System and method for predicting heart failure hospitalization
A system for predicting heart failure hospitalization, comprises a processing device configured to process data derived from signals sensed by an implantable medical device, the signals being indicative of cardiac activity, to obtain a heart failure prediction index. The processing device is configured to process a multiplicity of variables relating to different cardiac characteristics to obtain a multiplicity of processed variables, and to combine the processed variables using a mathematical model to compute the heart failure prediction index.
1 . System for predicting heart failure hospitalization, comprising:
an implantable medical device configured to determine one or more cardiac characteristic parameters and/or a patient activity parameter,
a processing device configured to process data derived from signals sensed by the implantable medical device, said processing including identifying intervals (I) in a series of data points containing monotonically increasing and/or decreasing values related to a mean ventricular heart rate, a heart rate variability, and/or a thoracic impedance,
wherein the processing device is configured to process variables associated with the series of data points containing monotonically increasing and/or decreasing values and which relate to different cardiac characteristics and/or a patient activity to obtain a multiplicity of processed variables, and to combine said processed variables using a mathematical model to compute a heart failure prediction index.
2 . System according to claim 1 , wherein said variables relate to at least a subset of a mean ventricular heart rate parameter, a mean number of extrasystoles parameter, a mean ventricular heart rate at rest parameter, a burden of atrial arrhythmias parameter, a heart rate variability parameter, a patient activity parameter, and a thoracic impedance parameter.
3 . System according to claim 1 , wherein for processing a particular variable of said variables, the processing device is configured to process a series(s) of data points (nW) derived from the signals sensed by the implantable medical device.
4 . System according to claim 3 , wherein the processing device is configured, for the processing, to apply a moving average filter to the series(s) of data points (nW).
5 . System according to claim 1 , wherein the processing device is configured, for the processing, to also determine a slope for data points within an interval of the intervals (I), determine a number of data points within the interval which fall outside a stability range, and/or determine a number of data points within the interval which exceed a predefined threshold (TH).
6 . System according to claim 1 , wherein a series(s) of data points (nW) of a first variable relating to a mean ventricular heart rate parameter is processed to obtain a first processed variable by identifying at least one interval within intervals (I) in a series(s) of data points (nW) containing monotonically increasing values, satisfying a condition s i ≥s i+1 ≥s i+2 ≥ . . . ≥s i+L-1 , wherein s i . . . s i+L-1 are data points in the at least one interval having a length of L data points.
7 . System according to claim 1 , wherein a series(s) of data points (nW) of a second variable relating to a mean number of ventricular extrasystoles parameter is processed to obtain a second processed variable by computing a slope of data points in at least one interval within intervals (I).
8 . System according to claim 1 , wherein a series(s) of data points (nW) of a third variable relating to a mean ventricular heart rate at rest parameter is processed to obtain a third processed variable by determining a number of data points within at least one interval within intervals (I) which fall outside a stability range (SR) associated with the at least one interval.
9 . System according to claim 1 , wherein a series(s) of data points (nW) of a fourth variable relating to a burden of atrial arrhythmias parameter is processed to obtain a fourth processed variable by determining a number of data points within at least one interval of intervals (I) which exceed a predefined threshold (TH).
10 . System according to claim 1 , wherein a series(s) of data points (nW) of a fifth variable relating to a heart rate variability parameter is processed to obtain a fifth processed variable by identifying at least one interval within intervals (I) in a series(s) of data points (nW) containing monotonically decreasing values, satisfying a condition s i ≤s i+1 ≤s i+2 ≤ . . . ≤s i+L-1 , wherein s i . . . s i+L-1 are data points in the at least one interval having a length of L data points.
11 . System according to claim 1 , wherein a series(s) of data points (nW) of a sixth variable relating to a patient activity parameter is processed to obtain a sixth processed variable by computing a slope of data points in at least one interval of intervals (I).
12 . System according to claim 1 , wherein a series(s) of data points (nW) of a seventh variable relating to a thoracic impedance parameter is processed to obtain a seventh processed variable by identifying at least one interval within intervals (I) in a series(s) of data points (nW) containing monotonically decreasing values, satisfying a condition s i ≤s i+1 ≤s i+2 ≤ . . . ≤s i+L-1 , wherein s i . . . s i+L-1 are data points in the at least one interval having a length of L data points.
13 . System according to claim 1 , wherein said mathematical model is a linear combination of said processed variables, wherein within the linear combination each processed variable is multiplied by an associated coefficient.
14 . System according to claim 13 , wherein the processing device is configured to determine a Seattle Heart Failure Model (SHFM) score and to include the SHFM score in the linear combination.
15 . Method for predicting heart failure hospitalization, comprising:
determining one or more cardiac characteristic parameters and/or a patient activity parameters,
processing data derived from signals sensed by an implantable medical device, said processing including identifying intervals (I) in a series of data points containing monotonically increasing and/or decreasing values related to a mean ventricular heart rate, a heart rate variability, and/or a thoracic impedance,
wherein the step of processing data comprises processing variables associated with the series of data points containing monotonically increasing and/or decreasing values and which relate to different cardiac characteristics and/or a patient activity to obtain a multiplicity of processed variables, and combining said processed variables using a mathematical model to compute a heart failure prediction index.