A Decision Tree Model-Based Classification of Abnormal Red Blood Cells

The diagnosis of a number of hematological illnesses depends on the ability to recognize and categorize aberrant red blood cells (RBCs). Furthermore, blood is composed of several components, namely platelets, plasma, red blood cells, white blood cells, etc. A red blood cell abnormalities can be classified by size (exocytosis), shape (poikilocytosis), color, or even the existence of bodies for inclusion. Finding these variations in red blood cell morphologies is important for one's well-being as it might reveal whether or not the blood is in good condition. To categorize aberrant red blood cell morphologies, hematologists, pathologists, and medical technicians often employed a manually using a microscope technique. This approach is challenging and prone to human mistake. To address this problem, a variety of strategies were used, including as convolutional neural networks for detection, machine learning-based classification, and image processing techniques. One machine-learning classification method is the Decision-Tree methodology. As a result, the system found and identified ten aberrant red blood cells. Hospital patients have already provided the images utilized in the system. The image was modified and categorized by the computer. The names of the aberrant red blood cells that the system detected in the picture are displayed in the results. The algorithm classified aberrant RBCs with a precision of 94.5% and recall of 93.2%, according to the results. The suggested decision tree technique provides an easy-to-use and effective method for automatically classifying aberrant red blood cells. In clinical contexts, it may help diagnose hematological problems and increase the effectiveness of blood analysis.