Severity analysis on vasopressin hormone secretion of smoker using Laser Doppler Flowmetry data

Smoking is a life threatening bad habit which is known to a leading cause of lung cancer and cardiovascular diseases. The secretion level of vasopressin hormone is closely associated with the health impact of smoking. This study presents an analysis on vasopressin hormone secretion using Laser Doppler Flowmetry (LDF) technique in micro-vascular network. Using the blood perfusion unit (BPU) datagram of LDF the severity of vasopressin hormone secretion is determined which can be used to examine the smoking related effects on health. 10 human subjects have been chosen to take the LDF data under smoking and nonsmoking conditions. From the acquired BPU signals, statistical analysis (mean, median, root mean square, variance, skewness) as well as spectral analysis (normalized peak, power spectral density) have been performed in order to extract specific features. 240 datasets of extracted features have been used to determine the severity level of vasopressin hormone secretion using Artificial Neural Network (ANN) where the severity level is considered as three specific secretion states Normal, Medium and High. It was observed that the secretion state of the randomly selected testing data can be accurately detected. We think the proposed method and the classifier will be helpful for clinical applications.