Medical Image Processing Using CNN

Malaria detection takes time. The sole approach that provides confirmation is blood sample analysis. A variety of computational techniques are currently being used to speed it up. Our proposed methodology employs the usage of image enhancement filters along with the Convolutional Neural Network (CNN) idea to reduce the time complexity of malaria identification. The standard model employs several deep-learning approaches while confirming stability on the same dataset. The model employs a self-built CNN architecture with multiple filters. The Gaussian filter, along with our dataset, produced the best accuracy of 98.13%. While the model with no filter, achieves 95% accuracy.