A big data scheme for heart disease classification in map reduce using jellyfish search flow regime optimization enabled Spinalnet.

Background: The disease related to the heart is serious and can lead to death. Precise heart disease prediction is imperative for the effective treatment of cardiac patients. This can be attained by machine learning (ML) techniques using healthcare data. Several models on the basis of ML predict and identify disease in the heart, but this model cannot manage a huge database because of the deficiency of the smart model. This paper provides an optimized SpinalNet with a MapReduce model to categorize heart disease.

Objective: The objective is to design a big data approach for heart disease classification using the proposed Jellyfish Search Flow Regime Optimization (JSFRO)-based SpinalNet.

Method: The binary image conversion is applied on Electrocardiogram (ECG) images for converting the image to binary image. MapReduce model is adapted, in which the mappers execute feature extraction and the reducer performs heart disease classification. In the mapper phase, the features like statistical features, shape features and temporal features are extracted and in reducer, the SpinalNet with JSFRO is considered. Here, the training of SpinalNet is done with JSFRO, which is produced by the unification of Jellyfish Search Optimization (JSO) and Flow Regime Optimization (FRO).

Method: The JSFRO-based SpinalNet offered effectual performance with the finest accuracy of 90.8%, sensitivity of 95.2% and specificity of 93.6%.