Cancer can be accurately diagnosed using urine tests with artificial intelligence

Image: The set of sensing signals collected for each patient was analyzed using ML to examine the patient for PCA. Seventy-six urine samples were measured three times, producing 912 … sequences. more

Sincerely: Korea Institute of Science and Technology (KIST)

Prostate cancer is one of the most common cancers in men. Patients are mainly prescribed based on prostate cancer * PSA, a cancer factor in the blood. However, as diagnostic accuracy is as low as 30%, a significant number of patients undergo additional invasive biopsies and thus suffer from the resulting side effects such as bleeding and pain.

* Prostate-specific antigen (PSA): A prostate-specific antigen (a cancer factor) is used as an index to examine prostate cancer.

The Korea Institute of Science and Technology (KIST) announced that Dr. from the Center for Biomaterials Research. A research team led by Kawan Hae Lee and Professor In Geb Jong of Asan Medical Center developed a technique for the diagnosis of prostate cancer within just twenty minutes. With almost 100% accuracy. The research team developed this technique by introducing an electrical AI analysis method to an electrical-signal-based ultrasensitive biosensor.

As a non-major method, a clinical trial using urine is convenient for patients and does not require invasive biopsy, leading to the diagnosis of cancer without side effects. Although ** factors are low in urine due to cancer concentrations, urine-based biosensors have been used to classify risk groups, rather than an accurate diagnosis.

** Cancer Factors: A cancer-related biological index that can objectively measure and evaluate drug response to common biological processes, disease progression, and a treatment modality.

Dr. in the Kist. Lee’s team is working towards developing a technique for diagnosing disease from urine using an electrical-signal-based ultrasensitive biosensor. The approach of using a single cancer factor associated with the diagnosis of cancer was limited in increasing the diagnostic accuracy by more than 90%. However, to overcome this limitation, the team simultaneously used a variety of cancer factors instead of using only one to increase diagnostic accuracy.

The team has developed an ultrasensitive semiconductor sensor system capable of simultaneously measuring the trace amounts of selected four cancer factors in the urine for the diagnosis of prostate cancer. They trained AI using correlation between four cancer factors, which were derived from developed sensors. The trained AI algorithm was then used to identify people with prostate cancer by analyzing complex patterns of identified signals. Diagnosis of prostate cancer using AI analysis successfully detected 76 urine samples with almost 100 percent accuracy.

“For patients who require surgery and / or treatment, cancer will be diagnosed with high accuracy using urine to reduce unnecessary biopsies and treatments, which can reduce medical costs and fatigue of medical staff. Can, ”said Professor Jeong of Asan Medical Center. “A smart biosensor has been developed in this research, which can rapidly diagnose prostate cancer with almost 100 percent accuracy through just one urine test, and use it in the accurate diagnosis of other cancers using a urine test Can be done, ”Dr. Lee said in Kist. .


This research was supported by the Korean National Research Foundation’s Midcare Researcher Grant Program, government departments (Ministry of Science and ICT, Ministry of Trade and Industry, Ministry of Health and Welfare and Food and Drug Safety), and the Korea Medical Equipment Development Fund, Science and ICT (MSIT) Ministry funded. Research results have been published in the latest issue of ACS Nano, A top international academic journal in the nano-field.

Disclaimer: AAAS and Eureclart! Are not responsible for the accuracy of news releases posted on EurekAlert! By contributing to the use of institutions or any information through the EurekAlert system.


Leave a Reply

Your email address will not be published.