Field validation of deep learning based Point-of-Care device for early detection of oral malignant and potentially malignant disorders.
Field validation of deep learning based Point-of-Care device for early detection of oral malignant and potentially malignant disorders.
复制标题
DOI：
10.1038/s41598-022-18249-x
复制
发表时间：
2022-08-22
影响因子：
4.6
通讯作者：
Kuriakose, Moni A.
中科院分区：
文献类型：
--
作者：
Birur, Praveen N.;Song, Bofan;Sunny, Sumsum P.;Keerthi, G.;Mendonca, Pramila;Mukhia, Nirza;Li, Shaobai;Patrick, Sanjana;Shubha, G.;Subhashini, A. R.;Imchen, Tsusennaro;Leivon, Shirley T.;Kolur, Trupti;Shetty, Vivek;Bhushan, Vidya R.;Vaibhavi, Daksha;Rajeev, Surya;Pednekar, Sneha;Banik, Ankita Dutta;Ramesh, Rohan Michael;Pillai, Vijay;Kathryn, O. S.;Smith, Petra Wilder;Sigamani, Alben;Suresh, Amritha;Liang, Rongguang;Kuriakose, Moni A.
在资源匮乏的环境中，早期发现口腔癌需要一种即时筛查工具，使一线卫生工作者（FHW）有能力。本研究旨在验证卷积神经网络（CNN）支持的移动健康设备与FHWs一起用于描述可疑口腔病变（恶性/潜在恶性疾病）的准确性。在印度的三级保健医院和资源匮乏的环境中测试了该装置的有效性。受试者由FHWs单独或与专家一起独立筛选。所有受试者也由口腔癌专家远程评估。该程序筛选了5025名受试者（图像：32,128），其中95% （n = 4728）进行了远程诊断。在现场专家评估的16% （n = 752）中，20% （n = 102）接受了活检。简单CNN和复杂CNN分别集成到手机和云端。与组织学相比，现场专家诊断显示出高灵敏度（94%），而远程诊断与现场专家相比显示出高准确性（灵敏度：95%；特异性：84%）。然而，与远程诊断相比，FHWs识别可疑病变的敏感性较低（60%）。结合手机，CNN （MobileNet）准确描绘病变（n = 1416，灵敏度82%），以远程诊断为参考标准，基于云的CNN （VGG19）准确度更高（灵敏度87%）。研究结果表明，自动化的移动医疗双图像系统是一种有用的分诊工具，并使FHWs能够在资源匮乏的环境中进行口腔癌筛查。
Early detection of oral cancer in low-resource settings necessitates a Point-of-Care screening tool that empowers Frontline-Health-Workers (FHW). This study was conducted to validate the accuracy of Convolutional-Neural-Network (CNN) enabled m(mobile)-Health device deployed with FHWs for delineation of suspicious oral lesions (malignant/potentially-malignant disorders). The effectiveness of the device was tested in tertiary-care hospitals and low-resource settings in India. The subjects were screened independently, either by FHWs alone or along with specialists. All the subjects were also remotely evaluated by oral cancer specialist/s. The program screened 5025 subjects (Images: 32,128) with 95% (n = 4728) having telediagnosis. Among the 16% (n = 752) assessed by onsite specialists, 20% (n = 102) underwent biopsy. Simple and complex CNN were integrated into the mobile phone and cloud respectively. The onsite specialist diagnosis showed a high sensitivity (94%), when compared to histology, while telediagnosis showed high accuracy in comparison with onsite specialists (sensitivity: 95%; specificity: 84%). FHWs, however, when compared with telediagnosis, identified suspicious lesions with less sensitivity (60%). Phone integrated, CNN (MobileNet) accurately delineated lesions (n = 1416; sensitivity: 82%) and Cloud-based CNN (VGG19) had higher accuracy (sensitivity: 87%) with tele-diagnosis as reference standard. The results of the study suggest that an automated mHealth-enabled, dual-image system is a useful triaging tool and empowers FHWs for oral cancer screening in low-resource settings.
登录
查看更多内容
影响因子：
3.7
作者：
Jeng, Ming-Jer;Sharma, Mukta;Chow, Lee
通讯作者：
Chow, Lee
影响因子：
7.2
作者：
Bhatt, Shreya;Isaac, Rita;Weller, David
通讯作者：
Weller, David
使用基于亮度比的客观自动荧光可视化方法的上皮发育不良的检测准确性。
影响因子：
14.9
作者：
Yamamoto N;Kawaguchi K;Fujihara H;Hasebe M;Kishi Y;Yasukawa M;Kumagai K;Hamada Y
通讯作者：
Hamada Y
影响因子：
3.7
作者：
Uthoff, Ross D.;Song, Bofan;Liang, Rongguang
通讯作者：
Liang, Rongguang
影响因子：
168.9
作者：
Sankaranarayanan, R;Ramadas, K;Rajan, B
通讯作者：
Rajan, B