Applying Poisson Equation for Omitting Uneven Contrasts in Brain Tomography Images
Rahim Safa, Alireza Nikravanshalmani
Background and Purpose of the study: Several important studies have been carried out so far regarding the use of image reconstruction through Poisson equation. During the recent years, after reaching an acceptable level of quality in medical images, the lowering of the calculation complexities was gradually considered to making advancements in the algorithms within a short period of time. Improving the contrast of the medical images was also among the considerations. The present study aimed at applying Poisson equation for the elimination of uneven contrasts in brain tomography images. Study Method: CT-scan and MRI images existent in the dataset were utilized as the proposed algorithm’s input. These were transformed into gray scale images in the first stage. Poisson algorithm was recalled and used for improving the quality of the images. In the end, the proposed algorithm was compared with two methods, namely back-propagation neural network and genetic algorithm. Findings: Considering the results of comparing the proposed algorithm’s efficiency with those of genetic algorithm and back-propagation neural network, it can be clearly discerned that the proposed algorithm offered defendable results in contrast to back-propagation neural network plus its considerably faster pace of calculations. However, it was not that much successful in comparison to genetic algorithm. Conclusion: The biggest advantage of the proposed algorithm based on screened Poisson equation was its speed of action. The proposed algorithm caused an improvement in the medical images’ contrast, especially in the tomographic images taken from such soft tissues as cerebral tissues.