Patel Palak, Ragland Katelyn, Robertson Brianna, Ragusa Gabriel, Wiley Christine, Miller Jacob, Jullens Robert, Dunham Michael, Richter Gresham
Louisiana State University Health Sciences Center, Department of Otolaryngology, Head and Neck Surgery, New Orleans, Louisiana, USA.
University of Arkansas for Medical Sciences, Department of Otolaryngology, Head and Neck Surgery, Little Rock, Arkansas, USA.
Int J Pediatr Otorhinolaryngol. 2022 May;156:111096. doi: 10.1016/j.ijporl.2022.111096. Epub 2022 Mar 18.
Design and validate a novel handheld device for the autonomous diagnosis of pediatric vascular anomalies using a convolutional neural network (CNN).
Retrospective, cross-sectional study of medical images. Computer aided design and 3D printed manufacturing.
We obtained a series of head and neck vascular anomaly images in pediatric patients from the database maintained in a large multidisciplinary vascular anomalies clinic. The database was supplemented with additional images from the internet. Four diagnostic classes were recognized in the dataset - infantile hemangioma, capillary malformation, venous malformation, and arterio-venous malformation. Our group designed and implemented a convolutional neural network to recognize the four classes of vascular anomalies as well as a fifth class consisting of none of the vascular anomalies. The system was based on the Inception-Resnet neural network using transfer learning. For deployment, we designed and built a compact, handheld device including a central processing unit, display subsystems, and control electronics. The device focuses upon and autonomously classifies pediatric vascular lesions.
The multiclass system distinguished the diagnostic categories with an overall accuracy of 84%. The inclusion of lesion metadata improved overall accuracy to 94%. Sensitivity ranged from 88% (venous malformation) to 100% (arterio-venous malformation and capillary malformation).
An easily deployed handheld device to autonomously diagnose pediatric skin lesions is feasible. Large training datasets and novel neural network architectures will be required for successful implementation.
设计并验证一种使用卷积神经网络（CNN）对小儿血管异常进行自主诊断的新型手持设备。
对医学图像进行回顾性横断面研究。计算机辅助设计和3D打印制造。
我们从一家大型多学科血管异常诊所维护的数据库中获取了一系列小儿患者的头颈部血管异常图像。该数据库还补充了来自互联网的其他图像。数据集中识别出四个诊断类别——婴儿血管瘤、毛细血管畸形、静脉畸形和动静脉畸形。我们团队设计并实施了一个卷积神经网络，以识别这四类血管异常以及第五类（即不存在任何血管异常的类别）。该系统基于使用迁移学习的Inception-Resnet神经网络。为了进行部署，我们设计并制造了一种紧凑的手持设备，包括中央处理器、显示子系统和控制电子设备。该设备专注于小儿血管病变并进行自主分类。
多类别系统区分诊断类别的总体准确率为84%。纳入病变元数据后，总体准确率提高到94%。敏感性范围从88%（静脉畸形）到100%（动静脉畸形和毛细血管畸形）。
一种易于部署的用于自主诊断小儿皮肤病变的手持设备是可行的。成功实施需要大型训练数据集和新颖的神经网络架构。
