Leukemia Cancer detection and classification

What is it about?

Leukemia detection through automated analysis of blood cell images is an emerging research area. This paper reviews image processing, machine learning and deep learning techniques applied for the detection of Leukemia Cancer. Recent advancements in deep CNNs and transfer learning for improved feature representation and classification are covered. Key contributions include preprocessing techniques, segmentation through clustering, thresholding and region growing approaches, feature extraction using morphological, statistical and transform methods, classification using SVM, neural networks and deep CNNs like AlexNet and ResNet, and transfer learning using fine-tuned models like VGGNet. Major challenges highlighted include lack of large standardized datasets, overlapping cell segmentation, subtype classification constraints, and detection accuracy versus computational complexity tradeoffs. Future work should focus on building stable public datasets, inventing hybrid segmentation methods, investigating cutting-edge deep network designs, and smoothly fusing deep learning and machine learning. In general, the review examines the architectures, constraints, and potential paths forward in automated analysis of blood cell images for fast and accurate leukemia screening.

Featured Image

Photo by National Cancer Institute on Unsplash

Photo by National Cancer Institute on Unsplash

Why is it important?

From older techniques to advanced techniques for leukemia detection are covered in paper,