Lasers in Surgery and Medicine · 2003 · 84 citations · 39 references
DentistryOral BiologyOral AnatomyOral DiagnosisOral OncologyFluorescence SpectroscopyOral Cavity CancerOral MucosaAutofluorescenceAutofluorescence ImagingOral Lesion DetectionReference Database ComparisonReference Database ConstructionReference Database MatchingTissue AutofluorescenceTissue Fluorescence Spectroscopy
BACKGROUND AND OBJECTIVES: Autofluorescence spectroscopy is a promising tool for oral cancer detection. Its reliability might be improved by using a reference database of spectra from healthy mucosa. We investigated the influence of anatomical location on healthy mucosa autofluorescence. STUDY DESIGN/MATERIALS AND METHODS: Spectra were recorded from 97 volunteers using seven excitation wavelengths (350-450 nm), 455-867 nm emission. We studied intensity and applied principal component analysis (PCA) with classification algorithms. Class overlap estimates were calculated. RESULTS: We observed differences in fluorescence intensity between locations. These were significant but small compared to standard deviations (SD). Normalized spectra looked similar for locations, except for the dorsal side of the tongue (DST) and the vermilion border (VB). Porphyrin-like fluorescence was observed frequently, especially at DST. PCA and classification confirmed VB and DST to be spectrally distinct. The remaining locations showed large class overlaps. CONCLUSIONS: No relevant systematic spectral differences have been observed between most locations, allowing the use of one large reference database. For DST and VB separate databases are required.
39
Multilayer feedforward networks are universal approximators
Kurt Hornik, Maxwell B. Stinchcombe, Halbert White · Neural Networks · 1989 · 21.2K citations
Multilayer feedforward networks are universal approximators
HornikK., StinchcombeM., WhiteH. · Neural Networks · 1989 · 9.4K citations
Rebekah A. Drezek, Konstantin Sokolov, Urs Utzinger et al. · Journal of Biomedical Optics · 2001 · 286 citations · Full text