Optical imaging is an emerging field of clinical diagnostics that can address the growingudmedical need for early cancer detection and diagnosis. Various human cancers areudamenable to better prognosis and patient survival if found and treated during earlyuddisease onset. Besides providing wide-field, macroscopic diagnostic information similarudto existing clinical imaging techniques, optical imaging modalities have the addedudadvantage of microscopic, high resolution cellular-level imaging from in vivo tissues in realudtime. This comprehensive imaging approach to cancer detection and the possibility ofudperforming an ‘optical biopsy’ without tissue removal has led to growing interest in theudfield with numerous techniques under investigation. Three optical techniques areuddiscussed in this thesis, namely multispectral fluorescence imaging (MFI), hyperspectraludreflectance imaging (HRI) and fluorescence confocal endomicroscopy (FCE). MFI andudHRI are novel endoscopic imaging-based extensions of single point detection techniques,udsuch as laser induced fluorescence spectroscopy and diffuse reflectance spectroscopy.udThis results in the acquisition of spectral data in an intuitive imaging format that allowsudfor quantitative evaluation of tissue disease states. We demonstrate MFI and HRI onudfluorophores, tissue phantoms and ex vivo tissues and present the results as an RGBudcolour image for more intuitive assessment. This follows dimensionality reduction of theudacquired spectral data with a fixed-reference isomap diagnostic algorithm to extract onlyudthe most meaningful data parameters. FCE is a probe-based point imaging techniqueudoffering confocal detection in vivo with almost histology-grade images. We perform FCEudimaging on chemotherapy-treated in vitro human ovarian cancer cells, ex vivo humanudcancer tissues and photosensitiser-treated in vivo murine tumours to show the enhanceduddetection capabilities of the technique. Finally, the three modalities are applied inudcombination to demonstrate an optical viewfinder approach as a possible minimally-invasiveudimaging method for early cancer detection and diagnosis.
展开▼