Clinical Validation of DystoniaNet Deep Learning Platform for Diagnosis of Isolated Dystonia

DystoniaNet 深度学习平台诊断孤立性肌张力障碍的临床验证

• 批准号: 10556419
• 负责人: Kristina Simonyan
• 金额: $ 67.63万
• 依托单位: MASSACHUSETTS EYE AND EAR INFIRMARY
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-06-01 至 2027-04-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10556419
• 关键词: Address; Affect; Agreement; Anterior; Biological Markers; Brain imaging; Characteristics; Chronic; Clinic; Clinical; Clinical Management; Complex; Corpus Callosum; Data; Development; Diagnosis; Diagnostic; Diagnostic tests; Differential Diagnosis; Disease; Disease Management; Dystonia; Early Diagnosis; Early treatment; Evaluation; Functional disorder; General Population; Healthcare; Hospitals; Human; Inferior; Knowledge; Machine Learning; Methodology; Mission; Movement; Movement Disorders; Muscle Contraction; Occupational; Outcome; Output; Patients; Pattern; Performance; Phenotype; Posture; Predictive Value; Prognosis; Prospective Studies; Provider; Psychological Stress; Public Health; Quality of life; Randomized; Research; Resources; Retrospective Studies; Smooth Muscle; Social isolation; Specificity; Structure of inferior temporal gyrus; Symptoms; Telemedicine; Testing; Time; Translations; United States National Institutes of Health; Validation; accurate diagnosis; accurate diagnostics; brain magnetic resonance imaging; comparative; data registry; deep learning; deep learning algorithm; diagnostic algorithm; diagnostic biomarker; diagnostic platform; diagnostic strategy; disability; executive function; improved; innovation; motor control; motor impairment; nervous system disorder; neural; neural network; novel; novel diagnostics; prospective; psychiatric comorbidity; research clinical testing; standard of care; thalamocortical tract; tool; translational applications; underserved area; web based interface

PROJECT SUMMARY / ABSTRACT Isolated dystonia is a movement disorder of unknown pathophysiology, which causes involuntary muscle contractions leading to abnormal, typically patterned, twisting movements and postures. A significant challenge in the clinical management of dystonia is due to the absence of a biomarker and associated ‘gold’ standard diagnostic test. Currently, the diagnosis of dystonia is guided by clinical evaluations of its symptoms, which lead to a low agreement between clinicians and a high rate of diagnostic inaccuracies. It is estimated that only 5% of patients receive an accurate diagnosis at symptom onset, and the average diagnostic delay extends up to 10.1 years. There is, therefore, an urgent unmet clinical need to establish an objective and pathophysiologi- cally relevant diagnostic test for isolated dystonia and to determine its clinical validity for accurate and fast di- agnosis of dystonia. The objective of this project is to conduct parallel retrospective and prospective studies to clinically validate the performance of DystoniaNet, a biomarker-based deep learning platform for diagnosis of isolated dystonia. Supported by our strong preliminary data, our central hypothesis is that validated perfor- mance characteristics of DystoniaNet are acceptable for its translation and implementation in the clinical set- ting as an objective, accurate, and fast platform for diagnosis of isolated dystonia. We postulate that the avail- ability of DystoniaNet platform in the clinical setting will significantly increase the accuracy of dystonia diagno- sis and significantly decrease the time to diagnosis, especially in phenotypically complex and uncertain cases. We will pursue the following two specific aims: (1) retrospective clinical validation of DystoniaNet for dystonia diagnosis, and (2) prospective randomized clinical validation of DystoniaNet. The proposed research is innova- tive because it is built on the novel conceptual and methodological concepts for clinical validation of a bi- omarker-based diagnostic platform that specifically addresses the current unmet clinical need for dystonia management. The proposed research is significant because it will advance the first objective diagnostic plat- form for dystonia diagnosis from its discovery and analytical validation to clinical use, thus filling the critical clin- ical gap in the standard of care of this disorder. Early detection and diagnosis of dystonia will enable its early therapy and improved prognosis, having an overall positive impact on healthcare and patient’s quality of life.

项目总结/摘要 孤立性肌张力障碍是一种病理生理学未知的运动障碍， 宫缩导致异常的、典型的模式化的、扭曲的运动和姿势。一个重大挑战 是由于缺乏生物标志物和相关的“金”标准 诊断测试目前，肌张力障碍的诊断是通过对其症状的临床评估来指导的， 导致临床医生之间低一致性和高的诊断不准确率。估计只有 5%的患者在症状发作时得到准确诊断，平均诊断延迟延长至 10.1年。因此，迫切需要建立一个客观的和病理生理学的临床需求， 孤立性肌张力障碍的临床相关诊断测试，并确定其临床有效性，以准确和快速诊断。 肌张力障碍的诊断。本项目的目的是进行平行的回顾性和前瞻性研究， 临床验证DystoniaNet的性能，DystoniaNet是一种基于生物标记的深度学习平台，用于诊断 孤立性肌张力障碍在我们强有力的初步数据的支持下，我们的中心假设是，经过验证的性能- DystoniaNet的mance特征对于其在临床环境中的转换和实施是可接受的- 为孤立性肌张力障碍的诊断提供了一个客观、准确、快速的平台。我们假设- DystoniaNet平台在临床环境中的能力将显著提高肌张力障碍诊断的准确性， SIS和显着减少诊断时间，特别是在表型复杂和不确定的情况下。 我们将追求以下两个具体目标：（1）对肌张力障碍的DystoniaNet进行回顾性临床验证 诊断，和（2）DystoniaNet的前瞻性随机临床验证。本研究具有创新性- 有效，因为它是建立在新的概念和方法的概念，临床验证的双， 基于omarker的诊断平台，专门解决目前未满足的肌张力障碍临床需求 管理这项研究意义重大，因为它将推动第一个客观诊断平台， 肌张力障碍诊断的形式，从它的发现和分析验证到临床使用，从而填补了关键的临床， 这种疾病的治疗标准存在着巨大的差距。早期发现和诊断肌张力障碍将使其早期 治疗和改善预后，对医疗保健和患者的生活质量产生总体积极影响。