Automated diagnosis of Coronary Artery Disease affected patients using LDA, PCA, ICA and Discrete Wavelet Transform

Recent works in diagnosis/decision making involve estimating severity of health episodes of patients with chronic disease, sleep issues such as polysomnography and apnea, estimation and classification of health conditions, and emotion recognition.

“Some examples of recent works in this area involve estimating the severity of health episodes of patients suffering chronic disease [44–46], sleep issues such as polysomnography and apnea [47,48], estimation and classification of health conditions [49,50], and emotion recognition [51].”

Papers working with large-scale and continuous data necessarily employed a preprocessing step because gathered sensor data is often unreliable and massive.

“Since the gathered sensor data is often unreliable and massive, the papers working with large scale and continuous data necessarily employed a preprocessing step [45].”

Reducing the amount of sensor network data is another task in feature extraction and feature selection phases, leading to an arranged vector of features as input to data mining techniques like classifiers.

“Moreover, reducing the amount of sensor network data is another task in feature extraction and feature selection phases which leads to have an arranged vector of features as an input of data mining techniques like classifier methods [45,60].”

For extra knowledge about periodic behavior of time series, medical research has given more attention to features from frequency domains such as power spectral density, low-pass/high-pass filters, spectral energy, and wavelet coefficients.

“However, for extra knowledge about the periodic behavior of time series, research in the medical field has given more attention to the features acquired from frequency-domains such as power spectral density, low-pass/high-pass filters, spectral energy, and wavelet coefficients of the signal [29,45].”

Three most popular approaches for dimension reduction in medical domain are PCA, ICA, and LDA, which statistically select the subset of most significant features.

“Three most popular approaches for dimension reduction in medical domain are PCA, ICA, and LDA [66] which are statistically selecting the subset of the most significant features [45,67].”

Most works used categorized and complex multivariate data sets with formal definitions and annotations by domain experts to evaluate quantitatively and test the performance of output decisions.

“the most of the works used categorized and complex multivariate data sets with formal definitions and annotations by domain expert [30,40,45].”