Fluorescence spectroscopy incorporating a ratiometric approach for the diagnosis and classification of urothelial carcinoma
Authors: Anand S., Cicchi R., Crisci A., Nesi G., Carini M., Pavone F.S.
Autors Affiliation: European Laboratory for Non-Linear Spectroscopy (LENS), University of Florence, Via Nello Carrara, I, Sesto Fiorentino, 50019, Italy; Department of Physics, University of Florence, Via Giovanni Sansone I, Sesto Fiorentino, 50019, Italy; National Institute of Optics, National Research Council (INO-CNR), Largo Enrico Fermi 6 I, Florence, 50125, Italy; Division of Urology, Department of Surgical and Medical Critical Area, University of Florence, Florence, I-50100, Italy
Abstract: The current most popular clinical method for the screening of urothelial carcinoma is white light cystoscopy. This method has inherent disadvantages making a strong genesis towards developing more powerful diagnostic techniques. Laser induced intrinsic fluorescence spectroscopy has been studied as an adjunct to current methods for the detection of tumors. This technique allows real time results based on the changes in spectral profile between normal and tumor tissues. We conducted a pilot study based on fluorescence spectroscopy at two wavelengths 378 and 445 nm excitation for the differentiation of urothelial carcinoma. At both the excitation wavelengths, the measured fluorescence signal showed an increased intensity at wavelengths greater than 520 nm. In addition, the emission profile showed modulation at 580 nm which is due to the reabsorption of emitted fluo- rescence due to hemoglobin. Additionally, we developed a tissue characterizing algorithm, based on fluorescence intensity ratios, F510/F600 and F520/F580 at 378 and 445 nm excitation wavelengths respectively. Further, the results were correlated with the pathologists assessment of urothelial carcinoma. This ratiometric classification algorithm yielded 81% sensitivity and 83% specificity at 378 nm and while at 445 nm excitation we achieved a sensitivity and specificity of 85% and 86% for classifying normal and tumor bladder tissues. In this study we have demonstrated the potential of a simple ratiometric algorithm based on fluorescence spectroscopy could be an alternative tool to tissue biopsy. Furthermore, this technique based fiber-based fluorescence spectroscopy could be integrated into an endoscopy system for use in the operating room.
KeyWords: Computer aided diagnosis; Diagnosis; Endoscopy; Fluorescence spectroscopy; Histology; Tissue; Tumors, Classification algorithm; Diagnostic techniques; Excitation wavelength; Fluorescence intensity ratio; Ratiometric; Sensitivity and specificity; uorescence; Urothelial carcinoma, Fluorescenc