Williams Brendan, Nguyen Dan, Vidal Julie P, Saranathan Manojkumar
Centre for Integrative Neuroscience and Neurodynamics, University of Reading, Reading, United Kingdom.
School of Psychology and Clinical Language Sciences, University of Reading, Reading, United Kingdom.
Imaging Neurosci (Camb). 2024 May 8;2:1-16. doi: 10.1162/imag_a_00166. eCollection 2024 May 1.
The thalamus and its constituent nuclei are critical for a broad range of cognitive, linguistic, and sensorimotor processes, and are implicated in many neurological and neurodegenerative conditions. However, the functional involvement and specificity of thalamic nuclei in human neuroimaging work is underappreciated and not well studied due, in part, to technical challenges of accurately identifying and segmenting nuclei. This challenge is further exacerbated by a lack of common nomenclature for comparing segmentation methods. Here, we use data from healthy young (Human Connectome Project, n = 100) and older healthy adults, plus those with mild cognitive impairment and Alzheimer's disease (Alzheimer's Disease Neuroimaging Initiative, n = 540), to benchmark four state-of-the-art thalamic segmentation methods for T1 MRI (FreeSurfer, histogram-based polynomial synthesis [HIPS]-THOMAS, synthesized contrast segmentation [SCS]-convolutional neural network [CNN], and T1-THOMAS) under a single segmentation framework. Segmentations were compared using overlap and dissimilarity metrics to the Morel stereotaxic atlas, a widely accepted thalamic atlas. We also quantified each method's estimation of thalamic nuclear degeneration across Alzheimer's disease progression, and how accurately early and late mild cognitive impairment, and Alzheimer's disease could be distinguished from healthy controls. We show that the HIPS-THOMAS approach produced the most effective segmentations of individual thalamic nuclei relative to the Morel atlas, and was also most accurate in discriminating healthy controls from those with mild cognitive impairment and Alzheimer's disease using individual nucleus volumes. This latter result was different when using whole thalamus volumes, where the SCS-CNN approach was the most accurate in classifying healthy controls. This work is the first to systematically compare the efficacy of anatomical thalamic segmentation approaches under a unified nomenclature. We also provide recommendations of which segmentation method to use for studying the functional relevance of specific thalamic nuclei, based on their overlap and dissimilarity with the Morel atlas.
丘脑及其组成核团对于广泛的认知、语言和感觉运动过程至关重要，并与许多神经和神经退行性疾病有关。然而，丘脑核团在人类神经成像研究中的功能参与情况和特异性尚未得到充分认识，且研究较少，部分原因是准确识别和分割核团存在技术挑战。由于缺乏用于比较分割方法的通用命名法，这一挑战进一步加剧。在此，我们使用来自健康年轻人（人类连接组计划，n = 100）和健康老年人的数据，以及患有轻度认知障碍和阿尔茨海默病的患者（阿尔茨海默病神经成像计划，n = 540），在单一分割框架下对四种用于T1 MRI的丘脑分割方法（FreeSurfer、基于直方图的多项式合成[HIPS]-THOMAS、合成对比分割[SCS]-卷积神经网络[CNN]和T1-THOMAS）进行基准测试。使用重叠和差异度量将分割结果与莫雷尔立体定向图谱（一种广泛接受的丘脑图谱）进行比较。我们还量化了每种方法对阿尔茨海默病进展过程中丘脑核团变性的估计，以及早期和晚期轻度认知障碍以及阿尔茨海默病与健康对照的区分准确性。我们表明，相对于莫雷尔图谱，HIPS-THOMAS方法对单个丘脑核团的分割效果最为有效，并且在使用单个核团体积区分健康对照与轻度认知障碍和阿尔茨海默病患者方面也最为准确。当使用全丘脑体积时，后一个结果有所不同，此时SCS-CNN方法在对健康对照进行分类方面最为准确。这项工作首次在统一命名法下系统地比较了解剖学丘脑分割方法的功效。我们还根据它们与莫雷尔图谱的重叠和差异，提供了用于研究特定丘脑核团功能相关性的分割方法建议。
