Arthroscopic near infrared spectroscopy enables simultaneous quantitative evaluation of articular cartilage and subchondral bone in vivo

摘要: Arthroscopic assessment of articular tissues is highly subjective and poorly reproducible. To ensure optimal patient care, quantitative techniques (e.g., near infrared spectroscopy (NIRS)) could substantially enhance arthroscopic diagnosis initial signs post-traumatic osteoarthritis (PTOA). Here, we demonstrate, for the first time, potential NIRS to simultaneously monitor progressive degeneration cartilage subchondral bone in vivo Shetland ponies undergoing different experimental repair procedures. Osteochondral adjacent sites were evaluated using an probe significant (p < 0.05) degenerative changes observed tissue properties when compared with from healthy joints. Artificial neural networks (ANN) enabled reliable (ρ = 0.63–0.87, NMRSE = 8.5–17.2%, RPIQ = 1.93–3.03) estimation biomechanical properties, plate thickness mineral density (BMD), trabecular thickness, volume fraction (BV), BMD, structure model index (SMI) vitro spectral data. The trained ANNs also reliably predicted independent test group (ρ = 0.54–0.91, NMRSE = 5.9–17.6%, RPIQ = 1.68–3.36). However, predictions based on NIR spectra less (ρ = 0.27–0.74, NMRSE = 14.5–24.0%, RPIQ = 1.35–1.70), possibly due errors introduced during acquisition. Adaptation address limitations conventional arthroscopy through lesion severity extent, thereby enhancing detection PTOA. This would be high clinical significance, example, conducting orthopaedic surgeries.