Improvement in Accuracy of Red Blood Cells (RBC), White Blood Cells (WBC), and Platelets Detection Using Artificial Neural Network and Comparison with Hybrid Convolution Neural Network
A. Sai Abhishek, T. J. Nagalakshmi · 2025
In this paper, artificial neural network methods are used to identify red blood cells (RBCs), white blood cells (WBC), and platelets in blood samples more accurately than using hybrid convolution neural networks. Forty samples were gathered for this study in total. There are twenty samples in each category. The two categories are artificial neural networks and hybrid convolutional neural network algorithms. The dataset was imported in accordance with the research protocol, and Google Collab software was utilized. The code for hybrid convolutional neural networks and artificial neural networks was developed specifically for the investigation. The two groups were compared using the statistical analysis program SPSS. Sample sizes were calculated using an alpha of 0.05 and a pretest power of 80%. According to the simulation findings, the hybrid convolution neural network method yields an accuracy of 98.28% with a significance of 0.00, which is less than 0.05 (p < 0.05), while the artificial neural network algorithm yields an accuracy of 97.30%. Thus, alpha=0.05 is the conventional significance. Thus, it is noted that there is a meaningful difference between the two groups. In blood sample detection, the hybrid convolutional neural network outperforms the artificial neural network by a large margin when it comes to identifying red blood cells (RBCs) and white blood cells (WBCs).