Supervised Machine Learning
February 1, 2022
Owlet Smart Sock
February 1, 2022Neural Analytics aims robotic transcranial doppler at COVID-19 patients
FDA permits marketing of artificial intelligence algorithm for detection of wrist fractures
The OsteoDetect is a software intended to recognize many types of wrist fractures in adult patients. The software utilizes an AI algorithm to examine and highlight areas of the distal radius during the review of posterior-anterior and medial-lateral X-ray images for signs of distal radius fractures.
The software identifies the location of the fracture in the image to aid the provider make a better assessment and diagnosis. A retrospective investigation of 1,000 radiographs consisted on the assessment of the independent performance of the image analysis algorithm for identifying wrist fractures and accuracy of its localization versus the performance of three board-certified orthopedic hand surgeons with aid from OsteoDetect. The research demonstrated that the providers’ performance in recognizing wrist fractures was increased using the software, including improved sensitivity, specificity, positive and negative predictive values, in contrast with their unaided performance according to standard clinical practice.(1,2,)
Bay Labs’ EchoMD AutoEF Software Receives FDA Clearance for Fully Automated AI echocardiogram analysis
Ejection fraction is the single most widely utilized metric of cardiac function and is utilized in many clinical settings. The EchoMD AutoEF’s algorithm removes the need to manually select views, keep the best clips, and manage them for quantification, a regularly time-consuming and highly variable method.
Bay Labs’ application of artificial intelligence for image choice and automated EF determination will allow clinicians with varying levels of experience to receive an accurate evaluation of ventricular function and description of the echocardiograms.
Unlike popular technologies, EchoMD AutoEF automatically analyzes all the relevant digital video clips of cardiac cycles from a patient’s echocardiography study, ranks them
according to image features, and selects the most suitable ones to determine the EF. The software algorithm “determines” its choice of the clip and EF estimation after being trained on a carefully curated dataset of over 4,000,000 images, representing 9,000 patients, creating a massive archive of meticulously curated studies optimized for deep training algorithm construction.(3)
Neural Analytics aims robotic transcranial doppler at COVID-19 patients
The algorithms aid in recognizing problems, including blood clots, changes in velocity, occlusions, the size of clots, and whether emboli are moving through the arteries. Patients can be scanned in less than five minutes at the bedside, an added advantage due to the current pandemic as patients in isolation units can be evaluated with minimal exposure to clinical workers.
(4,5)
References