Detection of Lesions in Medical Images Using the Computational Model of Visual Attention

Detecting abnormalities in medical images is one of the basic rules in medical diagnosis, which makes it easier to understand the illness and also to detect early illness. Human , Based on the mechanism of visual attention, can easily recognize salient of image. The computational model of saliency takes the image as an input, and then produces a topographic map of how the saliency or parts of each area are attracted by a human observer. In this research, a powerful algorithm for detecting medical lesions (saliencies) is presented based on the human computational visual attention model. In this model, the rapid sensitivity of the human visual system to stimuli of intensity, orientation and color are used to reveal the saliency map of the image. The proposed model was compared with two models of saliency detection methods. The results obtained from the comparison of visual results of the images show the optimal performance of the proposed method compared to other methods

کلیدواژه ها:

Computational Model ، Visual Attention ، Saliency Map ، Medical Images