Non-Contact Solution for Quantitative Clinical Management of MTD

这一阶段的SBIR将开发一种新的临床软件，可供言语语言病理学家(SLP)使用 通知评估和治疗肌肉紧张性发音障碍(MTD)是最常见的声音之一 临床实践中的精神障碍。MTD的特征是喉内和喉周围肌肉异常协同激活。 这会导致声音紧张-导致不适和疼痛，阻碍声音交流，并可能导致 做手术。而发声疗法可以通过减少喉部紧张和恢复肌肉来缓解这些健康风险 平衡，目前指导治疗的标准是通过手动触诊和听觉-知觉印象 -方法主观性强，评分者间可靠性差。量化衡量的最直接方法 喉部肌肉张力是对来自内窥镜的声带图像的手动分析-一种侵入性和时间- 对于MTD的常规评估和管理，临床上不可行的消耗方法。这些方法 也与最近对促进远程评估的技术的需求不相容 新冠肺炎。我们将通过开发一种新的非接触式软件工具来满足这些需求，该工具可提供准确、 在涉及发声偏移和发声的语音练习期间，MTD的有效声学测量。这 在我们之前由美国国立卫生研究院资助的研究中已经显示了一种方法，可以引发与以下相关的可测量的声学变化 MTD的存在和不存在以及声音治疗后。利用这一临床基础，我们的团队 Altec Inc.的信号处理专家将与波士顿的语音研究人员、SLP和耳鼻喉科医生合作 大学将我们基于实验室的初步软件转换为交钥匙临床工具，以支持这两个 MTD的面对面和远程评估和管理。我们将通过转换我们现有的时间来做到这一点- 将测量喉部张力的耗电手动方法输入使用全自动的临床软件 在N=450个人的现有数据库上进行测试时，使用和不使用的算法可以实现高精度 MTD(目标1)。自动算法将得到软件的补充，以指导患者遵守 MTD特定的语音练习，并为SLP提供量化的结果测量，以获得稳健的、临床上- 可行的实施(目标2)。Aim 3将在N=12名患者中通过N=4个SLP来评估我们的MTD软件原型 使用MTD来演示概念证明：1)与已验证的相比，它提供了准确的测量 手动操作；2)它检测由于发声增加或喉部按摩而引起的喉部张力变化 95%的治疗符合手动程序；3)它获得了良好的可行性评级， 患者和SLP的可接受性和感知价值。第二阶段将推进我们的第一阶段原型，真正- 时间计算、声音信号质量监测器和符合HIPPA的数据管理基础设施 用于临床护理点或远程使用。它依赖于低成本、现成的、非接触式的传感 一种实用而有效的工具，支持MTD患者的面对面和远程临床护理，从而推动 NIDCD优先考虑改善言语和语言障碍患者健康的技术。 1

This Phase I SBIR will develop a new clinical software that can be used by Speech Language Pathologists (SLPs) to inform the assessment and treatment of Muscle Tension Dysphonia (MTD), one of the most common voice disorders in clinical practice. MTD is characterized by abnormal coactivation of muscles in and around the larynx that causes vocal strain - leading to discomfort and pain that impede vocal communication and may lead to surgery. While voice therapy can mitigate these health risks by reducing laryngeal tension and restoring muscle balance, the current standard for guiding therapy is by manual palpations and auditory-perceptual impressions – highly subjective methods with poor inter-rater reliability. The most direct approach to quantitative measures of laryngeal muscle tension is manual analysis of vocal fold images from an endoscopy – an invasive and time- consuming method that is not clinically viable for routine assessment and management of MTD. These methods are also incompatible with the recent need for technology that facilitates remote assessment in the wake of COVID-19. We will address these needs by developing a new non-contact software tool that provides accurate, valid, acoustic measures of MTD during speech exercises involving the offset and onset of vocalization. This approach has been shown in our prior NIH-funded studies to elicit measurable acoustic changes associated with the presence and absence of MTD and following voice therapy. Leveraging this clinical foundation, our team of signal processing experts at Altec Inc. will partner with speech researchers, SLPs and otolaryngologists at Boston University to translate our preliminary laboratory-based software into a turn-key clinical tool to support both in-person and remote assessment and management of MTD. We will do so by translating our existing time- consuming manual approach for measuring laryngeal tension into clinical software that uses fully-automated algorithms to achieve high accuracy when tested on an existing database of N=450 individuals with and without MTD (Aim 1). The automated algorithms will be complemented by software to guide patient compliance with the MTD-specific speech exercises and provide SLPs with quantitative outcome measures for a robust, clinically- viable implementation (Aim 2). Aim 3 will evaluate our MTD software prototype by N=4 SLPs in N=12 patients with MTD to demonstrate the proof of concept that: 1) it provides accurate measures compared to validated manual procedures; 2) it detects changes in laryngeal tension due to increased vocal effort or laryngeal massage therapy with >95% agreement to manual procedures; and 3) it achieves favorable ratings of feasibility, acceptability and perceived value by patients and SLPs. Phase II will advance our Phase I prototype with real- time calculations, an acoustic signal quality monitor, and a HIPPA-compliant data management infrastructure for clinical point-of-care or remote use. It’s reliance on low-cost, readily available, non-contact sensing provides a practical and effective tool to support in-person and remote clinical care for those with MTD that advances the NIDCD priorities for technologies to improve health among individuals with speech and language disorders. 1