基于心音心电动态监测技术的运动性疲劳评估模型的建立

Establishment of an evaluation model for exercise-induced fatigue based on dynamic monitoring technology of heart sounds and electrocardiography

  • 摘要: 摘要:背景 疲劳是威胁健康的重要因素之一,准确评价疲劳程度并进行干预具有重要意义。目的 构建健康青年男性运动性疲劳的评估模型。方法 2023 年4 月解放军总医院第一医学中心麻醉科招募健康青年男性受试者192 例为建模训练集,另招募不同单位的青年男性受试者33 例为验证集。所有受试者利用《疲劳自觉症状调查表》进行评价,通过血压体位反射试验采集无创血压,利用可穿戴心音心电设备采集心音心电数据,并按照(平躺后扶起恢复2 分钟后-平躺前)/平躺前进行数据处理。采取单因素和多因素逻辑回归分析该人群中运动性疲劳发生的相关因素,建立对应的列线图评估模型。采用受试者工作特征(receiver operating characteristic,ROC)曲线分析模型效能。绘制校准曲线,评估模型的校准度,绘制决策曲线(decision curve analysis,DCA)评估模型的临床适用性。结果 训练集中多因素逻辑回归分析显示:△射血前期(OR=1.067,95% CI:1.011 ~ 1.139)、△第二心音能量(OR=1.019,95% CI:1.003 ~ 1.0390)、△心率(OR=1.115,95% CI:1.033 ~1.215)、(OR=0.972,95% CI:0.943 ~ 0.997)、△校正左室射血时间(OR=0.772,95% CI:0.613 ~ 0.932)均与中-重度疲劳独立关联。将以上因素纳入并建立列线图评估模型,该模型训练集的曲线下面积(area under the curve,AUC)为0.908(95% CI:0.825 ~ 0.991)。此模型在33 例受试者组成的验证集中预测中重度疲劳的AUC为0.940(95% CI:0.854 ~ 1)。校准曲线显示验证结果和模型结果有较高的一致性;决策曲线显示模型在低至中等风险阈值范围内表现较好。结论 本研究利用心音心电相关数据建立的逻辑回归模型在运动后疲劳程度的评估中表现良好,具有一定的应用价值。

     

    Abstract: Abstract: Background Fatigue is one of the crucial factors threatening people's health, and accurate assessment of fatigue levels and intervention are of great significance. Objective To construct and analyze the evaluation model for exercise-induced fatigue in healthy young male subjects.Methods In April 2023, 192 healthy young male subjects were recruited as the modeling training set by the Anesthesiology Department of the First Medical Center of Chinese PLA General Hospital. Another 33 young male subjects from different units were recruited as the validation set. All subjects were evaluated using the "Self-reported Symptoms of Fatigue Questionnaire". Non-invasive blood pressure was collected through the blood pressure postural reflex test, and heart sound and electrocardiogram data were collected using wearable heart sound and electrocardiogram devices. The data were processed according to (the value 2 minutes after being lifted up and recovered from lying flat - the value before lying flat) / the value before lying flat. Univariate and multivariate logistic regression analyses were performed to analyze the related factors for the occurrence of exercise-induced fatigue in this population, and the corresponding nomogram evaluation model was established. The receiveroperating characteristic (ROC) curve was used to analyze the efficacy of the model. The calibration curve was drawn to evaluate the calibration degree of the model, and the decision curve analysis (DCA) was drawn to evaluate the clinical applicability of the model. Results In the training set, multivariate logistic regression showed that △PEP (OR=1.067, 95% CI: 1.011-1.139), △S2E (OR= 1.019, 95% CI: 1.003-1.0390), and △HR (OR=1.115, 95% CI:1.033-1.215) were risk factors for moderate to severe fatigue; △S1E (OR = 0.972, 95% CI: 0.943-0.997) and △LVETc (OR= 0.772, 95% CI: 0.613-0.932) were independently associated with moderate to severe fatigue. Incorporating the above factors and establishing a nomogram evaluation model, the area under the curve (AUC) of the training set of this model was 0.908 (95% CI: 0.825-0.991). The AUC of the validation set was 0.940 (95% CI: 0.854-1). The calibration curve showed that there was a high degree of consistency between the verification results and the model results; The decision curve showed that the model performed well within the low to medium risk threshold range. Conclusion The logistic regression model in this study utilizing heart sound and electrocardiography data demonstrates good performance in assessing postexercise fatigue levels and has certain application value.

     

/

返回文章
返回