Number One Cancer Diagnostic

This non-invasive AI-based diagnostic tool detects prostate cancer with near-perfect accuracy while preventing unnecessary biopsies.

Nick Bild
3 years agoMachine Learning & AI
(📷: H. Kim et al.)

According to the Centers for Disease Control and Prevention, prostate cancer is the second most common cancer among men in the United States. Given the prevalence of this type of cancer, it may come as a surprise that screenings for the disease are notoriously inaccurate.

Initial screenings are done with the serum prostate-specific antigen test, which has a false positive rate of 80%. As a result, many unnecessary, invasive biopsies are performed which can lead to internal bleeding, pain, and unnecessary treatments.

The state-of-the-art in prostate cancer detection may soon be significantly advanced by the work of a team from the Korea Institute of Science and Technology. They have developed a non-invasive test that can detect prostate cancer in only 20 minutes, and with near-perfect accuracy.

The researchers made use of an electrical-signal-based ultrasensitive biosensor that is able to detect trace cancer factors present in urine. The minute quantities and complex patterns of these factors have previously prevented them from being useful diagnostic markers. To gain insight from these factors, the team trained an AI algorithm to recognize how their patterns correlate with the presence of prostate cancer.

In a test of 76 urine samples, greater than 99% were classified correctly by the new algorithm. Very impressive results, to say the least.

In addition to improving patient care, the researchers also see their new technique as a means to reduce medical costs. They hope to be able to extend this tool to diagnose other types of cancer in the future.

Nick Bild
R&D, creativity, and building the next big thing you never knew you wanted are my specialties.
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles