Evaluating Atlas-Driven Imaging for Clinical Applications in Cleft Palate Care

项目摘要 唇腭裂是美国最普遍的出生缺陷，每577人中就有一人患有唇腭裂。 出生尽管手术取得了进展，但据估计，25-38%的腭裂修复儿童将面临严重的并发症。 由于咽喉闭合不充分，继续有鼻音过多的讲话，需要额外的 手术虽然鼻内窥镜检查经常用于确定手术类型，但它仅提供了 对语言运动的有限认识。从当前诊断中获得的信息不足 技术来可视化肌肉的位置和他们的定量功能安排在孩子。 磁共振成像（MRI）提供了良好的软组织结构的可视化，如 肌肉，但传统上太慢，无法捕捉语音动态，其中完整的收缩可能 发生在100 ms内。 目前的提案开发了一种动态MRI成像方法，可以在 3D+时间，具有高空间和高时间分辨率，可实现每秒166帧（FPS）， 整个体积。具体而言，将开发一种动态语音成像技术， 适用于5-8岁儿童，这是腭裂二期再修复的关键年龄段 手术管理该技术集成了一个时空地图集，描述了平均运动 在健康儿童群体中寻找特定的语音样本。成像技术利用了 图谱：提高重建图像的质量，自动标记边缘进行定量 分析，并且将分析集中在与健康图谱相比来自一个对象的运动的变化上。 为了检验人口平均值地图集对性别、种族和方言的依赖性， 地图集将在三个不同的地理位置制作，有40名5-8岁的健康儿童 在每个地点，两性的数量相等，两个种族的数量相等。此外，我们将招募 45名5-8岁腭裂患者接受腭裂修复术（伴或不伴咽裂） 鼻音过浓的鼻音不足。我们将运用成像技术检查颞叶 腭咽肌功能特征，确定与 鼻音过重

PROJECT SUMMARY Cleft lip and palate is the most prevalent birth defect in the United States affecting one out of every 577 births. Despite advances in surgery, it is estimated that 25-38% of children with a repaired cleft palate will continue to have hypernasal speech due to insufficient velopharyngeal closure, requiring additional surgeries. Although nasal endoscopy is frequently used to determine the type of surgery, it provides only limited views of the movements of speech. Insufficient information is available from current diagnostic technologies to visualize the location of muscles and their quantitative functional arrangement in the child. Magnetic resonance imaging (MRI) provides excellent visualization of soft tissue structures, such as muscles, but has traditionally been too slow to capture speech dynamics where complete contractions may occur in 100 ms. The current proposal develops a dynamic MRI imaging method that can image the full vocal tract in 3D+time with high spatial and high temporal resolution, achieving 166 frames per second (FPS) for the entire volume. Specifically, a dynamic speech imaging technique will be developed that is suitable for application in children aged 5-8 years old, a critical age range for re-repair of cleft palate in secondary surgical management. The technique integrates a spatiotemporal atlas that describes the mean motions across a population of healthy children for particular speech samples. The imaging technique leverages the atlas to: improve the quality of the reconstructed images, automatically label edges for quantitative analysis, and focuses the analysis on variances in motion from one subject compared to the healthy atlas. In order to examine the dependency of the population mean atlas on sex, race, and dialect, spatiotemporal atlases will be developed at three different geographical sites, with 40 healthy children ages 5-8 years old at each site, with equal numbers of both sexes and equal numbers of two races. Additionally, we will recruit 45 cleft palate patients ages 5-8 years old following cleft palate repair with and without velopharyngeal insufficiency with resulting hypernasality. We will apply our imaging technique to examine temporal characteristics of velopharyngeal muscle function, determining movement differences related to hypernasality.