Validation of interventional fiber optic spectroscopy with MR spectroscopy, MAS-NMR spectroscopy, high-performance thin-layer chromatography, and histopathology for accurate hepatic fat quantification.

OBJECTIVES To validate near-infrared (NIR)-based optical spectroscopy measurements of hepatic fat content using a minimally invasive needle-like probe with integrated optical fibers, enabling real-time feedback during percutaneous interventions. The results were compared with magnetic resonance spectroscopy (MRS) as validation and with histopathology, being the clinical gold standard. Additionally, ex vivo magic angle spinning nuclear magnetic resonance spectroscopy and high-performance thin-layer chromatography were performed for comparison. MATERIALS AND METHODS Ten mice were used for the study, of which half received a regular chow diet and the other half received a high-fat diet to induce obesity and hepatosteatosis. The mice were imaged with a clinical 3-Tesla MR to select a region of interest within the right and left lobes of the liver, where MRS measurements were acquired in vivo. Subsequently, optical spectra were measured ex vivo at the surface of the liver at 6 different positions immediately after resection. Additionally, hepatic fat was determined by magic angle spinning nuclear magnetic resonance spectroscopy and high-performance thin-layer chromatography. Histopathologic analyses were performed and used as the reference standard. Pearson correlation and linear regression analyses were performed to assess the correlation of the various techniques with NIR. A 1-way analysis of variance including post hoc Tukey multiple comparison tests was used to study the difference in fat estimation between the various techniques. RESULTS For both the mice groups, the estimated fat fractions by the various techniques were significantly similar (P = 0.072 and 0.627 for chow diet and high-fat diet group, respectively). The Pearson correlation value between NIR and the other techniques for fat determination showed the same strong linear correlation (P above 0.990; P < 0.001), whereas for histopathologic analyses, which is a rather qualitative measure, the Pearson correlation value was slightly lower (P = 0.925, P < 0.001) . Linear regression coefficient computed to compare NIR with the other techniques resulted in values close to unity with MRS having the narrowest confidence interval (r = 0.935, 95% confidence interval: 0.860-1.009), demonstrating highly correlating results between NIR and MRS. CONCLUSIONS NIR spectroscopy measurements from a needle-like probe with integrated optical fibers for sensing at the tip of the needle can quickly and accurately determine hepatic fat content during an interventional procedure and might therefore be a promising novel diagnosing tool in the clinic.