Automatic identification of Chagas disease vectors using data mining and deep learning techniques

Chagas Disease (CD) is a vector-borne infectious disease transmitted from animals to humans and reversely. It is caused by the parasite Trypanosoma cruzi (abbr. as T. cruzi). CD is forcing an enormous social burden on public health and counts as one of the most significant threats to human health. CD has two phases of acute and chronic. The diagnostic of CD can be performed at both phases, but it is not entirely curable in the chronic phase. Therefore, diagnosing and treating it in the acute phase plays an essential role in overcoming the disease. There are many clinical trials dedicated to solving this problem, but they are expensive and time-consuming. Computational methods (automatic identification) to solving this problem are limited. Therefore, this work focuses on the automatic identification of CD. It proposes four different automated CD vector identification approaches that classify several vectors of kissing bugs with acceptable and promising accuracy rates. Three of the proposed frameworks are data mining-based approaches composed of preprocessing, feature extraction, feature selection, data balancing, and classification steps. The Principal component analysis (PCA) algorithm is used for feature extraction, and three different classification techniques, including Decision Tree (DT), Random Forest (RF), and Support Vector Machine (SVM), are employed in the classification step. In the fourth approach, two different convolutional neural networks (CNN) structures are applied to the preprocessed dataset. The results are optimistic and outperform previously developed systems. The accuracy of 100% is achieved using PCA + DT and PCA + RF. By achieving 100% accuracy rates for the current most common dataset of kissing bugs, we get one step closer to overcome this rising fatal disease. Such an automatic system could provide a great help for healthcare clinicians in identifying and treating CD disease in an early phase to cure it completely.