Medical image segmentation is critical for disease diagnosis, treatment planning, and prognosis assessment, yet the complexity and diversity of medical images pose significant challenges to accurate segmentation. While Convolutional Neural Networks capture local features and Vision Transformers excel in the global context, both struggle with efficient long-range dependency modeling. Inspired by Mamba's State Space Modeling efficiency, we propose ÆMMamba, a novel multi-scale feature extraction framework built on the Mamba backbone network. ÆMMamba integrates several innovative modules: the Efficient Fusion Bridge (EFB) module, which employs a bidirectional state-space model and attention mechanisms to fuse multi-scale features; the Edge-Aware Module (EAM), which enhances low-level edge representation using Sobel-based edge extraction; and the Boundary Sensitive Decoder (BSD), which leverages inverse attention and residual convolutional layers to handle cross-level complex boundaries. ÆMMamba achieves state-of-the-art performance across 8 medical segmentation datasets. On polyp segmentation datasets (Kvasir, ClinicDB, ColonDB, EndoScene, ETIS), it records the highest mDice and mIoU scores, outperforming methods like MADGNet and Swin-UMamba, with a standout mDice of 72.22 on ETIS, the most challenging dataset in this domain. For lung and breast segmentation, ÆMMamba surpasses competitors such as H2Former and SwinUnet, achieving Dice scores of 84.24 on BUSI and 79.83 on COVID-19 Lung. And on the LGG brain MRI dataset, ÆMMamba attains an mDice of 87.25 and an mIoU of 79.31, outperforming all compared methods.

The inherent characteristics of high-dimensional feature redundancy of biomedical data lead to the "curse of dimensionality" in bioinformatics, which brings new challenges to feature selection problems. Recently, the two-stage approach combining the filter and wrapper methods has become popular for feature selection tasks. However, these two-stage or previous one-stage algorithms suffer from blindness in the setting of thresholds, and the search methods tend to fall into local optimum solutions. To this end, we propose a three-stage feature selection method that combines ReliefF and Particle swarm optimization as a specific case, called TS-RePSO, including the filter stage, grouping stage, and wrapper stage. Specifically, in the filter stage, ReliefF is utilized to compute the weights of the features and sort them in descending order. In the grouping stage, the ranked features are grouped based on the density equalization strategy so that the weight of groups in all groups is equal. In the wrapper stage, the proposed grouping PSO is employed to search for the grouped features and select them according to the in-group and out-group evaluation strategies. Extensive experiments are conducted on 5 benchmark datasets and 6 real-world datasets, and experiment results show that the proposed method achieves the best performance.

High dimensional and small samples characterize gene expression data and contain a large number of genes unrelated to disease. Feature selection improves the efficiency of disease diagnosis by selecting a small number of important genes. Unfortunately, existing algorithms do not consider the correlation between features, and search algorithms tend to fall into the local optimal solution in the feature search process. To this end, this paper proposes a feature filter and group evolution hybrid feature selection algorithm (FG-HFS) for high-dimensional gene expression data. Unlike existing algorithms, we propose using spectral clustering to group redundant features into a group. Then, we propose a redundant feature filter algorithm. According to the principle of approximate Markov blanket, grouped feature groups are filtered to delete these redundant features. Among them, filtered features are evenly divided by density according to the feature exponential strategy. Most importantly, we propose using the group evolution multi-objective genetic algorithm to search the filtered feature subsets and evaluate the candidate feature subsets according to the in-group and out-group so as to select the feature subsets with the highest accuracy and the least number. Experimental results show that the average accuracy (ACC) and Matthews correlation coefficient (MCC) indexes of the selected feature subsets (FSs) by the FG-HFS algorithm on 5 gene expression datasets are 92.76% and 88.76%, respectively, which are significantly better than the existing algorithms. In addition, the FSs and ACC/FSs indexes of the FG-HFS algorithm are also better than the existing algorithms, which fully proves the superiority of the FG-HFS algorithm. More importantly, the Wilcoxon and Friedman statistical experiments results show that the feature selection effect of FG-HFS algorithm is significantly better than that of existing algorithms, no matter in pairwise comparison or multiple comparison.

Most of the previous machine learning-based models for multi-modal medical diagnosis, primarily designed for unimodal images, usually do not fully leverage the potential of multimodal medical images, leading to limited classification accuracy. These conventional methods typically focus only on the intermodality common information, neglecting the intra-modality specific information and assuming that the common information is more effective in disease diagnosis. Moreover, they do not adequately address the impact of fuzzy information between different medical imaging modalities on diagnostic results. To this end, we propose a Modality Fuzzy Information Separation Network for disease classification, which extracts both common and specific information from fuzzy information to construct a comprehensive representation of multi-modal medical images. Specifically, we extract modality invariant features as common information by explicitly modeling and maximizing loss constraints on mutual information. For specific information extraction, a constraint on feature space independence between specific and common information is imposed on each modality. Above two steps, we concatenate common information and specific information to construct a comprehensive multi-modal representation for separating fuzzy information. Finally, we purposely design a decoder network to reconstruct medical images from uni-modal specific information and common information to demonstrate the effectiveness of the modality fuzzy information separation network. We conducted a validation of the proposed method's performance in classifying cardiomegaly, pneumothorax, edema, and skin disease. The experimental results substantiate the effectiveness of our proposed approach.

Data imbalance is a common phenomenon in machine learning. In the imbalanced data classification, minority samples are far less than majority samples, which makes it difficult for minority to be effectively learned by classifiers. A synthetic minority oversampling technique (SMOTE) improves the sensitivity of classifiers to minority by synthesizing minority samples without repetition. However, the process of synthesizing new samples in the SMOTE algorithm may lead to problems such as "noisy samples" and "boundary samples." Based on the above description, we propose a synthetic minority oversampling technique based on Gaussian mixture model filtering (GMF-SMOTE). GMF-SMOTE uses the expected maximum algorithm based on the Gaussian mixture model to group the imbalanced data. Then, the expected maximum filtering algorithm is used to filter out the "noisy samples" and "boundary samples" in the subclasses after grouping. Finally, to synthesize majority and minority samples, we design two dynamic oversampling ratios. Experimental results show that the GMF-SMOTE performs better than the traditional oversampling algorithms on 20 UCI datasets. The population averages of sensitivity and specificity indexes of random forest (RF) on the UCI datasets synthesized by GMF-SMOTE are 97.49% and 97.02%, respectively. In addition, we also record the G-mean and MCC indexes of the RF, which are 97.32% and 94.80%, respectively, significantly better than the traditional oversampling algorithms. More importantly, the two statistical tests show that GMF-SMOTE is significantly better than the traditional oversampling algorithms.