Biochemical imaging of normal, adenoma, and colorectal adenocarcinoma tissues by Fourier transform infrared spectroscopy (FTIR) and morphological correlation by histopathological analysis: preliminary results

Introduction

The colorectal cancer is a major health problem worldwide. Histology is considered the gold standard for differential diagnosis. However, it depends on the observer's experience, which can lead to discrepancies and poor results. Spectroscopic imaging by Fourier transform infrared (FTIR) is a technique that may be able to improve the diagnosis, because it is based on biochemical differences of the structural constituents of tissue. Therefore, the main goal of this study was to explore the use of FTIR imaging technique in normal colon tissue, colorectal adenoma, and adenocarcinoma in order to correlate their morphological structures with their biochemical imaging.

Methods

Samples were collected from normal (n = 4), adenoma (n = 4), and adenocarcinoma human colorectal tissue (n = 4) from patients undergoing colonoscopy or surgical resection of colon lesions. The samples were sectioned with a cryostat in sequential sections; the first slice was placed on CaF₂ slide and the second slice was placed on glass slide for histological analysis (HE staining). The cluster analyses were performed by the software Cytospec (1.4.02)®.

Results

In normal samples, biochemical analysis classified six different structures, namely the lamina propria of mucous glands (epithelial cells and goblet cells), central lumen of the gland, mucin, and conjunctive tissue. In samples with adenoma and adenocarcinoma, altered regions could also be identified with high sensitivity and specificity.

Conclusion

The results of this study demonstrate the potential and viability of using infrared spectroscopy to identify and classify colorectal tissues.

Infrared imaging; Artificial neural network; Human colon cancer; Biochemical correlation; Tissue classification; FTIR