使用Cerviray®与醋酸肉眼观察法（VIA）检测宫颈癌早期的结果比较
Results comparison of cervical cancer early detection using cerviray ® with VIA test.
作者信息
Harsono Ali Budi, Susiarno Hadi, Suardi Dodi, Mantilidewi Kemala Isnainiasih, Wibowo Viko Duvadilan, Hidayat Yudi Mulyana
机构信息
Department of Obstetrics and Gynecology, Faculty of Medicine, Universitas Padjadjaran-Dr. Hasan Sadikin General Hospital, Bandung, Indonesia.
Faculty of Medicine, Universitas Padjadjaran - Dr. Hasan Sadikin General Hospital, Jl. Pasteur No. 38, Bandung, West Java, 40161, Indonesia.
出版信息
BMC Res Notes. 2025 Jan 22;18(1):30. doi: 10.1186/s13104-025-07086-6.
OBJECTIVES
This study investigates the performance of artificial intelligence (AI) technology, namely Cerviray AI, compared with Cerviray expert, aiming to compare its sensitivity, specificity, positive predictive value (PPV), and area under the receiver operating characteristic curve (AUC ROC). The Visual Inspection with Acetic Acid (VIA) test is used as the gold standard.
RESULTS
The study involved 44 patients from various health centers in West Java Province. Performance of Cerviray AI, or Cerviray expert, and lastly VIA tests were compared in their ability to detect pre-cancerous cervical lesions in high-risk women of childbearing age. The current study indicated that Cerviray AI had a sensitivity of 42.9%, specificity of 100%, PPV of 100%, and ROC AUC values of 71.4%. In comparison, the evaluation of the Cerviray expert demonstrated a sensitivity of 71.4%, specificity of 97.3%, PPV of 83.3%, and ROC AUC values of 84.4%. In conclusion, the evaluation of Cerviray expert outperformed Cerviray AI in ROC AUC values.
TRIAL REGISTRATION
Clinical Trials.gov Identifier NCT06518070 Retrospectively registered.
目的
本研究调查了人工智能（AI）技术Cerviray AI与Cerviray专家的性能，旨在比较其敏感性、特异性、阳性预测值（PPV）和受试者操作特征曲线下面积（AUC ROC）。以醋酸目视检查（VIA）试验作为金标准。
结果
该研究纳入了来自西爪哇省各健康中心的44名患者。比较了Cerviray AI、Cerviray专家以及最后VIA试验在检测育龄高危女性宫颈癌前病变方面的能力。当前研究表明，Cerviray AI的敏感性为42.9%，特异性为100%，PPV为100%，ROC AUC值为71.4%。相比之下，对Cerviray专家的评估显示其敏感性为71.4%，特异性为97.3%，PPV为83.3%，ROC AUC值为84.4%。总之，在ROC AUC值方面，对Cerviray专家的评估优于Cerviray AI。
试验注册
ClinicalTrials.gov标识符NCT06518070，回顾性注册。
Zotero 插件