Measurement of the correlation coefficients between extracted features from CT-scan and MRI images

Background and aims: Nowadays applying computer in image processing is rapidly increasing to resolve shortcomings of medical images. Image features modify its image contained image information. The aim of the present study was to find correlation between CT-scan and MRI images' features. Methods: In this descriptive study, it was used 1458 CT and MRI images of 6 patients (3 females and 3 males) referred to Golestan Hospital in Ahwaz, Iran. After collecting image, pre-processing and feature extract were performed. Then, the images' features were analyzed and correlation coefficients were calculated using Pearson correlation. Results: There was significant relation between most of the extracted features of the CT-scan and the MR (T1-weighted) images (P<0.05). The correlation coefficient between CT-scan images and MR (T1-weighted) images was higher than those of CT-scan images and MRI (T2-weighted). Furthermore, the correlation coefficient between CT-scan images and MRI (T1-weighted) images was higher than those between MR (T1-weighted) and MR (T2-weighted) features' images. Maximum value of the correlation coefficient (0.93) was related to the texture features and its minimum (0.004) was related to the morphological features. Conclusion: The results of this study revealed that there is a significant relationship between extracted features of CT-scan and MRI images, which leads to use a similar algorithm for classification and segmentation studies.