CellIdentifier: Classification of Peripheral Blood Cell Images using Deep Learning
Oyshik Ahmed Aadi, M Akash, Md. Asif Hossain, Fahim Ishraq, Abdullah Al Fahim, Dewan Ziaul Karim · 2023
Diagnosis and Identification of cells and disease infected cells are considered as very important in medical science all over the world. There are health implications which can often be identified by observing the morphological changes of cells. The traditional methods of identifying such issues can be expensive and time consuming to the extent that only certain medical centers can perform the task at hand, or take days to receive a report of. However, deep Learning with convolutional neural networks (CNNs) can take over most of this tedious process. This paper aims towards creating a custom CNN model that can quickly classify different kinds of peripheral blood cells such as different white blood cell times and platelets. Such a model can be used in blood cell counts which can be used to identify cases like leukemia. Moreover, such a method can be extended into other fields such as red blood cell detection or even infected cell detection, which includes identifying diseases from Sickle Cell Anemia to cells affected by COVID19. For this research, a total of 32768 images were used. Eventually, the custom CNN model has performed exceptionally well, achieving accuracies as high as 99.1% and 98.9% in training and validation respectively, which is significantly higher than using pretrained models such as DenseNet or NasNet. The result also proves to be higher compared to the previously completed similar tasks to the best of our knowledge. The performance of the model has also been evaluated based upon confusion matrix and classification report.