In a new study, it was found that AI outperforms human experts in detecting ovarian cancer.

An AI tool may be a lifesaver for the nearly 20,000 women in the U.S. who receive an ovarian cancer diagnosis each year.

Human doctors were outperformed by AI models in detecting ovarian cancer in a new study led by researchers at Karolinska Institutet in Sweden.

An AI model's ability to distinguish between benign and malignant lesions on the ovaries was tested in a research published in Nature Medicine, according to a press release.

Over 17,000 ultrasound images from 3,652 patients across 20 hospitals in eight countries were used to train the AI model, according to the release.

The AI model outperformed both human experts and those with less expertise in detecting ovarian cancer, achieving an accuracy rate of 86% compared to 82% and 77%, respectively.

Study author Elisabeth Epstein, a professor at the Department of Clinical Science and Education at Karolinska Institutet, was surprised that AI models outperformed all 33 expert examiners, according to a study she conducted.

According to Epstein, ovarian tumors are common and are often detected by chance.

"Our research shows that AI-assisted diagnostic support can significantly improve ovarian cancer diagnosis by increasing triage speed, decreasing diagnostic mistakes, and addressing the shortage of expert evaluators," she stated to Planet Chronicle Digital.

"Experienced examiners can become more accessible in regions with limited access to high-quality diagnostics," Epstein went on.

"By reducing waiting times, avoiding unnecessary interventions, and facilitating earlier cancer detection, this will ultimately improve patient outcomes and ensure more accurate diagnoses."

The researchers found that AI could potentially decrease the need for referrals and minimize the likelihood of misdiagnosis.

"I was surprised that the AI models outperformed all 33 of the expert examiners."

The "holy grail" to reduce deaths from ovarian cancer is early detection and screening, according to Dr. Brian Slomovitz, director of gynecologic oncology and co-chair of the Cancer Research Committee at Mount Sinai Medical Center in Florida.

Slomovitz, who was not involved in the study, stated to Planet Chronicle Digital that the large retrospective trial clearly shows that AI-driven support can improve the interpretation of ultrasound findings in patients with a pelvic mass.

"The radiologic findings' accuracy is enhanced by decreasing the false positive and false negative rates, which in turn reduces the need for unnecessary surgeries and ensures the detection of cancer cases."

The doctor pointed out that there are alternative methods for evaluating ovarian cancer patients aside from radiology scans.

"Slomovitz stated that menopausal status, symptoms, and blood test results are used to determine whether patients require surgery. Incorporating these factors into AI modeling would be beneficial."

To gain widespread use, the AI technology must demonstrate both improved accuracy of ultrasound analysis and a survival benefit.

"I believe that AI will be a valuable tool in enhancing the care we provide to our patients."

A national speaker on AI in health care, Dr. Harvey Castro, based in Dallas, Texas, is a board-certified emergency medicine physician who agreed that technology has the potential to enhance cancer diagnostics, but also warned of the limitations that remain.

He informed Planet Chronicle Digital that the effectiveness of the AI is limited by bias and the need for diverse, high-quality data.

"Transparency and regulatory concerns are still unresolved, and the routine clinical use of it has not been fully validated."

Castro emphasized the need for further investigation into the effectiveness of AI in real-world scenarios, its potential effects on healthcare expenses and outcomes, and its ability to accommodate various populations and clinical settings.

For more Health articles, visit planetchronicle.net/health

The researchers also acknowledged the potential limitations of the study.

Epstein informed Planet Chronicle Digital that the study is not prospective, so additional data is required to determine its performance in a real clinical setting.

She announced that the team will soon commence clinical studies at Stockholm South Hospital in Sweden.

"It is still the physician who remains responsible for the patient’s diagnosis and treatment."

Epstein emphasized that AI should serve as a diagnostic aid rather than a substitute for human physicians.

The physician remains responsible for the patient's diagnosis and treatment, as she stated.

The Karolinska Institutet research team collaborated with the KTH Royal Institute of Technology. Funding was provided by the Swedish Research Council, the Swedish Cancer Society, the Stockholm Regional Council, the Cancer Research Funds of Radiumhemmet and the Wallenberg AI, Autonomous Systems and Software Program (WASP), as stated in the release.