Development of a mobile medical app for diagnosis and treatment of benign paroxysmal positional vertigo (BPPV)

良性阵发性位置性眩晕(BPPV)是导致前庭功能障碍的最常见原因， 以与特定头部位置(例如，抬头或翻滚)相关的短暂但致残的眩晕为特征 躺在床上)。BPPV可以有效地采用管内结石复位术(CRT)治疗，该治疗方法使用 将颗粒(从耳石器官移入半规管)移出的一系列头部位置 半规管。然而，许多BPPV患者获得护理的机会很少，因为 没有接受过识别和治疗培训的一线医疗保健提供者未充分使用或不当使用治疗 BPPV。BPPV的成功诊断和治疗需要准确的眼球运动识别 (眼球震颤)在测试过程中和患者的头部在测试和 治疗。 这个项目的目的是检验这样一个假设，即智能手机摄像头和惯性技术可以 用于BPPV的准确诊断和治疗。我们提出了四个具体目标：(1)开发移动医疗 用于BPPV自动诊断和指导治疗的应用程序(APP)，(2)确定 用于BPPV自动诊断的APP，(3)确定APP用于自动诊断和 以及(4)确定APP对BPPV的引导治疗的有效性。 对于目标1，提出了三项任务：(1)开发记录和分析眼睛和头部的软件 用于诊断和指导BPPV治疗的运动，(2)开发与智能手机相连的头盔 用于准确记录眼睛和头部运动的面部，以及(3)开发教程视频，以指导 临床医生正确执行Dix-Hallpike手法和治疗程序。的发展。 该应用程序将是一个迭代过程，工程师将与专业临床医生密切合作，这些医生将 在试运行前提供反馈以修改和改进软件算法、头盔设备和教程视频 该应用程序的测试。 为了确定BPPV自动诊断应用程序(AIM 2)的准确性，将对该应用程序进行试点测试 主诉运动性眩晕的退伍军人使用专家前庭的BPPV诊断 临床医生被视为黄金标准。不和谐专家与APP患者的眼动记录 将检查诊断以确定错误的来源，并修改应用程序/设备 相应地。确定APP在BPPV自动诊断和指导治疗中的可用性 (目标3)，该应用程序将使用十名天真用户(临床研究生课程的一年级学生或 医学院)和10名健康志愿者(模拟病人)。可用性将使用该系统进行衡量 可用性等级来确定产品的可用性和可学习性的百分位数等级，数据将是 在对患者进行试点测试之前，用于提高应用程序的可用性。确定应用程序的有效性 对于BPPV的引导治疗(Aim 4)，该应用程序将在10名天真的用户身上进行试点测试，这些用户将执行 退伍军人运动性眩晕的诊断和治疗程序。前庭临床专家 将使用绩效评分表对每个用户的诊断和治疗程序的充分性进行评级。 检查APP对BPPV指导治疗的有效性，从绩效评分 将被计算，适当的性能将被定义为80%的幼稚用户达到中位数 80%。此外，天真用户的单次治疗成功率将与成功进行比较 在退伍军人群体中由专家临床医生进行一次治疗的比率(86%；Akin等人，2017年) 初级保健医生的成功率(40%，Munoz等人，2007年)。

Benign Paroxysmal Positional Vertigo (BPPV) is the most common cause of vestibular dysfunction, characterized by brief but disabling vertigo associated with specific head positions (e.g., looking up or rolling over in bed). BPPV can be treated effectively with the Canalith Repositioning Treatment (CRT) which uses a series of head positions to move particles (displaced from the otolith organs into the semicircular canals) out of the semicircular canals. Nevertheless, access to care for many patients with BPPV is poor because the treatment is underused or mis-used by frontline healthcare providers who are not trained to recognize and treat BPPV. Successful diagnosis and treatment of BPPV requires accurate identification of eye movements (nystagmus) during the test procedure and accurate positioning of the patient's head during testing and treatment. The purpose of this project is to test the hypothesis that a smartphone camera and inertial technology can be used to accurately diagnose and treat BPPV. We propose four specific aims: (1) develop a mobile medical application (app) for automated diagnosis and guided treatment of BPPV, (2) determine the accuracy of the app for automated diagnosis of BPPV, (3) determine the usability of the app for automated diagnosis and guided treatment of BPPV, and (4) determine the effectiveness of the app for guided treatment of BPPV. For Aim 1, three tasks are proposed: (1) development of software to record and analyze eye and head movements to diagnose and guide treatment of BPPV, (2) development of headgear to couple the smartphone to the face for accurate eye and head movement recording, and (3) development of tutorial videos to guide the clinician in the proper execution of the Dix-Hallpike maneuver and treatment procedure. The development of the app will be an iterative process in which the engineers will work closely with the expert clinicians who will provide feedback to revise and refine the software algorithm, headgear device, and tutorial videos prior to pilot testing of the app. To determine the accuracy of the app for automated diagnosis of BPPV (Aim 2), the app will be pilot tested on Veterans with complaints of motion-provoked dizziness using the BPPV diagnoses of expert vestibular clinicians as the gold standard. Eye movement recordings of patients with discordant expert versus app diagnoses will be examined to determine the source of the errors and the app/device will be modified accordingly. To determine the usability of the app for automated diagnosis and guided treatment of BPPV (Aim 3), the app will be pilot tested using ten naïve users (first year students in clinical graduate programs or medical school) and ten healthy volunteers (mock patients). Usability will be measured using the System Usability Scale to identify the percentile rank of the product's usability and learnability, and the data will be used to improve usability of the app prior to pilot testing on patients. To determine the effectiveness of the app for guided treatment of BPPV (Aim 4), the app will be pilot tested on ten naïve users who will perform the diagnostic and treatment procedures on Veterans with motion-provoked dizziness. Expert vestibular clinicians will use a performance rubric to rate each user on the adequacy of the diagnostic and treatment procedures. To examine the effectiveness of the app for guided treatment of BPPV, the rating from the performance rubric will be calculated, and adequate performance will be defined as 80% of naïve users achieving a median score of 80%. In addition, the success rate of a single treatment by naïve users will be compared with the success rate of a single treatment by expert clinicians in the Veteran population (86%; Akin et al., 2017) and with the success rate by primary care physicians (40%, Munoz et al., 2007).