Classification of Pattern using Spline Function and Statistical Data Mining: Review Paper

Abstract

ABSTRACT
Many people are affected by lung diseases. Respiratory diseases are curable in early detection. Spirometry is valuable for diagnosing specific lung disorders as well as detecting lung disease at an early stage. Spirometry is most widely used pulmonary function test that is essential to measure the volume of air exhaled from the fully inflated lung as a function of time. The output of spirometry is in
the form of graphs i.e., flow-volume loop and volume-time curve. The various parameters taken from these two graphs are modeled using spline function. In this paper, three types of pulmonary diseases such as obstructive, restrictive and mixed lung disorders are classified using statistical data mining approach. The obstructive lung diseases include asthma, chronic bronchitis, chronic obstructive
pulmonary disease (COPD) and emphysema while the restrictive lung diseases include asbestosis, sarcoidosis and pulmonary fibrosis. This classification helps a physician in diagnosis process of various diseases. After modeling, the classification is performed in which a statistical analysis is used to determine the influence of health conditions on the model parameter values. This is done using
ANOVA technique. This approach is used to increase the efficiency of classification.

Keywords: Respiratory diseases, spirometry, obstructive, restrictive and mixed lung disorders, modeling, statistical data mining