Massachusetts General Hospital, Harvard Medical School, Boston, USA; McGill Centre for Studies in Aging & Douglas Mental Health Institute, Faculty of Medicine, McGill University, Montreal, Canada.
Department of Psychology, Charite Berlin, Berlin, Germany.
Neuroimage. 2019 Dec;203:116108. doi: 10.1016/j.neuroimage.2019.116108. Epub 2019 Aug 28.
Disentangling age-related changes from developmental variations in hippocampal volume has proven challenging. This article presents a manual segmentation protocol for the hippocampal-to-ventricle ratio (HVR), a measure combining the assessment of hippocampal volume with surrounding ventricular volume. By providing in a single measure both a standard volumetric assessment of the hippocampus and an approximation of volume loss, based on ventricular enlargement, we believe the HVR provides a superior cross-sectional estimation of hippocampal structural integrity. In a first attempt to validate this measure, we contrasted the HVR and standard hippocampal volume in their associations with age and memory performance in two independent cohorts of healthy aging individuals. The first cohort consisted in 50 cognitively normal subjects (mean age: 66.8 years, SD: 4.96, range: 60-75 years), while the second cohort included 88 cognitively normal subjects (mean age: 65.06 years, SD: 6.42, range: 55-80 years). We showed that the manual segmentation protocol for the HVR can be implemented with high reliability. In both cohorts, the HVR showed stronger negative associations with age than standard hippocampal volume. Correlations with memory performance were also numerically superior with the HVR than standard hippocampal volume, across the two cohorts. These findings support an added benefit of using the HVR over standard hippocampal volume when examining relationships with age or memory function in aging individuals. Although further validation is required, we propose that the computation of the HVR is a promising method to improve the evaluation of hippocampal integrity from cross-sectional MR images.
从海马体积的发育变化中分辨出与年龄相关的变化一直具有挑战性。本文提出了一种手动分割海马-脑室比（HVR）的方案，该方案将海马体积的评估与周围脑室体积相结合。通过在单一指标中同时提供海马的标准体积评估和基于脑室扩大的体积损失估计，我们认为 HVR 提供了海马结构完整性的优越的横截面估计。在首次尝试验证该指标时，我们在两个独立的健康老龄化个体队列中比较了 HVR 和标准海马体积与年龄和记忆表现的相关性。第一队列由 50 名认知正常的受试者组成（平均年龄：66.8 岁，标准差：4.96，范围：60-75 岁），而第二队列包括 88 名认知正常的受试者（平均年龄：65.06 岁，标准差：6.42，范围：55-80 岁）。我们表明，HVR 的手动分割方案可以实现高度可靠性。在两个队列中，HVR 与年龄的负相关关系均强于标准海马体积。与记忆表现的相关性也以 HVR 优于标准海马体积，在两个队列中均如此。这些发现支持在检查与年龄或记忆功能的关系时，使用 HVR 而不是标准海马体积的额外益处。尽管需要进一步验证，但我们提出计算 HVR 是一种从横截面磁共振图像中改善海马完整性评估的有前途的方法。
