AI news that drew views, Oct. 19–23

Google Health all but invited the blowback when its AI developer-researchers suggested their breast-cancer model may be superior to radiologists’ eyes and generalizable across differing demographics.

Analyzing the scientific literature on medical AI published over the past 46 years, researchers in the U.K. have found the U.S. far ahead of the field for sheer quantity.

The FDA has greenlit a European company’s AI software package designed to streamline workflow for radiologists reading prostate MRIs.

Low-back imaging and interventions are often named among healthcare’s iffiest expenses due to the hit-or-miss predictability of the outcomes.

If big data is to fulfill its potential for advancing the state of modern healthcare, developers of medical AI must be willing to show their work. 

The bytes are accumulating far too fast for any human effort to connect whatever dots might be hiding in the massive muddle. Enter 40 researchers at 11 sites armed with AI and almost $18 million in new NIH funding.

If you’re failing to keep social distance, wear a mask and wear it properly, a red X lights up as a reminder. If you’re doing things right, it’s a green check mark affirming your compliance.

Compared with other major industries, healthcare believes itself to be at significantly lower risk for data breaches now than when the COVID crisis began prompting employers to let staff work from home.

In response to the COVID pandemic, most hospitals have adopted a policy of curtailing elective clinical and surgical services while preserving urgent/semi-urgent and inpatient care in order to conserve personal protective equipment (PPE), and to reduce patient and personnel exposure to infection, while providing for surge capacity in the event of wider dissemination.