Non-destructive monitoring of cartilage tissue engineering via near-infrared (NIR) spectroscopic assessment of culture medium

Abstract

Degenerative joint diseases are major contributors to pain, impaired mobility, and an increasing socioeconomic burden in aging populations. Tissue Engineering (TE) offers alternative solutions to restore joint functions. However, current TE approaches often fail to produce biocompatible, patient-specific cartilage. This could be due to the lack of a rapid, non-destructive assessment for monitoring tissue growth. Biological markers such as hyaluronan, lactate, and collagen are released in culture medium during tissue maturation, and measuring their levels indicates tissue growth and development. We performed near-infrared spectroscopy (NIRS) on conditioned medium collected during culture and differentiation of Na purified gellan gum (NaGG) embedded bone marrow stem/stromal cell (BMSC) constructs. We also employed NIR preprocessing using Nippy and an unbiased machine learning (ML) approach, where multiple ML regression models like linear regression, ridge regression, random forest regressor, and support vector regressor were used to correlate the spectral variations to the biomarkers released into culture medium. The analysis revealed strong correlations between near infrared (NIR) spectral data and biomarkers, with R² values of 0.976 for hyaluronan, 0.701 for lactate, and 0.444 for total collagen, and mean absolute percentage errors (MAPE) of 0.276, 0.090, and 0.158 respectively, showing the efficacy of NIRS in monitoring these key indicators. Feature importance further determined key regions of interest in NIR spectra for each of the biomarkers. In conclusion, this research is the first to demonstrate the potential of using NIRS for non-destructive assessment of tissue growth through the culture medium.