Detecting Ventolin-induced Cardiac abnormalities using Neural Vision Sensors
DOI:
https://doi.org/10.31224/2526Keywords:
Machine Learning, Deep Learning, Heart Rate, blood pressureAbstract
The objective of this work is to characterize how neural vision sensors, particularly neuromorphic vision sensors an be used to recognize abnormalities in heart rate of an individual at rest. We also introduce controlled studies between different subjects to design and induce a variety of rest states of cardiovascular regulation system. For this purpose, we employ intraveneous ventolin to safely induce the same. Additionally, we measure critical psychological signatures such as the blood pressure in the arteries during the systolic cycle, a time series of the heart rate of the individual in order to keep a check on the heart rate variability and any other abnormalities which may or may not occur in the process. Additionally, we also look into other measures such as entropy, the sample entropy in a specified time period, the dimensionality of the correlation, the dynamic cross-entropy, symbol Lympezel entropy and the Alhazaar entropy. Furthermore, we also apply a few tests such as the stationarity test and the collinearity test in order to ensure that the results we obtain are statistically significant for each patient. Our experimental results indicated that the pressure systems in the arteries of the patient during the systolic cycle as well as the LMS interval are mutually related in a sense, although they are governed by different properties. we empirically found this relationship to be direct in nature, with exceptions in case of large abnormalities in the collinearity tests. When the patients are present without the influence of the ventolin, we found that these systems share a lot of common properties. Furthermore, When the rigor of the refloflex feedback loop is changed with the presense of ventolin, the LMS intervals and the pressure of the blood in the arteries under the systolic cycle lose the relationship which was present previously and move toward the previous states of change. We empirically find that the systems in this state are extremely complicated and cannot be exactly governed by any previously known properties, although the pressure of the blood in the system is much simpler that it is under the ventolin-induced state.
Downloads
Downloads
Posted
License
Copyright (c) 2022 Aditya Tare

This work is licensed under a Creative Commons Attribution 4.0 International License.