Near-Infrared Spectroscopy in Peripheral Artery Disease: Differentiating Upper and Lower Limbs and its Correlation with the Ankle-Brachial Index

Abstract

Background: Peripheral artery disease (PAD) represents a global health challenge. Current diagnostic methods, including the ankle-brachial index (ABI), have limitations, particularly in patients with arterial calcification. Near-infrared spectroscopy (NIRS) offers potential advantages as a non-invasive assessment tool, yet its clinical utility in PAD remains underexplored. Objective: To evaluate NIRS in PAD by 1) assessing its ability to differentiate non-ischemic upper limbs from ischemic lower limbs, and 2) evaluating its correlation with ABI. Methods: This observational, cross-sectional study included 51 PAD patients (86.3% male, mean age 65.5±8.3 years). A portable spectrometer (900-1700 nm) recorded NIRS measurements from the right thumb and both halluces at rest. Data underwent Standard Normal Variates transformation for pre-processing. Random Forest classification was implemented to differentiate upper and lower limbs, with data split into training (2/3) and testing (1/3) sets. Interval Partial Least Squares regression (iPLS) identified wavelength regions correlating with ABI. Results: Random Forest analysis revealed distinct NIRS patterns between upper and lower limbs, with an area under the ROC curve of 0.91 (95% CI 0.88-0.94). The strongest discrimination occurred below 1060 nm and between 1250-1375 nm, corresponding to oxygenated hemoglobin absorption pattern. iPLS identified 1429-1463 nm as the most informative interval for ABI prediction, with a modest correlation (R²=0.167, RMSECV=0.186). Conclusions: NIRS demonstrates strong discriminative capability between non-ischemic upper and ischemic lower limbs in PAD. While the correlation between NIRS and ABI was modest, it suggests potential clinical relevance. These findings indicate that NIRS could be a rapid, portable, non-invasive complementary tool for PAD assessment.