Ekem Lillian, Skerrett Erica, Huchko Megan J, Ramanujam Nimmi
Department of Biomedical Engineering, Duke University, Durham, NC 27708, USA.
Center for Global Reproductive Health, Duke Global Institute, Durham, NC, USA.
Biomed Signal Process Control. 2025 Feb;100(Pt B). doi: 10.1016/j.bspc.2024.106948. Epub 2024 Sep 27.
The proportion of women dying from cervical cancer in middle- and low-income countries is over 60%, twice that of their high-income counterparts. A primary screening strategy to eliminate this burden is cervix visualization and application of 3-5% acetic acid, inducing contrast in potential lesions. Recently, machine learning tools have emerged to aid visual diagnosis. As low-cost visualization tools expand, it is important to maximize image quality at the time of the exam or of images used in algorithms.
We present the use of an object detection algorithm, the YOLOv5 model, to localize the cervix and describe blur within a multi-device image database.
We took advantage of the Fourier domain to provide pseudo-labeling of training and testing images. A YOLOv5 model was trained using Pocket Colposcope, Mobile ODT EVA, and standard of care digital colposcope images.
When tested on all devices, this model achieved a mean average precision score, sensitivity, and specificity of 0.9, 0.89, and 0.89, respectively. Mobile ODT EVA and Pocket Colposcope hold out sets yielded mAP score of 0.81 and 0.83, respectively, reflecting the generalizability of the algorithm. Compared to physician annotation, it yielded an accuracy of 0.72.
This method provides an informed quantitative, generalizable analysis of captured images that is highly concordant with expert annotation.
This quality control framework can assist in the standardization of colposcopy workflow, data acquisition, and image analysis and in doing so increase the availability of usable positive images for the development of deep learning algorithms.
中低收入国家死于宫颈癌的女性比例超过60%，是高收入国家女性的两倍。消除这一负担的主要筛查策略是宫颈可视化并应用3% - 5%的醋酸，以使潜在病变产生对比度。最近，机器学习工具已出现以辅助视觉诊断。随着低成本可视化工具的推广，在检查时或算法中使用的图像中最大化图像质量非常重要。
我们展示了使用目标检测算法YOLOv5模型来定位宫颈并描述多设备图像数据库中的模糊情况。
我们利用傅里叶域对训练和测试图像进行伪标记。使用袖珍阴道镜、移动光学检测设备EVA和标准护理数字阴道镜图像对YOLOv5模型进行训练。
在所有设备上进行测试时，该模型的平均精度得分、灵敏度和特异性分别达到0.9、0.89和0.89。移动光学检测设备EVA和袖珍阴道镜的验证集的平均精度得分分别为0.81和0.83，反映了该算法的通用性。与医生标注相比，其准确率为0.72。
该方法为捕获的图像提供了明智的定量、可推广分析，与专家标注高度一致。
这种质量控制框架可以协助阴道镜检查工作流程、数据采集和图像分析的标准化，并借此增加用于深度学习算法开发的可用阳性图像的数量。
