Diagnosis of vertebral column disorders using a novel sprint algorithm

Abstract

Data mining in the medical field has witnessed huge popularity and their applications gained remarkable impact. In the past, disease diagnosis is considered as a tedious task because of the inaccuracy and time consumptions. This raises the necessity of a prominent disease analysis system which would save thousands of lives. To achieve this especially in data mining data mining is vital, its classification accuracy will be an efficient remedy for proper diagnosis of diseases. In the world, people are affected by various kinds of diseases, in which Lower back pain has gained attention in recent years. The main challenge is the detection of the healthy and unhealthy spine. In this paper, we proposed a SPRINT algorithm for achieving better classification results. The proposed concept is a key basis of Decision Tree which considers lumbar and sacral parameters that perform effectively on detecting unhealthy spines. The experimental result is carried out with three sets of datasets on WEKA, a perfect and popular data mining suite. The obtained results are compared with K-NN and rep-tree on the basis of several parameters. It is proven that on this comparison classification accuracy obtained by SPRINT algorithm is far better than k-nn and rep-tree thus ensuring its overall performance.