Nature Medicine research shows artificial intelligence can accurately diagnose prostate cancer

International accreditation is done through a competition called PANDA

Rapid innovation

Only 10 days into the competition, an algorithm was developed to match the average pathologist

One problem with prostate cancer diagnosis today is that different pathologists reach different conclusions even on the same tissue sample, meaning treatment decisions are based on uncertain information

accurate diagnosis

KI researchers have shown in earlier studies that, compared with international experts, an artificial intelligence system can show whether a tissue sample contains cancer, estimate the amount of tumor tissue in a biopsy, and grade the severity of prostate cancer

"PANDA's results show for the first time that, in an international setting, AI systems can diagnose and grade prostate cancer as accurately as human pathologists

He continued: "Future studies should also include a wider variety of rare but challenging biopsies, as well as samples from countries with different ethnic and demographic characteristics

Not a substitute for human experts

AI-based assessment of prostate cancer biopsies has the potential to improve diagnostic quality, thereby ensuring more consistent and equitable treatment for patients at a lower cost

“The idea is not for AI to replace human experts, but to act as a safety net to prevent pathologists from missing cancer cases and to help standardize assessments

The study was carried out in collaboration with colleagues from the University Medical Center Nijmegen in the Netherlands, Google Health in the United States, and the University of Turku in Finland

Facts about prostate cancer:

• In Sweden, prostate cancer is the leading cause of cancer death in men

  
  

• In Europe and the United States, more than 20 million prostate tissue samples are examined each year

  
  

• Globally, there is a shortage of pathologists, with many developing countries having less than one pathologist per 1 million population

  
  

• The difficulty of objectively and reproducibly evaluating prostate tissue samples is a bottleneck in reducing prostate cancer mortality

  
  

Journal Reference :

1. Wouter Bulten, Kimmo Kartasalo, Po-Hsuan Cameron Chen, Peter Ström, Hans Pinckaers, Kunal Nagpal, Yuannan Cai, David F.