Nerve detection using optical spectroscopy, an evaluation in four different models

Abstract

Objective: Identification of peripheral nerve tissue is crucial in both surgery and regional anesthesia. Recently, optical tissue identification methods are presented to facilitate nerve identification in transcutaneous procedures and surgery. Optimization and validation of such techniques require large datasets. The use of alternative models to human in vivo, like human post mortem, or swine may be suitable to test, optimize and validate new optical techniques. However, differences in tissue characteristics and thus optical properties, like oxygen saturation and tissue perfusion are to be expected. This requires a structured comparison between the models. Study Design: Comparative observational study. Methods: Nerve and surrounding tissues in human (in vivo and post mortem) and swine (in vivo and post mortem) were structurally compared macroscopically, histologically, and spectroscopically. Diffuse reflective spectra were acquired (400–1,600 nm) after illumination with a broad band halogen light. An analytical model was used to quantify optical parameters including concentrations of optical absorbers. Results: Several differences were found histologically and in the optical parameters. Histologically nerve and adipose tissue (subcutaneous fat and sliding fat) showed clear similarities between human and swine while human muscle enclosed more adipocytes and endomysial collagen. Optical parameters revealed model dependent differences in concentrations of β-carotene, water, fat, and oxygen saturation. The similarity between optical parameters is, however, sufficient to yield a strong positive correlation after cross model classification. Conclusion: This study shows and discusses similarities and differences in nerve and surrounding tissues between human in vivo and post mortem, and swine in vivo and post mortem; this could support the discussion to use an alternative model to optimize and validate optical techniques for clinical nerve identification. Lasers Surg. Med. 50:253–261, 2018.