Accepted for/Published in: Journal of Medical Internet Research
Date Submitted: Oct 24, 2024
Date Accepted: Jan 22, 2025
Artificial intelligence derived blood biomarkers for ovarian cancer diagnosis: A systematic review and meta-analysis
ABSTRACT
Background:
Emerging evidence underscores the potential application of artificial intelligence (AI) in discorvering non‐invasive blood biomarkers. However, the diagnostic value of AI-derived blood biomarkers for ovarian cancer (OC) remains inconsistent.
Objective:
Therefore, this study aim to evaluate the research quality as well as the strength and validity of AI-based blood biomarkers in OC diagnosis.
Methods:
A systematic searching was performed in the Medline, Embase, IEEE, PubMed, Web of Science, and the Cochrane library. Studies examining the diagnostic accuracy of AI in OC blood biomarkers discovery were identified. The risk of bias was assessed using the Quality Assessment of Diagnostic Accuracy Studies-AI tool. Pooled sensitivity (SE), specificity (SP), and area under the curve (AUC) were estimated using a bivariate model for the diagnostic meta-analysis (PROSPERO: CRD42023481232).
Results:
A total of forty studies were ultimately included. The majority of included studies were evaluated as low risk of bias (n=31, 77·5%). Overall, the pooled SE, SP, and AUC were 85% (95%CI: 83 - 87%), 91% (90 - 92%), and 0·95 (0·92, 0·96), respectively. For the highest accuracy contingency tables, the pooled SE, SP, and AUC were 95% (90 - 97%), 97% (95 - 98%), and 0·99 (0·98, 1·00), respectively. Stratification by AI algorithms revealed higher SE and SP in studies employing machine learning study (n=36, SE=85%, SP=92%) compared to studies utilizing deep learning (n=4, SE=77%, SP=85%). Additionally, studies using serum reported significantly higher SE and SP (n=27, SE=94%, SP=96%) than those utilizing plasma (n=8, SE=83%, SP=91%). Stratification by external validation, demonstrated significantly higher SP in studies with external validation (n=7, SP=94%) compared to those without (n=33, SP=89%), while the reverse was observed for SE (74% vs. 90%). No publication bias was detected in the present meta-analysis.
Conclusions:
AI algorithms demonstrate satisfactory performance in the diagnosis of OC using blood biomarkers, and are anticipated to become an effective diagnostic modality in the future, potentially avoiding unnecessary surgeries. Future researches are warranted to incorporate external validation into AI diagnostic models, as well as to prioritize the adoption of deep-learning methodologies. Clinical Trial: NA
Citation
Per the author's request the PDF is not available.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.