定量海马体积与结构化评分量表在预测阿尔茨海默病诊断中的比较
回顾性分析2018-2021年328例接受痴呆方案MR检查的成人患者(阿尔茨海默病118例、轻度认知障碍172例、主观认知下降38例),比较自动 volumetric 软件得到的标准化海马总体积与手动内侧颞叶萎缩、内嗅皮层萎缩评分的诊断效能。内嗅皮层萎缩评分在区分阿尔茨海默病与主观认知下降时特异度较高(cutoff 2,87%),但敏感度较低(69%);作者认为内嗅皮层萎缩和内侧颞叶萎缩评分可能是有用的辅助手段。
要点
- 内嗅皮层萎缩评分cutoff 2区分阿尔茨海默病与主观认知下降:特异度87%、敏感度69%。
- 标准化海马总体积区分阿尔茨海默病与主观认知下降:敏感度84%、特异度74%。
- 内侧颞叶萎缩评分区分阿尔茨海默病与主观认知下降:敏感度84%、特异度66%。
核心结论
- 内嗅皮层萎缩和内侧颞叶萎缩评分可能是区分阿尔茨海默病与主观认知下降的有用辅助手段,且相比自动 volumetric 软件成本和实施难度更低。
研究设计与样本
- 回顾性研究,纳入2018-2021年接受痴呆方案MR检查的成人患者328例(阿尔茨海默病118例、轻度认知障碍172例、主观认知下降38例),诊断依据ICD-10,采用多变量logistic回归和ROC曲线分析比较自动标准化海马总体积与手动内侧颞叶萎缩、内嗅皮层萎缩评分的诊断效能。
诊断标准与关键数值
- 区分阿尔茨海默病与主观认知下降:内嗅皮层萎缩cutoff 2敏感度69%、特异度87%;标准化海马总体积敏感度84%、特异度74%;内侧颞叶萎缩敏感度84%、特异度66%。
- 区分阿尔茨海默病与轻度认知障碍:内嗅皮层萎缩cutoff 3敏感度69%、特异度66%;标准化海马总体积敏感度71%、特异度67%;内侧颞叶萎缩敏感度84%、特异度54%。
- 阿尔茨海默病患者标准化海马总体积中位数0.35%,低于主观认知下降和轻度认知障碍患者(P < .001)。
- 阿尔茨海默病患者内侧颞叶萎缩评分中位数3、内嗅皮层萎缩评分中位数2,均高于主观认知下降和轻度认知障碍患者(P < .001)。
临床应用与阅片要点
- 作者认为内嗅皮层萎缩和内侧颞叶萎缩评分可能是区分阿尔茨海默病与主观认知下降的有用辅助手段。
- 内嗅皮层萎缩评分cutoff 2区分阿尔茨海默病与主观认知下降时特异度87%,高于标准化海马总体积(74%)和内侧颞叶萎缩(66%),但敏感度69%低于后两者(均84%)。
- 内嗅皮层萎缩评分cutoff 3区分阿尔茨海默病与轻度认知障碍时特异度66%,与标准化海马总体积(67%)相似、高于内侧颞叶萎缩(54%);敏感度69%与标准化海马总体积(71%)相似、低于内侧颞叶萎缩(84%)。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
BACKGROUND AND PURPOSE: Brain imaging plays an important role in investigating patients with cognitive decline and ruling out secondary causes of dementia. This study compares the diagnostic value of quantitative hippocampal volumes derived from automated volumetric software and structured scoring scales in differentiating Alzheimer disease, mild cognitive impairment, and subjective cognitive decline.
MATERIALS AND METHODS: Retrospectively, we reviewed images and medical records of adult patients who underwent MR imaging with a dementia protocol (2018-2021). Patients with postscanning diagnoses of Alzheimer disease, mild cognitive impairment, and subjective cognitive decline based on the International Statistical Classification of Diseases and Related Health Problems, 10th revision, were included. Diagnostic performances of automated normalized total hippocampal volume and structured manually assigned medial temporal atrophy and entorhinal cortical atrophy scores were assessed using multivariate logistic regression and receiver operating characteristic curve analysis.
RESULTS: We evaluated 328 patients (Alzheimer disease, n = 118; mild cognitive impairment, n = 172; subjective cognitive decline, n = 38). Patients with Alzheimer disease had lower normalized total hippocampal volume (median, 0.35%), higher medial temporal atrophy (median, 3), and higher entorhinal cortical atrophy (median, 2) scores than those with subjective cognitive decline (P < .001) and mild cognitive impairment (P < .001). For discriminating Alzheimer disease from subjective cognitive decline, an entorhinal cortical atrophy cutoff value of 2 had a higher specificity (87%) compared with normalized total hippocampal volume (74%) and medial temporal atrophy (66%), but a lower sensitivity (69%) than normalized total hippocampal volume (84%) and medial temporal atrophy (84%). In discriminating Alzheimer disease from mild cognitive impairment, an entorhinal cortical atrophy cutoff value of 3 had a specificity (66%), similar to that of normalized total hippocampal volume (67%) but higher than medial temporal atrophy (54%), and its sensitivity (69%) was also similar to that of normalized total hippocampal volume (71%) but lower than that of medial temporal atrophy (84%).
CONCLUSIONS: Entorhinal cortical atrophy and medial temporal atrophy may be useful adjuncts in discriminating Alzheimer disease from subjective cognitive decline, with reduced cost and implementation challenges compared with automated volumetric software.
引用
- Essien M, Lah J, Weinberg BD, et al. Comparison of Quantitative Hippocampal Volumes and Structured Scoring Scales in Predicting Alzheimer Disease Diagnosis. AJNR Am J Neuroradiol. 2023. PMID: 38050003. doi:10.3174/ajnr.A8049
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。