Convolutional Neural Network Modeling for Classification of Human Red Blood Cells in Sickle Cell Disease Diagnosis
Ahmad Sabry Mohamad, Muhammad Noor Nordin, Anasuha Che Senu · 2023
Sickle cell disease, commonly known as SCA or SCD, is a genetic disorder that affects red blood cells. Instructions for performing morphometry on slides can now represent a considerable amount of effort by medical professionals, which they often struggle with. This happens because manual reviews take longer to complete. The main purpose of this article is to create a way to use the Convolutional Neural Network method developed using the MATLAB application, which will yield more accurate results in less time. Convolutional neural networks are envisioned for use as a deep learning method following the products of this work and used in this investigation to classify and identify images in order to extract features from these blood samples. A multi-layer SVM classifier trained the blood sample images to obtain the training results of this model with a scale accuracy of about 0.9583. In addition, for the test results, the imaging assesses whether the CNN neural network correctly identifies and classifies the cells as normal red blood cells or sickle cells. In addition, this model will help pathologists determine if a patient has sickle cell disease by identifying complex shapes and overlapping cells, thereby minimizing mass amount of work.