半定量[¹²³I]FP-CIT SPECT指标结合机器学习改善帕金森病与非典型帕金森综合征的临床鉴别
本研究纳入487例帕金森病(PD)和219例非典型帕金森综合征(APS)患者,评估半定量[¹²³I]FP-CIT SPECT指标结合机器学习能否改善退行性帕金森综合征的临床鉴别。尾状核/壳核和感觉运动/边缘系统区域间比值鉴别PD与APS的AUC最高达0.95;随机森林总体准确率64%,各类特异性均>84%,并揭示了两条诊断路径。
要点
- 尾状核/壳核与感觉运动/边缘系统比值鉴别PD与APS的AUC最高达0.95。
- 随机森林总体准确率64%,各类特异性均>84%,并揭示两条诊断路径。
- 低尾状核/后壳核比值主要聚PSP、CBD和MSA-C,高比值聚PD和MSA-P。
核心结论
- 整合解剖与功能[¹²³I]FP-CIT SPECT指标于机器学习框架可增强退行性帕金森综合征的临床鉴别,并支持其作为稳健的体内疾病生物标志物。
研究设计与样本
- 横断面研究纳入487例PD和219例APS患者,APS包括PSP 127例、MSA-P 37例、MSA-C 12例和CBD 43例,所有参与者均行[¹²³I]FP-CIT SPECT。
- 采用解剖(尾状核、壳核、腹侧纹状体)和功能(边缘系统、执行、感觉运动)分区量化纹状体摄取,计算特异性结合比、不对称指数和区域间比值。
- 用ROC曲线评估各指标鉴别性能,并训练和验证整合所有半定量指标的随机森林分类器以数据驱动识别诊断路径。
主要发现
- 尾状核/壳核和感觉运动/边缘系统区域间比值鉴别PD与APS的判别性能最强,AUC最高达0.95。
- 随机森林总体准确率64%,各类特异性均>84%,并揭示两条诊断路径。
- 低尾状核/后壳核比值主要聚PSP、CBD和MSA-C,其中对侧感觉运动摄取较低指向PSP,较高值及同侧尾状核亚区摄取进一步区分CBD与MSA-C。
- 高尾状核/后壳核比值包括PD和MSA-P,其中同侧尾状核摄取较低联合较高的感觉运动/认知比值主要鉴别两组。
关键数值与推荐
- PD 487例,APS 219例(PSP 127例、MSA-P 37例、MSA-C 12例、CBD 43例)。
- 尾状核/壳核和感觉运动/边缘系统比值鉴别PD与APS的AUC最高达0.95。
- 随机森林总体准确率64%,各类特异性>84%。
临床怎么用
- 可利用尾状核/壳核和感觉运动/边缘系统等半定量区域间比值辅助鉴别PD与APS。
- 随机森林模型的高特异性有助于减少误判,但其总体准确率64%提示需结合临床综合判断。
- 数据驱动的诊断路径(基于尾状核/后壳核比值等)可为APS亚型鉴别提供参考。
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)。
原文
PURPOSE: To evaluate whether semiquantitative striatal [¹²³I]FP-CIT SPECT-derived metrics improve clinical differentiation of degenerative parkinsonism using an integrated machine learning approach. METHODS: This cross-sectional study included 487 patients with Parkinson's disease (PD) and 219 with atypical parkinsonisms (APS), classified as progressive supranuclear palsy (PSP, n = 127), multiple system atrophy parkinsonian type (MSA-P, n = 37), multiple system atrophy cerebellar type (MSA-C, n = 12), and corticobasal degeneration (CBD, n = 43). All participants underwent a [¹²³I]FP-CIT SPECT. Striatal [¹²³I]FP-CIT uptake was quantified using anatomical (caudate, putamen, ventral striatum) and functional (limbic, executive, sensorimotor) parcellations to calculate specific binding ratios, asymmetry indices, and inter-regional ratios. Discriminative performance of each metric was evaluated using receiver operating characteristic (ROC) curves analyses. A random forest classifier integrating all semiquantitative metrics was trained and validated, enabling data-driven identification diagnostic pathways. RESULTS: Caudate-to-putamen and sensorimotor-to-limbic inter-regional ratios showed the strongest discriminative performance (AUC up to 0.95) for differentiating PD from APS. The random forest achieved a 64% overall accuracy with high per-class specificities (> 84%) and revealed two diagnostic pathways. A lower caudate-to-posterior putamen ratio, primarily grouped PSP, CBD and MSA-C, where lower contralateral sensorimotor uptake pointing to PSP while higher values and ipsilateral caudate uptake subregion further distinguishing CBD from MSA-C. A higher caudate-to-posterior putamen ratio, included PD and MSA-P, where lower ipsilateral caudate uptake together with a higher sensorimotor-to-cognitive ratio mainly differentiate both groups. CONCLUSIONS: Integrating anatomical and functional [¹²³I]FP-CIT SPECT metrics within a machine learning framework enhances the clinical differentiation of degenerative parkinsonisms and supports [¹²³I]FP-CIT SPECT as a robust in vivo disease biomarker.
引用
- Fernández-Rodríguez P, Franco-Rosado P, Diaz-Galvan P, et al. Semiquantitative [¹²³I]FP-CIT SPECT metrics combined with machine learning improve clinical differentiation of Parkinson's disease and atypical parkinsonian syndrome. Eur J Nucl Med Mol Imaging. 2026. PMID: 42393224. doi:10.1007/s00259-026-08022-x
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。