A Novel Feature Reconstruction Method for Bone Marrow Cell Classification

Bone Marrow Cells (BMC) play a pivotal role in maintaining human health, particularly in hematopoiesis and immunity. In this paper, we propose a novel module named Reduced Redundancy Block (RRBlock) in this paper. The fundamental principle is to retain a greater quantity of high-frequency information while eliminating less significant characteristics. In addition, the RR-Block is a plug-and-play solution that can be directly incorporated into Convolutional Neural Networks (CNNs). The experimental findings illustrate that substituting the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$3 \times 3$</tex> convolutional layer in Residual Networks (ResNet) with this module leads to superior classification results. These results demonstrate an enhancement over certain up-todate models, underscoring the model's architecture efficacy.