Construction of adaptive pulse coupled neural network for abnormality detection in medical images
In this article, we propose a customized pulse coupled neural network for image segmentation to detect the different classes of skin lesions by minimizing the number of pixels as image harmonics. In addition, we have used a distribution similar to primary visual cortex for internal activation functi...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2018-05-01
|
| Series: | Applied Artificial Intelligence |
| Online Access: | http://dx.doi.org/10.1080/08839514.2018.1481818 |
| Summary: | In this article, we propose a customized pulse coupled neural network for image segmentation to detect the different classes of skin lesions by minimizing the number of pixels as image harmonics. In addition, we have used a distribution similar to primary visual cortex for internal activation function. This helps in transforming the visual behavior of the cortex. The developed neural synchrony of adaptive pulse coupled neural network helps in classifying the lesion patterns in dermoscopy images. We evaluate our proposed approach on 240 gold standard dermofit images of lesions. Our results have shown significant improvement in the accuracy and efficiency when compared with existing methods. |
|---|---|
| ISSN: | 0883-9514 1087-6545 |