文献速递脊髓损伤康复神经康复

创伤性脊髓损伤后独立步行能力的分级特异性临床预测规则

A Grade-Specific Clinical Prediction Rule for Independent Walking After Traumatic Spinal Cord Injury.
PubMed 文献 · Arch Phys Med Rehabil · 2026-08-20

本研究基于美国国家脊髓损伤统计中心2607例创伤性脊髓损伤患者数据,按AIS分级筛选最优预测变量组合并构建机器学习模型,预测伤后1年独立步行能力。结果显示不同AIS分级的最优预测变量组合显著不同,集成随机森林、人工神经网络和多任务学习的融合模型在各指标上至少不劣于任一基础模型,纳入出院变量可显著缩小AIS B+C与AIS A+D之间的预测差异。

要点

  • 不同AIS分级的最优预测变量组合显著不同,AIS A、C、D集中于腰骶段运动和感觉评分,AIS B集中于下胸段且以针刺觉为主。
  • 融合随机森林、人工神经网络和多任务学习的集成模型在所有指标上至少与任一基础模型相当,AIS B和C改善尤为明显。
  • 纳入出院时AIS和FIM评分可显著缩小AIS B+C与AIS A+D之间的预测差异。

核心结论

  • 所提出的最优变量集成模型显著改善了所有AIS分级的预测性能,尤其是AIS B和C。
  • 纳入出院变量显著缩小了不同分级之间的预测差异,该分级特异性临床预测规则对脊髓损伤康复的个体化与精准医疗具有重要意义。

研究设计与样本

  • 基于美国国家脊髓损伤统计中心数据集的分析,最终纳入2607例创伤性脊髓损伤患者。
  • 以年龄、性别、美国脊髓损伤协会(ASIA)评分和FIM评分为预测变量,预测伤后1年独立步行能力。
  • 按各AIS分级筛选最优变量组合并建立多种机器学习模型。

主要发现

  • 不同AIS分级的最优变量组合差异显著。
  • AIS A、C、D的显著预测因子集中于腰骶段运动和感觉评分,年龄也被验证为关键预测因子。
  • AIS B的关键预测因子集中于下胸段,以针刺觉为主。
  • 以最优变量组合建模后,准确率、F1分数和受试者工作特征曲线下面积值显著提高,AIS B尤为明显。
  • 增加模型复杂度可改善所有AIS分级的预测性能,但最优模型因分级而异。

关键数值与推荐

  • 融合随机森林、人工神经网络和多任务学习的集成模型在所有指标上至少与任一基础模型相当。
  • 纳入出院时ASIA和FIM评分等变量显著缩小了AIS B+C与AIS A+D之间的预测差异。

临床怎么用

  • 可按AIS分级采用不同的预测变量组合评估伤后1年独立步行预后。
  • AIS A、C、D应重点关注腰骶段运动与感觉评分及年龄,AIS B应重点关注下胸段针刺觉。
  • 出院时纳入ASIA和FIM评分有助于缩小不同分级间的预测差异。

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

OBJECTIVE: To identify significant predictors of individual American Spinal Injury Association Impairment Scale (AIS) grades and develop a clinical prediction rule with differentiated prognoses for patients with different spinal cord injury (SCI) grades. DESIGN: Age, sex, American Spinal Injury Association, and FIM scores were selected as predictive variables to predict independent walking 1 year after SCI. Optimal variable combinations for individual AIS classifications were screened, and various machine learning models were established. SETTING: Analysis of the National Spinal Cord Injury Statistical Centre dataset. PARTICIPANTS: A total of 2607 participants with traumatic SCI were included in the final analysis (N=2607). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Self-reported ability to walk both indoors and outdoors. RESULTS: Optimal variable combinations differed significantly across different AIS grades. For AIS A, C, and D, significant predictors were concentrated in motor and sensory scores at lumbosacral levels, with age also validated as a key predictor. However, for AIS B, key predictors were concentrated in the lower thoracic segments, with pinprick sensation being predominant. After modeling with optimal variable combinations, accuracy, F1 score, and area under the receiver operating characteristic curve value were significantly improved, especially for AIS B. Increasing model complexity improved prediction performance across all AIS grades, although the optimal model differed by grade. We therefore developed an ensemble model integrating random forest, artificial neural network, and multitask learning, which performed at least as well as any individual base model across all metrics. In addition, incorporating discharge variables, such as American Spinal Injury Association and FIM scores, significantly reduced the predictive discrepancy between AIS B+C and AIS A+D. CONCLUSIONS: The optimal-variable ensemble model we proposed significantly improved prediction performance for all AIS grades, especially for AIS B and C. Incorporating discharge variables significantly reduced predictive discrepancy among different grades. This grade-specific clinical prediction rule holds great significance for individualized rehabilitation and precision medicine in SCI recovery.

来源:Arch Phys Med Rehabil,查看原文。版权归原作者所有。

引用

  1. Li X, Hao F, Xiong Y, et al. A Grade-Specific Clinical Prediction Rule for Independent Walking After Traumatic Spinal Cord Injury. Arch Phys Med Rehabil. 2026. PMID: 42624188. doi:10.1016/j.apmr.2026.08.010

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。