Deep learning has become a powerful tool to automatically classify medical hyperspectral images (MedHSIs) for the diagnosis of various tumors such as cancer. These deep learning based classification approaches consist of both feature extraction and disease prediction, which are independent of each other. Therefore, the extracted features may be incompatible with the used classifier for prediction. To remedy such deficiency, in this work, we propose a novel deep method, i.e ., deep margin cosine autoencoder (DMCA), for the MedHSI classification, which provides solid support to diagnose tumors. To be specific, first , we consider deep autoencoder (DAE) network as the basic framework for the feature extraction of the MedHSI, and the soft-max classifier is introduced as the output layer to predict the results. Second , to report the features well compatible with the classifier, the importance of the soft-max is added as a constraint into the DAE network. Further, a cosine margin is introduced to enhance the discrimination of different feature clusters. In addition, we also design an optimization scheme to the solutions of the DMCA model. Third , a two-stage training strategy is presented to train the built DMCA network. After completing the training, unknown tissues ( i.e ., unlabeled samples) can be specified to determine whether they are tumors or not. Sufficient experimental results have been provided to validate that our DMCA method achieves better classification performance and higher tumor diagnosis accuracy compared with some advanced approaches.

This article presents a novel deep network with irregular convolutional kernels and self-expressive property (DIKS) for the classification of hyperspectral images (HSIs). Specifically, we use the principal component analysis (PCA) and superpixel segmentation to obtain a series of irregular patches, which are regarded as convolutional kernels of our network. With such kernels, the feature maps of HSIs can be adaptively computed to well describe the characteristics of each object class. After multiple convolutional layers, features exported by all convolution operations are combined into a stacked form with both shallow and deep features. These stacked features are then clustered by introducing the self-expression theory to produce final features. Unlike most traditional deep learning approaches, the DIKS method has the advantage of self-adaptability to the given HSI due to building irregular kernels. In addition, this proposed method does not require any training operations for feature extraction. Because of using both shallow and deep features, the DIKS has the advantage of being multiscale. Due to introducing self-expression, the DIKS method can export more discriminative features for HSI classification. Extensive experimental results are provided to validate that our method achieves better classification performance compared with state-of-the-art algorithms.

Hyperspectral image (HSI) classification is a current research hotspot. Most existing methods usually export discriminative features with low-quality distribution and low information utilization, which may induce classification performance degeneration. To remedy such deficiencies, we propose a diagonalized low-rank learning (DLRL) model for HSI classification in this study. Specifically, a classwise regularization is used to capture the classwise block-diagonal structure of low-rank representation, which can further cluster the represented HSI pixels from one class into the same subspace and extract features with well-ordered distribution. Such a regularization assists to easily and correctly classify HSIs. In addition, we combine sparsity and collaboration to extract more discriminative features for guaranteeing high information utilization, i.e., a tradeoff of sparsity and collaboration is sought to acquire both correlations among HSI pixels and characteristics of each pixel. By this way, rich information in the HSI can be fully used for good feature extraction. Further, the estimated feature representation is used as an input to the support vector machine (SVM) classifier for HSI classification. Extensive experiments have been done to validate that the proposed DLRL method achieves much classification performance in contrast to several state-of-the-art algorithms.