Infected Malaria Cell Image Classification using Convolutional Neural Networks
Malaria is a life-threatening disease caused by parasites that are transmitted to people through the bites of infected female Anopheles mosquitoes. It is preventable and curable. Malaria causes symptoms that typically include fever, tiredness, vomiting, and headaches. In severe cases, it can cause yellow skin, seizures, coma, or death.
Malaria has been one of the major causes of death in Africa, and being able to classify malaria-infected cell images using convolutional neural networks is pivotal to the quick diagnosis of Malaria.
A total of 27,558 images were gathered for the development of a deep learning model that can accurately predict the cell that is infected with malaria and the one that is not infected with malaria. The total number of images used to train the model is 22,076 and a total of 5,512 images to validate the model.
One of the important things in the data preprocessing was to resize the malaria cell images as the images were of various sizes. The images were resized to 50 by 50.
From the images of the cells infected with malaria and uninfected cell images, it can be seen that one can observe the small clot inside the cellular images for the infected cells and that uninfected cells are clean without any clot in the cellular image.
The Keras deep learning framework was used for building the convolutional neural network. Necessary libraries were imported from Keras to train the model.
The optimizer used for training the model was “Adam” and the model was trained and validated over twenty-five (25) epochs.
The model had an accuracy of 1.00 on the train set and an accuracy of 1.00 on the validation set. This shows that the model predicts and can predict if a cell is infected or uninfected with malaria
The confusion matrix and the Classification report reiterates that the model accurately predicts and can accurately predict if a cell is infected or not infected with malaria.
The whole programming was done on Kaggle kernels as it offered free GPU and there was no need to download the dataset because it is already on Kaggle. I would encourage you to use Kaggle kernels because of its free GPU or you can use Google Colab. The full link to the code can be found here and you can access the dataset on Kaggle here.
Thank you for reading the article. Please share your comments and feedback on it.