World Journal of Engineering Research and Technology

M. Nirmala* and C. Karunya

ABSTRACT

The traditional method of diagnosing blood problems relies on the physical acuity of the hematologist and is time-consuming, error-prone, and has limitations. To facilitate clinical decision-making, an automated optical image processing system is needed. The objective of deep learning blood cell classification is to build a machine learning system that can correctly classify different blood cell types from digital images of blood samples. The project makes use of deep learning techniques, which have been shown to outperform traditional machine learning algorithms in a wide range of computer vision applications. The project aims to improve the quality of blood cell images and build a deep learning model for categorizing blood cells using convolutional neural networks (CNN). It has been shown that the deep learning architecture CNN excels at categorization tasks involving images. Training a deep learning model: The project's objective is to train the CNN model using a big dataset of images of tagged blood cells. During training, the back propagation technique is utilized to modify the weights and biases of the network's neurons, enabling the model to correctly classify different types of blood cells. Evaluation of the deep learning model: The project's goal is to evaluate how well the trained CNN model performed on a different dataset of images of blood cells. F1-score, recall, accuracy, and precision are components of evaluation measures. The project's goal is to develop a method for classifying blood cells and make it accessible as a web or mobile application so that medical personnel may do so more accurately and efficiently. The major objective of the project is to develop a system that will assist physicians in appropriately diagnosing and treating blood-related diseases. Deep learning is used to categorize blood cells.

[Full Text Article] [Download Certificate]