Background: Raman spectroscopy is a non-invasive optical technique which can measure molecular vibrational modes within tissue. A large-scale clinical study (n = 518) has demonstrated that real-time Raman spectroscopy could distinguish malignant from benign skin lesions with good diagnostic accuracy; this was validated by a follow-up independent study (n = 127). Objective: Most of the previous diagnostic algorithms have typically been based on analyzing the full band of the Raman spectra, either in the fingerprint or high wavenumber regions. Our objective in this presentation is to explore wavenumber selection based analysis in Raman spectroscopy for skin cancer diagnosis. Methods: A wavenumber selection algorithm was implemented using variably-sized wavenumber windows, which were determined by the correlation coefficient between wavenumbers. Wavenumber windows were chosen based on accumulated frequency from leave-one-out cross-validated stepwise regression or least and shrinkage selection operator (LASSO). The diagnostic algorithms were then generated from the selected wavenumber windows using multivariate statistical analyses, including principal component and general discriminant analysis (PC-GDA) and partial least squares (PLS). A total cohort of 645 confirmed lesions from 573 patients encompassing skin cancers, precancers and benign skin lesions were included. Lesion measurements were divided into training cohort (n = 518) and testing cohort (n = 127) according to the measurement time. Result: The area under the receiver operating characteristic curve (ROC) improved from 0.861–0.891 to 0.891–0.911 and the diagnostic specificity for sensitivity levels of 0.99-0.90 increased respectively from 0.17–0.65 to 0.20–0.75 by selecting specific wavenumber windows for analysis. Conclusion: Wavenumber selection based analysis in Raman spectroscopy improves skin cancer diagnostic specificity at high sensitivity levels.
Jianhua Zhao, Haishan Zeng, Sunil Kalia, and Harvey Lui, "Wavenumber selection based analysis in Raman spectroscopy improves skin cancer diagnostic specificity at high sensitivity levels (Conference Presentation)," Proc. SPIE 10037, Photonics in Dermatology and Plastic Surgery, 1003708 (Presented at SPIE BiOS: January 28, 2017; Published: 19 April 2017); https://doi.org/10.1117/12.2256375.5369935391001.
Conference Presentations are recordings of oral presentations given at SPIE conferences and published as part of the conference proceedings. They include the speaker's narration along with a video recording of the presentation slides and animations. Many conference presentations also include full-text papers. Search and browse our growing collection of more than 14,000 conference presentations, including many plenary and keynote presentations.