AI detects ovarian cancer better than human experts in new study
The use of artificial intelligence (AI) in the field of medicine is proving to be revolutionary, especially when it comes to the early detection of ovarian cancer. A recent study conducted by researchers at Karolinska Institutet in Sweden has shown that AI models are more effective in detecting ovarian cancer than human doctors.
The study, published in Nature Medicine, tested an AI model’s ability to differentiate between benign and malignant ovarian lesions using over 17,000 ultrasound images from 3,652 patients across 20 hospitals in eight countries. The results were impressive, with the AI model achieving an accuracy rate of 86% in detecting ovarian cancer, compared to 82% for human experts and 77% for those with less expertise.
Dr. Elisabeth Epstein, a professor at Karolinska Institutet, expressed surprise at the AI model outperforming all 33 expert examiners involved in the study. She emphasized the potential impact of AI on improving ovarian cancer diagnosis, particularly in regions with limited access to experienced examiners.
Dr. Brian Slomovitz, a gynecologic oncology expert, highlighted the importance of early detection in reducing ovarian cancer deaths and noted that incorporating AI-driven support could enhance the accuracy of radiologic findings. However, he also emphasized the need for further research to demonstrate a survival benefit before widespread implementation.
Dr. Harvey Castro, an AI in healthcare expert, acknowledged the potential of AI in improving cancer diagnostics but cautioned about limitations such as bias, lack of validation for routine clinical use, and regulatory concerns. Additional research is needed to determine AI’s adaptability to real-world settings, impact on healthcare costs, and ability to handle diverse populations.
Despite the promising results of the study, the researchers emphasized that AI should only be used as a diagnostic support tool and not a replacement for human physicians. They are planning to conduct clinical studies to further validate the AI model’s performance in real clinical settings.
Overall, the integration of AI in the early detection of ovarian cancer holds great promise for improving patient outcomes and reducing unnecessary interventions. With continued research and development, AI could become a valuable tool in the fight against ovarian cancer.
