A low-cost miniaturized near infrared spectrometer in combination with advance data analysis can discriminate between normal and fibrotic tissue in seconds under physiological conditions. The major differences between spectra of healthy and fibrotic tissue were seen in the following absorption bands; 1509, 1581, 1689 and 1725 nm, which can be attributed to disruption to changes in the collagen network.
Clinical diagnosis of fibrosis is currently reliant on conventional methods. The current “gold standard” for fibrosis diagnosis is histological examination of a biopsy, which is labour intensive and requires extensive sample preparation. Here we show that a portable handheld near-infrared spectrometer coupled with machine learning algorithms can discriminate between kidney and cardiac fibrosis in a rat model of kidney failure compared to healthy rats without kidney failure. The most significant changes in the spectra of fibrotic tissue included shifts in absorption bands at 1509, 1581, 1689 and 1725 nm attributed to collagen components. The best discrimination of fibrosis was achieved in kidney tissue (AUC=0.962), which showed a higher level of fibrosis compared to cardiac tissue (AUC=0.882). The results show the potential of the NIR spectroscopy to detect and to quantify fibrosis in the heart and kidney that in the future could be applied as an intraoperative surgical tool to guide surgical procedures.Zum Volltext