Endoscopic NIRAF-LSCI multimodal system with pseudo-color visualization for parathyroid autofluorescence localization and tissue perfusion assessment: preclinical validation

Postoperative hypoparathyroidism remains a prevalent complication of thyroid surgery, primarily caused by unrecognized parathyroid gland (PG) injury or devascularization. While near-infrared autofluorescence (NIRAF) enables accurate PG localization, it provides no functional perfusion information; conversely, laser speckle contrast imaging (LSCI) offers label-free real-time perfusion assessment. Herein, we validated a multimodal endoscopic system integrating NIRAF and LSCI for simultaneous PG identification and vascular evaluation, focusing on LSCI perfusion performance and the clinical practicability of its pseudo-color semi-quantitative visualization. The system was tested in 8 Sprague–Dawley rats, where rat thyroids with homologous blood supply served as perfusion surrogates due to the minute volume of rat PGs. Imaging was performed under three strictly controlled in situ conditions: intact native vascularization, partial vascular occlusion, and complete vascular occlusion. LSCI pseudo-color visualization was generated via a standardized pipeline including temporal laser speckle contrast analysis (tLASCA)-based blood flow index (BFI) calculation, 5th–95th percentile normalization, and Jet colormap mapping. Intraoperative probe-based quantitative NIRAF was used for PG localization, and all results were validated against dual gold standards: intraoperative NIRAF probe detection and postoperative histopathology. NIRAF clearly identified PGs with a fluorescence intensity of 140.0 ± 5.0 mV, significantly higher than thyroid (94.5 ± 0.7 mV), trachea (83.0 ± 4.0 mV), and muscle (52.0 ± 2.0 mV) (all P < 0.001), achieving 100% positive/negative predictive values and zero false results. LSCI speckle contrast increased progressively with decreasing perfusion, exhibiting a precise 1:1 correspondence with the continuous blue-to-red pseudo-color perfusion gradient, with histopathology confirming all fluorescent tissues as PGs. This multimodal system addresses two core clinical needs for intraoperative PG protection, and its standardized pseudo-color semi-quantitative workflow provides real-time label-free perfusion data with strong potential to guide clinical PG preservation decisions.