Miftari Arben, Pizzagalli Fabrizio, Bommarito Giulia, Armand Stéphane, Assal Frederic, Van De Ville Dimitri, Griffa Alessandra, Allali Gilles
Leenaards Memory Centre, Department of Clinical Neurosciences, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland.
Department of Neurosciences "Rita Levi Montalcini", University of Turin, Turin, Italy.
Neuroimage Clin. 2025 Jun 5;47:103816. doi: 10.1016/j.nicl.2025.103816.
Idiopathic normal pressure hydrocephalus (iNPH), the leading cause of reversible dementia in older adults, is characterized by urinary incontinence, gait impairment, cognitive decline, and ventriculomegaly. Neuroradiological features rely on visual assessment, including sulcal characteristics. This study applies automatic sulcal-based morphometry to characterize the sulcal phenotype in iNPH and to distinguish responders from non-responders. We analyzed the sulcal morphology in 32 iNPH patients and 41 healthy controls. Patients were categorized as responders (Resp) or non-responders (nResp) based on gait improvement following a cerebrospinal fluid tap test. A generalized linear model identified the iNPH sulcal phenotype, and a Support Vector Machine (SVM) classifier was applied to distinguish iNPH patients from controls, as well as Resp from nResp. We found that sulcal depth and widening were the key descriptors of the iNPH brain phenotype. Eight core sulci contributed the most, including compressed central, superior frontal, and frontal intraparietal bilateral sulci, and flattened left calcarine and posterior lateral fissures. An SVM classifier trained on these features effectively differentiated iNPH patients from controls (AUC: 0.933) but had limited accuracy for Resp vs. nResp (AUC: 0.556). Post-hoc analyses showed smaller superior frontal sulcal opening in nResp than in Resp. This study identified an iNPH neuroradiological phenotype based on sulcal morphology, emphasizing depth and opening as key markers. SVM classifiers trained on different sulci features performed well in differentiating healthy controls from iNPH patients but was less effective for Resp vs. nResp. Future research should investigate more advanced anatomical landmarks in iNPH.
特发性正常压力脑积水（iNPH）是老年人可逆性痴呆的主要原因，其特征为尿失禁、步态障碍、认知衰退和脑室扩大。神经放射学特征依赖于视觉评估，包括脑沟特征。本研究应用基于脑沟的自动形态测量法来表征iNPH的脑沟表型，并区分反应者和无反应者。我们分析了32例iNPH患者和41名健康对照者的脑沟形态。根据脑脊液引流试验后的步态改善情况，将患者分为反应者（Resp）或无反应者（nResp）。广义线性模型确定了iNPH的脑沟表型，并应用支持向量机（SVM）分类器来区分iNPH患者与对照者，以及反应者与无反应者。我们发现脑沟深度和增宽是iNPH脑表型的关键描述指标。八个核心脑沟贡献最大，包括受压的中央沟、额上沟以及双侧额顶内沟，还有变平的左侧距状沟和外侧后裂。基于这些特征训练的SVM分类器能有效区分iNPH患者与对照者（曲线下面积：0.933），但区分反应者与无反应者的准确率有限（曲线下面积：0.556）。事后分析显示，无反应者的额上沟开口比反应者小。本研究基于脑沟形态确定了一种iNPH神经放射学表型，强调深度和开口是关键标志。基于不同脑沟特征训练的SVM分类器在区分健康对照者与iNPH患者方面表现良好，但在区分反应者与无反应者方面效果较差。未来的研究应调查iNPH中更先进的解剖标志。
