A Model for Leishmaniasis Disease Classification Based on Machine Learning and Deep Learning Algorithms

Leishmaniasis is caused by leishmania parasites and presents considerable diagnostic challenges. This work uses an approach for identifying leishmaniasis images as positive or negative. The proposed model comprises three main modules, namely, simple machine learning, convolutional neural network (CNN), and transfer learning modules. In the first module, statistical features have been extracted from microscopic images of skin lesions. The information extracted is then given as inputs to the machine learning algorithms such as Random Forest (RF), XGBoost, Decision Tree (DT), Naïve Bayes (NB), K-Nearest Neighbor (KNN) and Support Vector Machine (SVM). With the given data the SVM performs better than other machine learning algorithms. However, in the convolutional neural network module, the images are directly fed to CNN algorithm. Similarly, in the transfer learning module, the images are also directly given as inputs to the VGGNet-16 algorithm. Different performance metrics have been used to measure the performance of the algorithms used in the proposed study. The comparison shows that the performance of SVM is better than simple machine learning algorithms, CNN and VGGNet-16 used in different modules.