Abstract:Background and objectives Compared with traditional RGB images, medical hyperspectral imagery (HSI) has numerous continuous narrow spectral bands, which can provide rich information for cancer diagnosis. However, the abundant spectral bands also contain a large amount of redundancy information and increase computational complexity. Thus, dimensionality reduction (DR) is essential in HSI analysis. All vector-based DR methods ignore the cubic nature of HSI resulting from vectorization. To overcome the disadvantage of vector-based DR methods, tensor-based techniques have been developed by employing multi-linear algebra. Methods To fully exploit the structure features of medical HSI and enhance computational efficiency, a novel method called unsupervised dimensionality reduction via tensor-based low-rank collaborative graph embedding (TLCGE) is proposed. TLCGE introduces entropy rate superpixel (ERS) segmentation algorithm to generate superpixels. Then, a low-rank collaborative graph weight matrix is constructed on each superpixel, greatly improving the efficiency and robustness of the proposed method. After that, TLCGE reduces dimensions in tensor space to well preserve intrinsic structure of HSI. Results The proposed TLCGE is tested on cholangiocarcinoma microscopic hyperspectral data sets. To further demonstrate the effectiveness of the proposed algorithm, other machine learning DR methods are used for comparison. Experimental results on cholangiocarcinoma microscopic hyperspectral data sets validate the effectiveness of the proposed TLCGE. Conclusions The proposed TLCGE is a tensor-based DR method, which can maintain the intrinsic 3-D data structure of medical HSI. By imposing the low-rank and sparse constraints on the objective function, the proposed TLCGE can fully explore the local and global structures within each superpixel. The computational efficiency of the proposed TLCGE is better than other tensor-based DR methods, which can be used as a preprocessing step in real medical HSI classification or segmentation.

Patch Tensor-Based Sparse and Low-Rank Graph for HyperspectralImages Dimensionality Reduction

Dimension Reduction of Hyperspectral Images with Sparse Linear Discriminant Analysis

Enhanced Tensor Low-Rank Representation Learning for Hyperspectral Anomaly Detection

Hyperspectral Restoration Via $L_0$ Gradient Regularized Low-Rank Tensor Factorization

Dimensionality Reduction of Hyperspectral Imagery Using Sparse Graph Learning

Dimensionality Reduction of Hyperspectral Image Based on Local Constrained Manifold Structure Collaborative Preserving Embedding

Hyperspectral Imagery Denoising Via Reweighed Sparse Low-Rank Nonnegative Tensor Factorization

Tensor Discriminative Locality Alignment for Hyperspectral Image Spectral–Spatial Feature Extraction

Learning Tensor Low-Rank Representation for Hyperspectral Anomaly Detection

Dimensionality Reduction and Classification Based on Lower Rank Tensor Analysis for Hyperspectral Imagery

An Efficient Method of Hyperspectral Image Dimension Reduction Based on Low Rank Representation and Locally Linear Embedding

Constrained Manifold Learning for Hyperspectral Imagery Visualization

Spatial-Spectral Graph-Based Nonlinear Embedding Dimensionality Reduction for Hyperspectral Image Classificaiton.

Hyperspectral Image Dimensionality Reduction via Graph Embedding in Core Tensor Space

Irregular Tensor Low-Rank Representation for Hyperspectral Image Representation

Unsupervised Dimensionality Reduction of Medical Hyperspectral Imagery in Tensor space

Multimanifold Bistructured Low Rank Representation of hyperspectral images

Tensor Cascaded-Rank Minimization in Subspace: A Unified Regime for Hyperspectral Image Low-Level Vision

Spectral Super-Resolution via Deep Low-Rank Tensor Representation

Weighted Low-rank Tensor Recovery for Hyperspectral Image Restoration

Robust Dimensionality Reduction via Low-rank Laplacian Graph Learning