STUDY OF MICROSCOPIC BLOOD SAMPLES FOR EARLY DETECTION OF WHITE BLOOD CELL DISEASES

Abstract

The immune system is the natural guard of the human body against various diseases due to viruses and bacteria. White Blood Cells are centrally involved in the body's defense mechanism. WBC features extracted from blood samples can be used in diagnosing blood disorders. These features include chromatic, geometric, and textural features of WBC nucleus. Traditional manual diagnosis is error-prone, subjective, and time-consuming. Considering these challenges, this research introduces a more advanced automated algorithm in the classification of microscopic blood sample datasets. This proposed algorithm turned out to have a high accuracy in classification of 90.79%. The results obtained from the experiment showed that it had been trained even faster at higher accuracy of 94%. This algorithm, utilizing artificial intelligence techniques, significantly enhances WBC type classification efficiency and hence assists hematologists in screening several blood disorders at an early stage, thus significantly improving the limitations of manual diagnosis.