For the approximately 20,000 women diagnosed with ovarian cancer in the United States each year, artificial intelligence is emerging as a potentially lifesaving tool.
In a new study led by researchers at Sweden's Karolinska Institute, an AI model performed better than human doctors at detecting ovarian cancer.
The study, published in Nature Medicine, tested the ability of an AI model to distinguish between benign and malignant lesions of the ovary, according to a press release.
What is artificial intelligence?
The AI model was trained on more than 17,000 ultrasound images from 3,652 patients from 20 hospitals in eight countries, according to the release.
“High-quality diagnostics may become more accessible, especially in areas where access to experienced medical examiners is limited,” the study authors said. (St. Petersburg)
The AI model was 86% accurate in detecting ovarian cancer, compared to 82% for human experts and 77% for non-experts.
“We were surprised that the AI model outperformed all 33 expert judges,” study author Elizabeth Epstein, a professor in the Department of Clinical Sciences and Education at Karolinska Institutet, told FOX News Digital.
Parents trust AI for medical advice more than doctors, researchers find
Ovarian tumors are common and “often discovered incidentally,” Epstein said.
“Our study shows that AI-driven diagnostic support can significantly improve ovarian cancer diagnosis by increasing triage efficiency, reducing diagnostic errors, and addressing the shortage of specialized examiners. '' she said in a statement to FOX News Digital.
The AI model was 86% accurate in detecting ovarian cancer, compared to 82% for human experts and 77% for non-experts. (St. Petersburg)
“High-quality diagnostics could become more accessible, especially in areas where access to experienced medical examiners is limited,” Epstein continued.
“This reduces waiting times, avoids unnecessary interventions, and facilitates early detection of cancer, ultimately improving patient outcomes and ensuring more accurate diagnosis.”
AI may also reduce the need for referrals and the chance of misdiagnosis, the researchers noted.
“We were surprised that our AI model outperformed all 33 expert judges.”
Dr. Brian Slomowitz, director of gynecologic oncology at Mount Sinai Medical Center in Florida and co-chair of the Committee on Cancer Research, said screening and early detection of ovarian cancer can help reduce deaths from the disease. “The Holy Grail,” he said.
Ovarian cancer treatment is on track for FDA approval as chemotherapy alternatives emerge: 'We're making progress'
“This large retrospective study clearly shows that there may be a role for incorporating AI-driven support to better interpret ultrasound findings in patients with pelvic masses.” , Slomowitz, who was not involved in this particular study, told FOX News Digital.
“This in itself increases the accuracy of radiological findings by reducing both the false-positive rate (leading to unnecessary surgery) and the false-negative rate (missing cancer).”
Screening and early detection of ovarian cancer is the “holy grail” for reducing deaths from the disease, one oncologist said. (St. Petersburg)
The doctor pointed out that there are other ways to evaluate ovarian cancer patients besides radiology.
“Menopausal status, presence of symptoms, and blood test results are also some of the factors used to determine which patients require surgery,” Slomowitz said. “It would be great to be able to incorporate these elements into AI modeling.”
Ovarian Cancer Signs, Symptoms, Diagnosis and Treatment Options
He added that for AI technology to be widely used, it needs to not only show improved accuracy in ultrasound analysis, but also demonstrate survival benefits.
“I am optimistic that artificial intelligence will become part of the arsenal used to improve the care we provide to patients.”
“I am optimistic that artificial intelligence will become part of our arsenal to improve the care we provide to patients,” the oncologist said. (St. Petersburg)
Dr. Harvey Castro Based in Dallas, Texas The board-certified emergency medicine physician and national speaker on AI in healthcare agreed that the technology has the potential to improve cancer diagnosis, but warned that limitations remain.
Click here to sign up for our health newsletter
“AI relies on diverse, high-quality data, and bias can limit its effectiveness,” he told FOX News Digital.
“It has also not been fully validated for routine clinical use, and transparency and regulatory concerns remain unresolved.”
Castro said it will be necessary to determine how well AI adapts to real-world situations, the long-term impact on health care costs and outcomes, and whether the technology can accommodate diverse populations and different clinical settings. It reiterated that additional research is needed.
For more health articles, visit: www.foxnews.com/health
The researchers also acknowledged potential limitations of the study.
“This is not a prospective study, so we need additional data to know how it works in a real-world clinical setting,” Epstein told Fox News Digital.
Study finds ovarian cancer may be detected early with new blood test
She added that the team plans to begin clinical studies soon at Stockholm South Hospital in Sweden.
“Physicians remain responsible for diagnosing and treating patients.”
Epstein pointed out that AI should only be used as diagnostic support and not as a replacement for human doctors.
“Physicians are still responsible for diagnosing and treating patients,” she said.
CLICK HERE TO GET THE FOX NEWS APP
The Karolinska Institute research team collaborated with KTH Royal Institute of Technology. Funding was provided by the Swedish Research Council, the Swedish Cancer Society, the Stockholm Regional Council, the Radium Hemet Cancer Research Fund, and Wallenberg AI, Autonomous Systems and Software Programs (WASP), according to the release.
