Multivariate Parkinson's Disease Prediction System

描述（由申请人提供）：目标是开发一个多变量系统，能够记录、量化、分析和呈现与帕金森病（PD）诊断相关的多种症状。该临床系统将通过客观量化PD的几个关键症状，包括运动特征、嗅觉、睡眠参数、声音调节和其他相关的神经学症状，来帮助早期筛查。该系统将有助于PD筛查和诊断，定义诊断标准，量化治疗方案的效果。目的是开发一个多变量系统，能够记录、量化、分析和呈现与帕金森病（PD）诊断相关的多种症状。目前尚不存在死前PD诊断的生物学标志物。诊断依赖于临床特征的存在和进展以及神经病理学的证实。临床病理研究显示假阳性和假阴性诊断率显著。一些临床特征与早期PD相关，包括运动功能、嗅觉、心率变异性和睡眠时的肌电图、声音调节和动眼肌活动。提出的多变量帕金森预测系统（MPPS）将是一种无创、易于使用的小、轻、无线联网模块系统，可量化多种帕金森症状。模块将包括运动、生理、语言和嗅觉。MPPS将有助于PD筛查和诊断，定义诊断标准，量化治疗方案的效果。该系统应允许全科医生筛查帕金森病。MPPS将由小而轻的遥测硬件模块和一个临床基站组成。一个运动模块将感知三维运动。一个生理模块将捕获标准的电生理输入。语音模块将利用无线麦克风捕捉定量语音特征。嗅觉模块将集成一个可靠的、现成的系统。临床基站将由一台小型、轻便的笔记本电脑组成，该笔记本电脑集成了无线电和临床接口软件。该基站将检测区域模块、处理数据并报告临床细节。临床系统将通过客观量化PD的几个关键症状，包括运动特征、嗅觉、睡眠参数、声音调节和其他相关的神经学症状，来帮助早期筛查。患者数据库将连接临床组，以指导诊断标准并跟踪症状进展。它将最大限度地提高患者的安全性和舒适性，通过一个小的，非侵入性的，不显眼的，不受束缚的设计，可以在诊所或家庭使用。它将通过大型的、精心设计的、多中心的临床试验来说明MPPS准确地捕捉PD症状，并区分PD和非PD受试者。具体来说，在第一阶段，我们将集成原型硬件，设计临床特征提取算法，开发软件界面，并在PD和非PD受试者中进行临床试验。我们假设我们可以准确地记录数据，提取客观临床特征，并准确地预测PD和非PD受试者之间使用多个客观临床指标作为输入。最后一期原型装置的临床效用也将由几位运动障碍专家进行评估。

DESCRIPTION (provided by applicant): The objective is to develop a multivariate system capable of recording, quantifying, analyzing, and presenting multiple symptoms related to the diagnosis of Parkinson's disease (PD). The clinical system will aid in early screening by objectively quantifying several key PD symptoms including motor features, sense of smell, sleep parameters, voice modulation, and other relevant neurological signs. The system will aid in PD screening and diagnosis, defining diagnostic criteria, and quantifying treatment protocol effects. The objective is to develop a multivariate system capable of recording, quantifying, analyzing, and presenting multiple symptoms related to the diagnosis of Parkinson's disease (PD). No biological markers currently exist for antemortem PD diagnosis. Diagnosis relies upon presence and progression of clinical features and confirmation on neuropathology. Clinicopathologic studies have shown significant false- positive and false-negative diagnosis rates. Several clinical features have been correlated with early PD including motor function, olfaction, heart rate variability and electromyography during sleep, voice modulation, and oculomotor activity. The proposed Multivariate Parkinson's Prediction System (MPPS) will be a non-invasive, easy to use system of small, lightweight, wirelessly networked modules to quantify multiple PD symptoms. Modules will include motor, physiological, speech, and olfaction. The MPPS will aid in PD screening and diagnosis, defining diagnostic criteria, and quantifying treatment protocol effects. The system should allow general practitioners to screen for PD. The MPPS will consist of small, lightweight, telemetry hardware modules and a clinical base station. A motor module will sense three-dimensional motion. A physiologic module will capture standard electrophysiology inputs. A voice module will utilize a wireless microphone to capture quantitative speech features. An olfaction module will integrate a reliable, off the shelf system. The clinical base station will consist of a small, lightweight laptop computer with an integrated radio and clinical interface software. The base station will detect area modules, process data, and report clinical details. The clinical system will aid in early screening by objectively quantifying several key PD symptoms including motor features, olfaction, sleep parameters, voice modulation, and other relevant neurological signs. A patient database will link clinical groups to guide diagnostic criteria and track symptom progression. It will maximize patient safety and comfort through a small, non-invasive, unobtrusive, untethered design that can be used in the clinic or home. It will illustrate through large, well-designed, multi-center clinical trials that the MPPS accurately captures PD symptoms and differentiates between PD and non-PD subjects. Specifically, for Phase I we will integrate prototype hardware, design algorithms for clinical feature extraction, develop a software interface, and conduct a clinical trial with PD and non-PD subjects. We hypothesize that we can accurately record data, extract objective clinical features, and accurately predict between PD and non-PD subjects using multiple objective clinical measures as inputs. The clinical utility of the final Phase I prototype device will also be evaluated by several movement disorder experts.