252 SHORT-TERM ANTI-REMODELING EFFECTS OF GLIFLOZINS IN DIABETIC PATIENTS WITH REDUCED EJECTION FRACTION: AN EXPLAINABLE ARTIFICIAL INTELLIGENCE APPROACH.

Abstract Background Sodium glucose cotransporter type 2 inhibitors (SGLT2i), also called gliflozins, are playing an emerging role for the treatment of heart failure with reduced left ventricle ejection fraction (HFrEF). However, direct effects SGLT2i on and right ventricular remodeling function have not been completely clarified. We therefore aimed to assess clinical response gliflozins focusing echocardiographic evaluation identify any predictive factors a machine learning approach. Methods Based Random Forest, robust consolidated approach, we carried out single subject analysis evaluate which extent patients treated can effectively be distinguished from undergoing non-gliflozins treatments. Besides, eXplainability using Shapley values outline parameters mostly took advantage by gliflozins. Finally, experiments were designed highlight presence specific patterns undermining effectiveness. Results 5-fold cross-validation analyses showed that gliflozin was identified 0.70 ± 0.03% accuracy; most important supporting such accuracy Right Ventricle S’ Velocity (RV S’), Left End Systolic Diameter (LVESD) E/e’ ratio. Low Tricuspid Annular Plane Excursion (TAPSE) along high LVESD Diastolic Volume (EDV) likely impair Conclusions Treatment resulted in improvement several related biventricular remodeling. Several parameters, as simple may accurately predict cardiovascular treatment.