Improvement of MRI images through heterogeneous interpolation techniques

This paper presents the heterogeneous implementation and evaluation of five algorithms to improve the resolution in magnetic resonance images, these are Nearest-Neighbor, Bilinear, Bicubic, Catmull-Rom and Lanczos. The experiments were carried out using the database BraTS 2017 considering T 1, T 2, T 1ce and F lair acquisition modes. To measure the efficiency and quality of the results, the metrics PSNR, MAE, SSIM, as well as the Runtime were used. The results show that for all modalities the Bicubic algorithm has a PSNR > 35, MAE <; 1.089, SSIM > 0.958 and RT <; 5.205 seconds per whole database. By other hand, Catmull-Rom achieves favorable results compared with Bicubic algorithm, maintaining the balance between quality and a less processing time than mentioned method.