Introduction/Background Ovarian cancer is the 6th most common malignancy with a 35% survival rate across all stages at 10 years. Ultrasound is a widely used tool for ovarian tumour diagnosis and accurate pre-operative diagnosis is essential for appropriate patient management. Artificial intelligence is an emerging field within gynaecology and has been shown to aid in the ultrasound diagnosis of ovarian cancers.
Methodology EMBASE and MEDLINE databases were searched. All type of clinical studies that used artificial intelligence in ultrasound for the diagnosis of ovarian malignancies were screened. Studies with the histopathological findings as standard were used. The diagnostic performance of each study was analysed, and the pooled diagnostic performance was assessed.
Results The initial search identified 3726 papers of which 166 were suitable for abstract screening. In the final analysis, 16 papers were included with different sample sizes and different methods used. There was a combined total of 18451 ultrasound images examined through the final included studies. The overall sensitivity was 85% (95% CI 0.84–0.85) and specificity was 93% (95% CI 0.93–0.94).
Conclusion Artificial intelligence plays an important role in aiding the ultrasound diagnosis of ovarian cancer. Further prospective work is required to further validate AI for use in clinical practice.
Disclosures nothing to declare
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.