The incidence and mortality rates of malignant tumors, such as acute leukemia, have risen significantly. Clinically, hospitals rely on cytological examination of peripheral blood and bone marrow smears to diagnose malignant tumors, with accurate blood cell counting being crucial. Existing automated methods face challenges such as low feature expression capability, poor interpretability, and redundant feature extraction when processing highdimensional microimage data. We propose a novel finegrained classification model, SCKansformer, for bone marrow blood cells, which addresses these challenges and enhances classification accuracy and efficiency. The model integrates the Kansformer Encoder, SCConv Encoder, and Global-Local Attention Encoder. The Kansformer Encoder replaces the traditional MLP layer with the KAN, improving nonlinear feature representation and interpretability. The SCConv Encoder, with its Spatial and Channel Reconstruction Units, enhances feature representation and reduces redundancy. The Global-Local Attention Encoder combines Multi-head Self-Attention with a Local Part module to capture both global and local features. We validated our model using the Bone Marrow Blood Cell FineGrained Classification Dataset (BMCD-FGCD), comprising over 10,000 samples and nearly 40 classifications, developed with a partner hospital. Comparative experiments on our private dataset, as well as the publicly available PBC and ALL-IDB datasets, demonstrate that SCKansformer outperforms both typical and advanced microcell classification methods across all datasets.

As the volume of data available for analysis grows, feature selection is becoming a vital part of ensuring accurate classification results. In classification problems, selecting a small number of features reduces computational complexity, but selecting the right features is important to maintain a high level of accuracy. In this paper, we present a feature selection method based on hybrid improved quantum-behavior particle swarm optimization, called HI-BQPSO. HI-BQPSO combines a filtering method with an improved quantum-behavior particle swarm optimization algorithm to greatly reduce the dimensionality of the data so as to overcome some of the shortcomings of BQPSO. Tests were conducted on nine gene expression datasets and 36 UCI datasets to evaluate and compare the classification accuracy of HI-BQPSO’s selected feature subsets against four other algorithms. The results, using a variety of different classifiers, show that HI-BQPSO significantly reduces the number of features required for classification while maintaining higher levels of accuracy in many cases.