| A novel system that can discern between normal and malignant tissues has been developed here. This system is capable of dual-modality hyperspectral imaging (DMHSI). The system consists of a laser, an endoscope, a liquid crystal tunable filter (LCTF), and two cameras coupled with optics and electronics. Preliminary results show that the system can delineate normal and malignant mouse tissues. This dissertation provides the first known multispectral and hyperspectral real-time images taken with an LCTF of the autofluorescence of cancerous and normal tissues in-vivo. In addition, a novel software program was developed to analyze the hyperspectral data in real-time. After taking several complete data sets, the hyperspectral analysis shows that a subset of the data, a mouse predictive tumor wavelength set, can predict the behavior of the total set. It is hoped that this approach can lead to a similar human predictive tumor wavelength set to predict esophageal cancer after clinical trials are complete. |
