Validating an objective assessment of speech outcomes of children with cleft palate pre and post secondary surgery

项目摘要/摘要 对鼻音过度的感觉评估被认为是评估语音的关键组成部分 唇腭裂儿童(CLP)。然而，大多数语言病理学家(SLP)并不这样认为 接受正式的语言感知评估培训，结果，研究表明， 主观评级在本质上偏向于感知者，并表现出相当大的变异性。在这个项目中， 我们的目标是验证一种人工智能(AI)算法，该算法可以自动评估语音 美国左派演讲结果小组(ASOG)认为至关重要的维度，即 过度鼻音和言语可接受性。在NIH资助的R21(DE026252， PiS Scherer&Berisha)基于美国左翼演讲中5-7岁患有CLP的现有数据库 成果项目。人工智能工具在5-7岁患有CLP的小样本中进行了验证。 这项提议有两个目的。第一个目标是将语音样本扩展到3-5岁的儿童 中电将涵盖考虑二次腭部管理决定时的典型年龄和 收集该年龄段非唇裂儿童的对比语音样本。第二个目标是 使用接受第二次治疗的儿童的横断面和纵向样本验证人工智能工具 多中心临床试验中的腭部管理(VPI-OPS，Thomas Sitzman博士，PI)。结果将会是 提供对自动工具的验证，以客观量化和跟踪儿童的言语产出 与中电合作。

Project Summary/Abstract Perceptual assessment of hypernasality is considered a critical component when evaluating the speech of children with cleft lip and/or palate (CLP). However, most speech-language pathologists (SLPs) do not receive formal training for perceptual evaluation of speech and, as a result, research shows that the subjective ratings are inherently biased to the perceiver and exhibit considerable variability. In this project, we aim to validate an artificial intelligence (AI) algorithm that automatically evaluates speech along two dimensions deemed to be critically important by the Americleft Speech Outcomes Group (ASOG), namely hypernasality and speech acceptability. An AI algorithm was developed in an NIH-funded R21 (DE026252, PIs Scherer & Berisha) based on an existing database of 5-7 year olds with CLP from the Americleft Speech Outcomes Project. The AI tool was validated on a small prospective sample of 5-7 year olds with CLP. This proposal has two aims. The first aim will extend the speech samples to children 3-5 years of age with CLP to cover the typical ages when decisions about secondary palatal management are considered and to collect comparison speech samples from noncleft children within this age range. The second aim will validate the AI tool with a cross-sectional and longitudinal sample of children who are receiving a secondary palatal management in a multi-center clinical trial (VPI-OPS, Dr. Thomas Sitzman, PI). The results will provide validation of the automated tool to objectively quantify and track speech production in children with CLP.