Spectral-Spatial Classification Using Tensor Modeling for Cancer Detection with Hyperspectral Imaging

• Type of Material: Conference Paper
• Authors: Guolan Lu, Georgia Institute of Technology; Luma Halig, Emory University; Dongsheng Wang, Emory University; Georgia Chen, Emory University; Baowei Fei, Emory University
• Language: English
• Date: 2014-01-01
• Publisher: Emory University Libraries
• Publication Version: Author Accepted Manuscript (After Peer Review)
• Copyright Statement: © 2014 SPIE.
• Final Published Version (URL): http://dx.doi.org/10.1117/12.2043796
• Title of Journal or Parent Work: Proceedings of SPIE
• Conference or Event Name: Conference on Medical Imaging - Image Processing
• Volume: 9034
• Start Page: 903413
• End Page: 903413
• Keywords: Feature ranking; Physical Sciences; Technology; Optics; Science & Technology; Hyperspectral imaging; Head and neck cancer; Dimension reduction; Life Sciences & Biomedicine; Imaging Science & Photographic Technology; Radiology, Nuclear Medicine & Medical Imaging; Tensor modeling; Tucker tensor decomposition
• Research Categories: Computer Science; Health Sciences, Radiology; Engineering, Biomedical

Abstract

As an emerging technology, hyperspectral imaging (HSI) combines both the chemical specificity of spectroscopy and the spatial resolution of imaging, which may provide a non-invasive tool for cancer detection and diagnosis. Early detection of malignant lesions could improve both survival and quality of life of cancer patients. In this paper, we introduce a tensor-based computation and modeling framework for the analysis of hyperspectral images to detect head and neck cancer. The proposed classification method can distinguish between malignant tissue and healthy tissue with an average sensitivity of 96.97% and an average specificity of 91.42% in tumor-bearing mice. The hyperspectral imaging and classification technology has been demonstrated in animal models and can have many potential applications in cancer research and management.