Irregular Breathing Classification from Multiple Patient Datasets - Prediction and Classification of Respiratory Motion - page 107

various breathing patterns. For the analysis of breathing patterns, we extracted
breathing features, e.g. vector-oriented feature [ 18 , 19 ], amplitude of breathing
cycle [ 4 , 7 ] and breathing frequency [ 13 ], etc., from the original dataset, and then
classified the whole breathing data into classes based on the extracted breathing
features. To detect irregular breathing, we introduce the reconstruction error using
neural networks as the adaptive training value for anomaly patterns in a class.
The contribution of this study is threefold: First, we propose a new approach to
detect abnormal breathing patterns with multiple patients-breathing data that better
reflect tumor motion in a way needed for radiotherapy than the spirometry.
Second, the proposed new method achieves the best irregular classification per-
formance by adopting Expectation-Maximization (EM) based on the Gaussian
Mixture model with the usable feature combination from the given feature
extraction metrics. Third, we can provide clinical merits with prediction for
irregular breathing patterns, such as to validate classification accuracy between
regular and irregular breathing patterns from ROC curve analysis, and to extract a
reliable measurement for the degree of irregularity. This study is organized as
follows. In Sect. 6.2 , the theoretical background for the irregular breathing
detection is discussed briefly. In Sect. 6.3 , the proposed irregular breathing
detection algorithm is described in detail with the feature extraction method. The
evaluation criteria of irregular classifier and the experimental results are presented
in Sects. 6.4 and 6.5 . A summary of the performance of the proposed method and
conclusion are presented in Sect. 6.6 .
6.2 Related Work
Modeling and prediction of respiratory motion are of great interest in a variety of
applications of medicine [ 26 - 30 ]. Variations of respiratory motions can be rep-
resented with statistical means of the motion [ 28 ] which can be modeled with finite
mixture models for modeling complex probability distribution functions [ 31 ]. This
study uses EM algorithm for learning the parameters of the mixture model
[ 32 , 33 ]. In addition, neural networks are widely used for breathing prediction and
for classifying various applications because of the dynamic temporal behavior with
their synaptic weights [ 2 , 3 , 34 - 36 ]. Therefore, we use neural networks to detect
irregular breathing patterns from feature vectors in given samples.
6.2.1 Expectation-Maximization Based on Gaussian
Mixture Model
A Gaussian mixture model is a model-based approach that deals with clustering
problems in attempting to optimize the fit between the data and the model. The
joint probability density of the Gaussian mixture model can be the weighted sum