FIELD: medicine.
SUBSTANCE: invention relates to a method of predicting the effectiveness of cardiac resynchronization therapy in patients with chronic heart failure. The applicant for resynchronization therapy undergoes a medical examination, from which data on BMI, LVEF_LV, and LVEF_LV are taken. Based on segmented CT and MRI images, a digital geometric model of the torso, lungs and heart is built in the form of a finite element mesh of the ventricles of the heart. The course of myocardial fibers is specified in the geometric model. Non-conductive areas of post-infarction scars and areas of fibrosis with reduced conductivity are marked. Combinations of electrode placement on the LV surface are modeled. Based on the geometric model, the first calculation of a personalized electrophysiological model of excitation of the ventricles of the heart is carried out in the mode of simulating the natural rhythm during blockade of the left bundle branch, as well as in the mode of biventricular stimulation. The resulting characteristics are substituted into the ML classifier. The first set of ML-score values on the LV surface is calculated. A regression model that predicts the second set of ML-score values for the algorithm-corrected second set of combinations of electrode placement on the LV surface is built. Bayesian optimization is used to select the stimulation point on the LV surface where the ML score has the maximum value. For this point, the characteristics of the ML classifier are calculated. The regression model is retrained at least twice until the algorithm achieves performance at which the last two iterations predicted the same point for which the ML-score value is recorded. If the ML score is greater than or equal to 0.5, it is predicted that the applicant for resynchronization therapy will be a responder. If the ML score is less than 0.5, it is predicted that the applicant for resynchronization therapy will be a non-responder.
EFFECT: prognosis of the effectiveness, feasibility and justification of CRT using optimization of the location of stimulating electrodes is provided, thus reducing the number of patients without effect from CRT.
1 cl, 4 dwg, 3 ex
Title | Year | Author | Number |
---|---|---|---|
METHOD OF PREDICTING “SUPER RESPONSE” TO CARDIAC RESYNCHRONIZATION THERAPY IN PATIENTS WITH COMPLETE LEFT BUNDLE BRANCH BLOCK AND CHRONIC HEART FAILURE OF VARIOUS ETIOLOGIES | 2023 |
|
RU2813796C1 |
METHOD FOR PREDICTION OF RESPONSE TO CARDIAC RESYNCHRONIZING THERAPY IN PATIENTS WITH CHRONIC HEART FAILURE | 2019 |
|
RU2721821C1 |
METHOD FOR PREDICTING THE EFFECTIVENESS OF PACING RESYNCHRONIZATION THERAPY IN PATIENTS WITH DILATED CARDIOMYOPATHY IN THE EARLY POSTOPERATIVE PERIOD | 2020 |
|
RU2801127C2 |
METHOD FOR PREDICTING SUPER-LIGHT ON A CARDIAC RESYNCHRONISING THERAPY IN PATIENTS WITH CHRONIC HEART FAILURE | 2019 |
|
RU2746530C2 |
METHOD FOR PREDICTATION OF SUPER RESPONSE TO CARDIAC RESYNCHRONIZATION THERAPY FOR PATIENTS WITH CHRONIC HEART FAILURE | 2016 |
|
RU2623487C1 |
METHOD FOR PREDICTION OF RESPONSE TO CARDIAC RESYNCHRONISING THERAPY IN PATIENTS WITH CHRONIC HEART FAILURE | 2018 |
|
RU2716479C1 |
METHOD FOR PREDICTION OF POSITIVE RESPONSE TO CARDIO-RESYNCHRONIZING THERAPY IN PATIENTS WITH CHRONIC HEART FAILURE OF II AND III FUNCTIONAL CLASS | 2022 |
|
RU2786467C1 |
METHOD OF SELECTING SURGICAL METHOD OF DELIVERING LEFT VENTRICULAR ELECTRODE IN TREATMENT OF SYNDROME OF CHRONIC H | 2010 |
|
RU2493779C2 |
METHOD FOR DETERMINING SEVERITY OF FIBROUS MYOCARDIAL DAMAGE USING MRT FOR SELECTING CANDIDATES FOR CARDIAC RESYNCHRONISING THERAPY | 2018 |
|
RU2685643C1 |
METHOD FOR OPTIMISING ATRIOVENTRICULAR DELAY IN PATIENTS WITH CARDIAC RESYNCHRONISING THERAPY | 2014 |
|
RU2551636C1 |
Authors
Dates
2023-11-01—Published
2022-11-23—Filed