Ai blood smear analysis

Ravantheran, Herish Chaudray (2024) Ai blood smear analysis. Project Report. Melaka, Malaysia, Universiti Teknikal Malaysia Melaka. (Submitted)

Text (Full text) Ai blood smear analysis.pdf - Submitted Version Download (5MB)

Abstract

White blood cells (WBCs) play a crucial role in the blood circulatory system, serving as key defenders against infections and foreign invaders. Accurate identification and classification of WBCs are essential for diagnosing various hematological conditions. This project presents HaemLyst, an innovative AI-driven web application designed to analyze blood smears and classify different types of white blood cells. HaemLyst leverages state-of-the-art machine learning algorithms and image processing techniques to achieve high accuracy in WBC classification. The application accepts user-uploaded images of blood smears, processes these images, and provides detailed predictions, including the identification of specific WBC types. Three different predictive models are employed to ensure robust and reliable results, each offering unique insights based on the same preprocessed images. In addition to predictive accuracy, HaemLyst focuses on delivering an intuitive and modern user interface. This includes clear visualization of predictions, comparison of results across models, and the display of ground truth labels for verification. The application's user-friendly design makes it accessible to both medical professionals and researchers, facilitating the rapid and accurate analysis of blood smears. Through comprehensive testing and validation, HaemLyst has demonstrated its potential to significantly enhance the efficiency and accuracy of WBC classification in clinical settings. This thesis details the development process, underlying methodologies, and performance evaluation of HaemLyst, highlighting its contributions to the field of hematology.

Actions (login required)

View Item