dangerous cases to aid in prioritizing cases. Being that some of these hemorrhages can be life-threatening, this is an imperative task that the radiologist is required to fulfill. Aidoc can significantly reduce turnaround time by analyzing these images, and also increases the radiologists confidence in decision-making.
3. ProFound AI by iCAD
Also approved last August was the ProFound AI, produced by medical technology leader iCAD. This system provides a precise, AI-driven cancer detection system that analyzes breast tomosynthesis images, or 3D mammograms, for malignancy. The app was built on deep-learning technology, and assists radiologists in reading tomosynthesis images by increasing detection rates, reducing unnecessary recalls and false positives, and speeding up read times.
4. Zebra Medical Vision
Using ECG-gated CT scans, the Zebra Medical Vision platform calculates the degree of calcification in a patient’s coronary artery. This arterial defect narrows the blood vessel’s diameter and can lead to major adverse cardiovascular events. By using this FDA-approved diagnostic tool, radiologists can better diagnose the severity of the arterial defect to improve patient outcomes. This technology was approved in July 2018. Zebra has also generated AI-algorithms that can detect bone density, fat in the liver, and emphysema in the lungs from images.
5. EchoMD AutoEF by Bay Labs
Bay Labs received approval for their EchoMD AutoEF deep learning software last June. This technology was found to have less variability in analyzing left ventricular ejection fraction (EF) than most cardiologists in a study conducted with the Minneapolis Heart Institute. The AI platform automatically reviews digital video clips from a patient’s echocardiogram and selects the best ones for calculating EF. This system was trained using a dataset of over four million images from 9,000 patients, and aims to assist cardiologists with rapid image analysis. Typically expressed as a percentage, EF indicates how efficiently the heart pumps blood. For example, an EF of 50 percent would mean the left ventricle pumps out half of the blood within it each contraction. This indicates how well the heart pumps blood and can indicate heart failure.
Check out more including a list of the approved AI platforms from Dr. Topol and Nature Journal below.
When I recently (Jan ’19) reviewed #AI @US_FDA approvals/clearances there were 14.https://t.co/EvUweuCumF @NatureMedicine Now, there are at least 26!
Let me know if there are ones that I’m missing. pic.twitter.com/v0PUCvs7xP
— Eric Topol (@EricTopol) April 20, 2019
Source: Nature Medicine