White blood cells classification using multi-fold pre-processing and optimized CNN model.

Area of Science:

• Hematology
• Medical Imaging
• Computational Biology

Background:

• White blood cells (WBCs) are crucial for immune response, and their anomalies signal diseases like leukemia.
• Previous WBC classification research often lacks accuracy due to limited feature sets and focus on fewer cell types.

Purpose of the Study:

• To develop a highly accurate and computationally efficient method for classifying WBC types from microscopic images.
• To address limitations in existing WBC classification techniques by improving feature extraction and model performance.

Main Methods:

• Employed extensive pre-processing and data augmentation techniques to generate a robust feature set.
• Utilized both conventional deep learning and transfer learning models for WBC image classification.
• Compared the proposed method against state-of-the-art machine and deep learning models.

Main Results:

• The proposed method, utilizing a pre-processed feature set with a convolutional neural network, achieved a classification accuracy of 0.99.
• Demonstrated superior performance and computational efficiency compared to existing state-of-the-art approaches.
• Successfully addressed the challenge of classifying multiple WBC types with high precision.

Conclusions:

• The novel approach significantly improves WBC classification accuracy and efficiency.
• This method offers a promising advancement for automated diagnostic tools in hematology.
• Highlights the importance of advanced feature engineering in medical image analysis.