By Sy Mukherjee
May 20, 2019

Happy Monday, readers! I hope you enjoyed your weekend.

The machines just aced a major test in the world of cancer detection.

A joint collaboration between Google A.I. scientists and medical institutions such as Northwestern Medicine, Stanford Health Care and Palo Alto Veterans Affairs, and the NYU-Langone Medical Center found that neural networks and deep machine learning algorithms were just as good as (if not better than) doctors in detecting early stage lung cancers in CT scans.

The research published in the journal Nature is the latest example of A.I.’s potential in the field of radiology, where other studies have also shown that programs informed by massive data sets and neural networks can be a potent primary or secondary opinion to help the doctors charged with reading our medical scans.

According to the study, the algorithm was able to keep pace with doctors when there were available CT scans on hand – and outpace them when they weren’t. “When prior computed tomography imaging was not available, our model outperformed all six radiologists with absolute reductions of 11% in false positives and 5% in false negatives,” the wrote. “Where prior computed tomography imaging was available, the model performance was on-par with the same radiologists.”

Given lung cancer’s status as one of the deadliest around (the disease caused 160,000 American deaths in 2018 alone, making it the most fatal cancer), some firms have focused their efforts on catching it at the earliest possible stage. If study results such as these hold true, our software physician partners could play an important role in such detection.

Read on for the day’s news.

Sy Mukherjee


You May Like