Quantitative diagnosis of cervical precancer using fluorescence intensity and lifetime imaging from the stroma

Abstract

Fluorescence microscopy has been widely used in characterizing the pathological states of tissues because intensity and spectra arise from fluorescence emission can reveal structural and biochemical information of biological samples and the fluorescence excited state lifetime has been verified to identify tissue pathology due to its sensitivity to the fluorophore microenvironment. In this study, we have demonstrated that early cervical cancer can be quantitatively diagnosed using intensity and lifetime derived from the stroma fluorescence in conjunction with extreme learning machine (ELM) classifier which can result in a concurrently high sensitivity of 99.1% and specificity of 99.6%. The results suggest that the proposed technique can be used to aid and supplement the traditional histopathological examination of cervical precancer.