New research suggests artificial intelligence can outperform medical experts at spotting ovarian cancer in ultrasound images.
The technology can differentiate between benign and malignant lesions on ovaries, according to researchers.
The research team in Sweden developed machine learning models known as artificial neural networks and tested them on more than 17,000 ultrasound images from 3,652 patients.
READ MORE: BBC Homes Under the Hammer star Martin Roberts gives worrying update about ‘how long he has left’ after scare
READ MORE: The symptom that means you have Covid rather than flu, RSV or norovirus
The results, published in Nature Medicine, showed that AI models were better than both expert and less experienced ultrasound examiners, with an accuracy rate of 86.3%.
Expert and non-expert examiners achieved an accuracy rate of 82.6% and 77.7%, respectively. Professor Elisabeth Epstein, of the Karolinska Institute, said: “Ovarian tumours are common and are often detected by chance.
“There is a serious shortage of ultrasound experts in many parts of the world, which has raised concerns of unnecessary interventions and delayed cancer diagnoses. We therefore wanted to find out if AI can complement human experts.”
Epstein said the findings suggest AI can offer “valuable support in the diagnosis of ovarian cancer, especially in difficult-to-diagnose cases and in settings where there’s a shortage of ultrasound experts.”
The study found that the AI tool reduced the number of referrals by 63% and the misdiagnosis rate by 18%. Although further studies are needed to assess the effectiveness of neural network models, AI tools could lead to faster and more cost-effective care for patients with ovarian cancer.
The study’s joint first author Filip Christiansen, a doctoral student in the research group at Karolinska Institute, said: “With continued research and development, AI-based tools can be an integral part of tomorrow’s healthcare, relieving experts and optimising hospital resources.
“But we need to make sure that they can be adapted to different clinical environments and patient groups.”
The team says it is conducting clinical research to investigate the safety and usefulness of the AI technology. AI has also shown promise in detecting the early stages of breast cancer in subtle changes that occur in the bloodstream.
One patient’s breast cancer was successfully detected by AI despite receiving a ‘normal’ result in a routine scan.