[Study of a multivariate statistical 5-ALA-based discrimination method for fluorescence spectra of colonic tissue of...

In order to diagnosis colon early cancer with laser-induced 5-ALA-PpIX fluorescence spectra, a multivariate statistical method to distinguish these fluorescence spectra acquired in vivo was developed. 343 spectra were collected from 8 normal SD rats, and 20 1,2-DMH-induced SD colon cancer models, and 12 second generation rats of induced rats. 150 min after trail intravenous injections of 5-ALA at a dose of 25 mg x kg(-1) BW, fluorescence spectra excited with 370 nm Ti-laser were collected in vivo. All spectra were divided into a calibration group and a prediction group. After preprocessing, 4 principal components were extracted with PCA. And then, discrimination models were built by stepwise multivariate logistic regression (SMLR) on calibration group. 3 pathological styles were combined each other, and then 3 SMLR models were derived. Normal tissues were classified from early cancers and advanced cancers with sensitivity of 100% and 98.4%, and specificity of 96% and 100%, and accuracy of 98% and 99.2% on prediction group, respectively. The multivariate statistical discrimination method of PCA and SMLR together can effectively distinguish normal tissues from early cancers and advanced cancers with high sensitivity and specificity by means of systemic 5-ALA at low dose. Laser induced fluorescence 5-ALA-based technique is promising for the detection of colonic early cancer.