DATASET COLLECTION AND VALIDATION USING CORRELATION ANALYSIS

Main Article Content

Ruchi, Jimmy Singla

Abstract

Data collection is critical in the detection of lower spine diseases. Data collected could be in the form of text or image form. In the proposed work data collection is primarily dependent upon image dataset. MRI image dataset collected by visiting different laboratories in Punjab(Pathankot) and through online medium. Dataset verification is accomplished using correlation analysis. Collected dataset contains anomalies and must be eliminated. The pre-processing mechanism involving edge detection and contrast stretching applied and its result is also demonstrated through this proposed work. Comparison of result by applying techniques like histogram equivalence and slicing with contrast stretching yield best possible result of contrast stretching. Dataset contains 5000 samples collected from different labs out of which only 3032 shows positive correlation and were retained for demonstrating lower spine detection.

Article Details

Section
Articles