AI-Based Urine Test Diagnoses Prostate Cancer with Almost 100% Accuracy

Although prostate cancer is one the most common types of malignancy in men, diagnosis is typically made on the basis of the Prostate-specific antigen (PSA), with an accuracy as low as 30%. Given how unreliable PSA-based testing can often be, many patients require invasive biopsy which often leaves them with long-term side effects, such as pain and bleeding.

To address the situation, researchers from the Korea Institute of Science and Technology (KIST) have recently developed an AI algorithm which, coupled with an electrical-signal-based ultrasensitive biosensor, can diagnose prostate cancer within 20 minutes with almost 100% accuracy.

Advanced biosensor and AI algorithm might finally address the problem of wildly inaccurate prostate cancer diagnosis. Image: Darko Stojanovic via, free licence

Commenting on the findings, Professor In Gab Jeong at the Asan Medical Centre said their smart biosensor could also be used for the precise diagnosis of many other types of cancer based on urine sampling alone.

The semiconductor biosensor was engineered to simultaneously measure trace amounts of four different cancer factors in urine. Thus far, cancer factors – present in urine only at low concentrations – have been used for classifying risk groups, rather than for precise diagnosis.

Training of the AI system was performed using the correlation between the four cancer factors obtained from the novel biosensor. Once ready, the algorithm was deployed to analyse complex patterns of the detected signals.

After performing tests on 76 urinary samples, the researchers found the algorithm to be capable of diagnosing prostate cancer with near-perfect accuracy – an achievement that could eventually improve the lives of millions of men around the world.

“For patients who need surgery and/or treatments, cancer will be diagnosed with high accuracy by utilizing urine to minimize unnecessary biopsy and treatments, which can dramatically reduce medical costs and medical staff’s fatigue,” Jeong said.

A study describing the development of the new system was published in the latest issue of the journal ACS Nano.


Notify of
Inline Feedbacks
View all comments
Would love your thoughts, please comment.x