Analysis of Blood Cell Images to Aid in Medical Diagnosis and Disease Prediction Using AI

Abstract

Artificial intelligence is the process of imitating human intelligence. It is based on advanced techniques and designed for various applications in numerous sectors including: industry, agriculture, finance, education, particularly medicine. In this context, our study is based on the use of an AI tool for the implementation of a model, creation of a medical diagnosis support system and prediction of certain diseases by taking the case of application: Myelodysplastic Syndromes (MDS). MDS are the most common hematological diseases. These are groups of diseases that are more difficult to detect and treat. The use of conventional techniques for the analysis of this type of diseases is time- consuming; hence AI meets the need for acceleration, precision of treatment results, data analysis and prediction of anomalies. Our objective is divided into three essential phases: pre-treatment, treatment and post- treatment. For the first, we used microscopic image analysis techniques for patients with MDS and those without MDS in order to extract features. The following step involves the classification of images, the detection of pathological areas, the identification of the type of blood cells affected. The last step is the finalization step which involves predicting certain diseases and developing an action plan. Throughout the process, we used intelligent tools: DL (CNN, transfer learning), ML (KNN, SVM).