The FDA granted de novo marketing authorization for Paige Prostate, artificial intelligence (AI)-driven software that improves detection of prostate cancer. The clinical study submitted to the FDA demonstrated that using Paige Prostate resulted in a 7% improvement in sensitivity in correctly diagnosing cancer, increasing from 89.5% to 96.8%.
The question of whether an artificial intelligence (AI) algorithm can be an inventor has been making the rounds in the past couple of years, and the question came up again in the U.S. District Court for the Eastern District of Virginia. Stephen Thaler, who developed the Device for the Autonomous Bootstrapping of Unified Sentience (DABUS) algorithm that has been credited with two inventions, failed to persuade the court that an algorithm qualifies as an “individual,” and thus patents must still be assigned to humans, at least where the U.S. Patent and Trademark Office is concerned.
PARIS – Researchers from the department of radiation oncology at the European Hospital Georges Pompidou (HEGP) and Stanford University School of Medicine have together developed a new artificial intelligence (AI) prediction tool for patients diagnosed with prostate cancer. These researchers have just published a validation of this interpretable AI model in Cancers. “It’s a question of distinguishing patients at risk of mortality from aggressive cancer that is spreading rapidly, from patients who might have far less aggressive cancer and who are not likely to die from it in under 10, 15 or even 20 years,” Jean-Emmanuel Bibault, radiotherapy oncologist at HEGP, told BioWorld.
The World Health Organization’s (WHO) guidance for ethics and governance for artificial intelligence (AI) in health discusses several issues regarding regulation, including the question of transparency for the algorithm’s source code. The WHO paper is not prescriptive on this and several other issues, however, raising the prospect that regulatory entities will not be discouraged from adopting policies that run afoul of intellectual property concerns and thus impede advances in AI.
TORONTO – Femtherapeutics Inc. is combining artificial intelligence-driven machine learning and 3D design to manufacture a device for relieving urinary incontinence and discomfort in women suffering pelvic organ prolapse. The custom-made pessary is intended to support vaginal tissues displaced because of the condition, replacing conventional pessaries that company officials said can result in irritation and penetrate soft tissues causing bleeding.
PARIS – Gleamer SAS said Radiology published compelling results from a study evaluating the performance of its artificial intelligence (AI) system Boneview, which helps radiologists and emergency doctors detect and localize fractures. This is the first study to evaluate the performance of AI-assisted health care professionals in locating bone fractures on all appendicular X-rays.
Researchers at the University of Washington reported in the May 31, 2021, issue of Nature Medicine that artificial intelligence (AI) algorithms meant to recognize COVID-19 infections based on chest X-rays picked up on confounders, selecting “shortcuts” such as patient age or positioning in the X-ray as a basis for their predictions.
Edwards Lifesciences Corp. has won the FDA’s nod for its Acumen hypotension prediction index (HPI) software with the Acumen IQ finger cuff. Prior to receiving this 510(k) clearance, the low blood pressure prediction software was only available for patients using an invasive arterial line. The Acumen HPI software uses data gleaned from the Acumen IQ cuff, along with machine learning, to alert clinicians if a patient is at risk of hypotension, drops in pumping pressure that reduce the flow of oxygenated blood and nutrients to the body’s vital organs.
Becton, Dickinson & Co. (BD) is looking to ease the burden on microbiology labs with the U.S. launch of its urine culture analysis imaging application by automating microbiology processes. Powered by artificial intelligence (AI), the app is intended for use on the device manufacturer’s Kiestra lab automation incubation and imaging system.
LONDON – Ibex Medical Analytics Ltd. has secured a second CE marking for its Galen artificial intelligence decision support system for automated interpretation of tumor biopsies, adding breast cancer to the approval in prostate cancer secured just over a year ago. The CE mark follows results from a blinded, multicenter clinical study at Institut Curie in France and Maccabi Healthcare Services in Israel. Ibex says that in the trial Galen breast demonstrated very high accuracy in detecting various types of breast cancer.