The development of health technology and advances in diagnostic equipment has led to the highly challenging process of medical analysis and diagnostics in multiple dimensions of large data sets. Based on large amounts of data, and data analysis, very complex problems are raised, from medical reasoning automatic knowledge extraction. This is mainly due to methodological problems inherent in multidimensional data analysis, as well as due to limitations due to the performance of computer systems. Therefore, it is often called the "curse of dimensionality". It can reduce the number of large datasets by reducing the number of analysis parameters (sizes) or by reducing the number of analysis cases. This dimensionality reduction, or feature selection method, can be used by statistical methods, primarily principal component analysis (PCA).PCA is used in many medical data analyzes. PCA performs principal component analysis and transformation of ECG datasets. Dimensionality reduction is achieved by setting a threshold or designation for the score to maintain the highest number of attributes. The PCA algorithm takes the original function as input and generates a new function. These are linear combinations of the original features selected.