たこつぼ心筋症の臨床リスク層別化のための画像マーカーの統合

To develop machine learning (ML) models that integrate clinical and echocardiographic markers to predict adverse cardiovascular events in patients with Takotsubo syndrome (TTS).

ML-based prediction models were developed based on two retrospective datasets comprising 450 TTS patients (252 in an internal cohort and 198 in an external cohort). Baseline clinical characteristics included demographics, medical history, triggering factors, symptoms, and vital signs. All echocardiograms were performed within 24 hours of hospital admission. Feature selection and model development were performed using the Random Forest algorithm for prediction of 3-month major adverse cardiovascular events (MACE). Model performance was evaluated by area under the curve (AUC), calibration, sensitivity and specificity. Model interpretability was evaluated using SHapley Additive exPlanations (SHAP).

The established ML models integrating key clinical and echocardiographic variables demonstrated good discrimination for 3-month MACE prediction. The MLCE-4 model, which incorporated mitral valve inflow (E wave) deceleration time (DT), systolic blood pressure (SBP), right ventricular fractional area change (RVFAC), and heart rate (HR) showed an AUC of 0.80 (0.73-0.86) in the internal cohort and an AUC of 0.76 (0.69-0.84) in the external cohort. Calibration analyses showed good agreement between predicted and observed risks internally and acceptable calibration externally. SHAP analysis highlighted the dominant contributions of key features to MACE prediction. Restricted cubic spline analyses identified clinically relevant thresholds for elevated MACE risk: DT < 129.9ms, SBP < 125mmHg, RVFAC < 34.7%, and HR > 93bpm.

ML-based clinical-echocardiographic prediction models showed prognostic value in early risk stratification in hospitalized TTS patients.