Machine learning augmented near-infrared spectroscopy: In vivo follow-up of cartilage defects.

OBJECTIVE: To assess the potential of near-infrared spectroscopy (NIRS) for in vivo arthroscopic monitoring of cartilage defects. METHOD: Sharp and blunt cartilage grooves were induced in the radiocarpal and intercarpal joints of Shetland ponies and monitored at baseline (0 weeks) and at three follow-up timepoints (11, 23, and 39 weeks) by measuring near-infrared spectra in vivo at and around the grooves. The animals were sacrificed after 39 weeks and the joints were harvested. Spectra were reacquired ex vivo to ensure reliability of in vivo measurements and for reference analyses. Additionally, cartilage thickness and instantaneous modulus were determined via computed tomography and mechanical testing, respectively. The relationship between the ex vivo spectra and cartilage reference properties was determined using convolutional neural network. RESULTS: In an independent test set, the trained networks yielded significant correlations for cartilage thickness (ρ = 0.473) and instantaneous modulus (ρ = 0.498). These networks were used to predict the reference properties at baseline and at follow-up time points. In the radiocarpal joint, cartilage thickness increased significantly with both groove types after baseline and remained swollen. Additionally, at 39 weeks, a significant difference was observed in cartilage thickness between controls and sharp grooves. For the instantaneous modulus, a significant decrease was observed with both groove types in the radiocarpal joint from baseline to 23 and 39 weeks. CONCLUSION: NIRS combined with machine learning enabled determination of cartilage properties in vivo, thereby providing longitudinal evaluation of post-intervention injury development. Additionally, radiocarpal joints were found more vulnerable to cartilage degeneration after damage than intercarpal joints.

Near-infrared spectroscopy enables quantitative evaluation of human cartilage biomechanical properties during arthroscopy.

OBJECTIVE: To investigate the feasibility of near-infrared (NIR) spectroscopy (NIRS) for evaluation of human articular cartilage biomechanical properties during arthroscopy. DESIGN: A novel arthroscopic NIRS probe designed in our research group was utilized by an experienced orthopedic surgeon to measure NIR spectra from articular cartilage of human cadaver knee joints (ex vivo, n = 18) at several measurement locations during an arthroscopic surgery. Osteochondral samples (n = 265) were extracted from the measurement sites for reference analysis. NIR spectra were remeasured in a controlled laboratory environment (in vitro), after which the corresponding cartilage thickness and biomechanical properties were determined. Hybrid multivariate regression models based on principal component analysis and linear mixed effects modeling (PCA-LME) were utilized to relate cartilage in vitro spectra and biomechanical properties, as well as to account for the spatial dependency. Additionally, a k-nearest neighbors (kNN) classifier was employed to reject outlying ex vivo NIR spectra resulting from a non-optimal probe-cartilage contact. Model performance was evaluated for both in vitro and ex vivo NIR spectra via Spearman's rank correlation (ρ) and the ratio of performance to interquartile range (RPIQ). RESULTS: Regression models accurately predicted cartilage thickness and biomechanical properties from in vitro NIR spectra (Model: 0.77 ≤ ρ ≤ 0.87, 2.03 ≤ RPIQ ≤ 3.0; Validation: 0.74 ≤ ρ ≤ 0.84, 1.87 ≤ RPIQ ≤ 2.90). When predicting cartilage properties from ex vivo NIR spectra (0.33 ≤ ρ ≤ 0.57 and 1.02 ≤ RPIQ ≤ 2.14), a kNN classifier enhanced the accuracy of predictions (0.52 ≤ ρ ≤ 0.87 and 1.06 ≤ RPIQ ≤ 1.88). CONCLUSION: Arthroscopic NIRS could substantially enhance identification of damaged cartilage by enabling quantitative evaluation of cartilage biomechanical properties. The results demonstrate the capacity of NIRS in clinical applications.

Multimodality scoring of chondral injuries in the equine fetlock joint ex vivo.

OBJECTIVE: We investigate the potential of a prototype multimodality arthroscope, combining ultrasound, optical coherence tomography (OCT) and arthroscopic indentation device, for assessing cartilage lesions, and compare the reliability of this approach with conventional arthroscopic scoring ex vivo. DESIGN: Areas of interest (AIs, N = 43) were selected from equine fetlock joints (N = 5). Blind-coded AIs were independently scored by two equine surgeons employing International Cartilage Repair Society (ICRS) scoring system via conventional arthroscope and multimodality arthroscope, in which high-frequency ultrasound and OCT catheters were attached to an arthroscopic indentation device. In addition, cartilage stiffness was measured with the indentation device, and lesions in OCT images scored using custom-made automated software. Measurements and scorings were performed twice in two separate rounds. Finally, the scores were compared to histological ICRS scores. RESULTS: OCT and arthroscopic examinations showed the highest average agreements (55.2%) between the scoring by surgeons and histology scores, whereas ultrasound had the lowest (50.6%). Average intraobserver agreements of surgeons and interobserver agreements between rounds were, respectively, for conventional arthroscope (68.6%, 69.8%), ultrasound (68.6%, 68.6%), OCT (65.1%, 61.7%) and automated software (65.1%, 59.3%). CONCLUSIONS: OCT imaging supplemented with the automated software provided the most reliable lesion scoring. However, limited penetration depth of light limits the clinical potential of OCT in assessing human cartilage thickness; thus, the combination of OCT and ultrasound could be optimal for reliable diagnostics. Present findings suggest imaging and quantitatively analyzing the entire articular surface to eliminate surgeon-related variation in the selection of the most severe lesion to be scored.

Semi-automated International Cartilage Repair Society scoring of equine articular cartilage lesions in optical coherence tomography images.

BACKGROUND: Arthroscopic optical coherence tomography (OCT) is a promising tool for the detailed evaluation of articular cartilage injuries. However, OCT-based articular cartilage scoring still relies on the operator's visual estimation. OBJECTIVES: To test the hypothesis that semi-automated International Cartilage Repair Society (ICRS) scoring of chondral lesions seen in OCT images could enhance intra- and interobserver agreement of scoring and its accuracy. STUDY DESIGN: Validation study using equine cadaver tissue. METHODS: Osteochondral samples (n = 99) were prepared from 18 equine metacarpophalangeal joints and imaged using OCT. Custom-made software was developed for semi-automated ICRS scoring of cartilage lesions on OCT images. Scoring was performed visually and semi-automatically by five observers, and levels of inter- and intraobserver agreement were calculated. Subsequently, OCT-based scores were compared with ICRS scores based on light microscopy images of the histological sections of matching locations (n = 82). RESULTS: When semi-automated scoring of the OCT images was performed by multiple observers, mean levels of intraobserver and interobserver agreement were higher than those achieved with visual OCT scoring (83% vs. 77% and 74% vs. 33%, respectively). Histology-based scores from matching regions of interest agreed better with visual OCT-based scoring than with semi-automated OCT scoring; however, the accuracy of the software was improved by optimising the threshold combinations used to determine the ICRS score. MAIN LIMITATIONS: Images were obtained from cadavers. CONCLUSIONS: Semi-automated scoring software improved the reproducibility of ICRS scoring of chondral lesions in OCT images and made scoring less observer-dependent. The image analysis and segmentation techniques adopted in this study warrant further optimisation to achieve better accuracy with semi-automated ICRS scoring. In addition, studies on in vivo applications are required.