Detecting Articular Cartilage and Meniscus Deformation Effects Using Magnetization Transfer Ultrashort Echo Time (MT-UTE) Modeling during Mechanical Load Application: Ex Vivo Feasibility Study

Abstract

Objective

Ultrashort echo time (UTE) magnetic resonance imaging (MRI) sequences have improved imaging of short T2 musculoskeletal (MSK) tissues. UTE-MRI combined with magnetization transfer modeling (UTE-MT) has demonstrated robust assessment of MSK tissues. This study aimed to investigate the variation of UTE-MT measures under mechanical loading in tibiofemoral cartilage and meniscus of cadaveric knee joints.

Design

Fourteen knee joints from young (n = 8, 42 ± 12 years old) and elderly (n = 6, 89 ± 4 years old) donors were scanned on a 3-T scanner under 3 loading conditions: load = 300 N (Load1), load = 500 N (Load2), and load = 0 N (Unload). UTE-MT sequences were performed at each loading condition. Macromolecular proton fraction (MMF) was calculated from UTE-MT modeling. Wilcoxon rank sum test was used to examine the MRI data differences between loading conditions.

Results

For young donors, MMF increased in all grouped regions of interest (meniscus [M], femoral articular cartilage [FAC], tibial articular cartilage [TAC], articular cartilage regions covered by meniscus [AC-MC], and articular cartilage regions uncovered by meniscus [AC-UC]) when the load increased from 300 to 500 N. The increases in MMF were significant for M (13.3%, P < 0.01) and AC-MC (9.2%, P = 0.04). MMF decreased in all studied regions after unloading, which was significant only for AC-MC (−8.9%, P = 0.01). For elderly donors, MRI parameters did not show significant changes by loading or unloading.

Conclusion

This study highlights the potential of the UTE-MT modeling combined with knee loading in differentiating between normal and abnormal knees. Average tissue deformation effects were likely higher and more uniformly distributed in the joints of young donors compared with elderly donors.

Keywords

knee joint axial loading cartilage meniscus magnetic resonance imaging

Introduction

Osteoarthritis (OA) is one of the most prevalent human joint diseases worldwide, particularly affecting the knee joint. The tibiofemoral articular cartilage (TFC) and meniscus are 2 crucial components of the human knee joint which are under compressive mechanical loads in the standing position. TFC lines on the tibial and femoral subchondral bone surface to facilitate joint articulation.^1,2 Meniscus conducts a major portion of the load from the femur to the tibia.³

Magnetic resonance imaging (MRI) has been increasingly used for OA diagnosis. MRI can provide high soft tissue contrast and accurate morphological assessment of cartilage and meniscus. However, quantitative evaluation of early-stage OA is very challenging with conventional MRI.^4,5 A large portion of cartilage and meniscus possess short T2 values and therefore cannot be quantitatively assessed using conventional MRI sequences.^6-9 Ultrashort echo time (UTE) MRI sequences can image short T2 musculoskeletal (MSK) tissues such as deep layer cartilage, meniscus, and bone with a high signal.^9-18 Specifically, with UTE-MRI, signal can be acquired after radiofrequency (RF) excitation, as quickly as allowed by the RF hardware (e.g., 30 µs), before major decay in transverse magnetization.

Although UTE-MRI sequences have made significant progress in imaging cartilage and meniscus tissues, regular UTE biomarkers, such as UTE T1ρ and T2 are sensitive to the tissue’s orientation angle with respect to the scanner bore axis (B0). This MR orientation sensitivity has been explained by the magic angle effect.^19-21 UTE-MRI combined with magnetization transfer (UTE-MT) modeling has recently been shown to be insensitive to the tissue orientation and to provide robust assessment of MSK tissues.^22-24 This technique employs UTE-MT data acquired with a series of frequency offsets and MT powers to estimate the macromolecular proton fraction (MMF) and macromolecular relaxation time (T2mm) in both short and long T2 tissues.^25-27 T2mm is calculated indirectly using the UTE-MT modeling by separating the contributions of water pool and macromolecular pool in the magnetization transfer. It should be noted that a conventional T2 technique (e.g., CPMG sequence) measures the T2 value of the free water proton pool. Although conventional T2 sequences do not separate the contributions of different pools, their effective echo times are in the order of few milliseconds which miss the T2 decay of macromolecular (<20 µs) and bound water (<1 ms) pools.^9,14,28

MRI-based knee investigation usually takes place in a non-weight-bearing condition that does not mimic the actual physiological and functional conditions of the joint. This can result in delayed detection of early-stage OA. Structural changes in collagen fibrils and proteoglycan (PG) of cartilage and meniscus during early-stage OA can alter the mechanical properties of the tissue.^1,29,30 Joint mechanical malfunction due to early-stage OA can be revealed by recognizing distinct deformation patterns of cartilage and meniscus under known mechanical loads. Several MRI-based studies on knee joint have been performed during mechanical load application in order to detect disease-specific changes in the knee mechanics.³¹ In most of the reported MRI-based knee-loading studies subjects were positioned in a supine condition inside a clinical MRI scanner, and the mechanical load is applied using an MRI-compatible loading device.^32-42

The main objective of this study was to investigate the variation of UTE-MT modeling biomarkers in tibiofemoral cartilage and meniscus of cadaveric knee joints during mechanical load application. The variation pattern of UTE-MT biomarkers in knees during loading were compared between cadaver joints from young and elderly donors.

Materials and Methods

Mechanical Loading Device

An MRI-compatible loading device was manufactured from polyvinyl chloride (PVC) tubes, high-density polyethylene (HDPE) plates, and nylon bolts and screws. The final prototyped loading device is shown in Figure 1 , mounting a cadaveric knee joint wrapped in biohazard absorbent pads. Eight sets of plastic springs (Ultem* PEI resin, LL100125U40G, Lee Spring, NY) were used in the loading device. The applied compression load was manually adjustable using a 1-inch nylon screw. The applied load was calculated by the spring stiffness multiplied by the average spring deformation length. The average actual stiffness of springs was 3.2 N/mm, as measured by the manufacturer.

Figure 1.

Designed and fabricated MRI compatible loading device using 8 sets of plastic springs. The load is adjustable manually by an axially installed nylon screw (1″ diameter) and monitored by measuring the average length of the springs.

Sample Preparation

In total, 14 fresh-frozen human knee joints from young (n = 8, 42 ± 12 years old, <60 years old) and elderly (n = 6, 89 ± 4 years old, >75 years old) donors were provided by a nonprofit donation company (United Tissue Network, Phoenix, AZ). The knee specimens underwent 1 freeze-thaw cycle before scanning. The proximal and distal shafts (femoral and tibial) were trimmed in order to fit the joints into the loading device. The final length of the joints was 30 to 40 cm ( Fig. 1 ). The cadaveric joints were wrapped in biohazard absorbent pads and then sealed before being placed in the loading device.

UTE-MRI Scans and Loading

All 14 knee specimens were scanned under the following 3 loading conditions: load = 300 N (Load1), load = 500 N (Load2), and load = 0 N (Unload). In order to reduce scan time, an initial step without load was not performed, as we first assumed that the final Unload step would provide similar results. As this assumption was found plausible later, the last 4 of the specimens underwent an initial step with no load (Load0) to investigate the difference between initial Load0 and final Unload steps. A 20-minute rest time was incorporated between each loading condition.

Knee joints mounted in the loading device were placed parallel to B0 and scanned on a clinical 3-T MR scanner (MR750, GE Healthcare Technologies, Milwaukee, WI) using an 8-channel knee coil. For each loading stage, the following 2 UTE-MRI imaging protocols were performed: (A) an actual flip angle-variable TR (AFI-VTR) based 3D UTE-Cones sequence (AFI: TE = 0.032 ms, TRs = 20 and 100, flip angle [FA] = 45°; VTR: TE = 0.032 ms; TRs = 20, 30, 50, and 100 ms; FA = 45°; rectangular RF pulse with a duration of 150 µs) for T1 measurement,⁴³ which is a prerequisite for accurate UTE-MT modeling; and (B) a 3D UTE-Cones-MT sequence (Fermi saturation pulse powers = 500°, 1,000°, and 1,500°; frequency offset = 2, 5, 10, 20, and 50 kHz; FA = 7°; 9 spokes acquired after each MT preparation; rectangular RF excitation pulse with a duration of 26 µs) for UTE-MT modelling.^14,25 Field of view (FOV), matrix dimension, nominal in-plane pixel size, and slice thickness were 140 × 140 mm², 256 × 256 × 40, 0.54 × 0.54 mm², and 2 mm for tibial specimens, respectively.

MRI Data Analysis

UTE-MT and T1 datasets were analyzed within 5 regions of interest (ROIs) defined at both lateral and medial compartments of the knee joints. Figure 2 depicts schematically the location of the drawn ROIs on the medial compartment of a representative knee joint (54-year-old-male). Meniscus (M), femoral articular cartilage covered by meniscus (FAC-MC), tibial articular cartilage covered by meniscus (TAC-MC), femoral articular cartilage uncovered by meniscus (FAC-UC), and tibial articular cartilage uncovered by meniscus (TAC-UC) were the selected ROIs for MRI analysis. FAC ROI was generated by combining FAC-MC and FAC-UC. TAC ROI was generated by combining TAC-MC and TAC-UC. Articular cartilage region covered by meniscus (AC-MC) was generated by combining FAC-MC and TAC-MC. Articular cartilage region uncovered by meniscus (AC-UC) was generated by combining FAC-UC and TAC-UC. ROIs were defined by a graduate student with 2 years of experience in knee joint MRI analysis.

Figure 2.

Schematic ROIs defined at the lateral compartment of a representative cadaveric knee joint (54-year-old male). M refers to meniscus. FAC-MC and TAC-MC refer to femoral and tibial articular cartilage covered by meniscus. FAC-UC and TAC-UC refer to femoral and tibial articular cartilage uncovered by meniscus.

UTE-T1 data were fitted using single component exponential model. The acquired UTE-MT data were fitted using the UTE-MT model,²⁵ which results in an estimation of MMF and T2mm in cartilage and meniscus ROIs. All UTE-MRI measurement and the ROI defining process were performed using MATLAB (version 2017, The Mathworks Inc., Natick, MA) codes developed in-house.

Statistical Analysis

UTE-MRI quantifications were compared between the 3 loading conditions (Load1, Load2, and Unload) for the 14 scanned joints within M, FAC, TAC, AC-MC, and AC-UC regions. The comparisons were performed separately for joints from young and elderly donors on the results averaged over lateral and medial joint compartments. The Load0 (before loading) and Unload datasets were compared for the last 4 specimens with extra scans. Load0-Unload comparison was meant to examine the initial hypothesis that the final Unload step would provide similar results to Load0. Performing one-sample Kolmogorov-Smirnov test showed that the measured MRI parameters in this study were not normally distributed. Hence, Wilcoxon rank sum test was used to examine the MRI data differences between loading conditions (Load1, Load2, and Unload). P values below 0.05 were considered significant. All statistical analyses were performed in MATLAB.

Results

Figure 3 demonstrates the 2-pool MT modeling analyses performed in lateral meniscus (M in Fig. 2 ) of a 54-year-old-male donor at 3 different loading conditions (Load1, Load2, and Unload). MT modeling was performed using UTE-MT datasets for 5 off-resonance frequencies and 3 MT saturation pulse power levels. MMF increased from 20.7% to 23.9% when the applied load increased from 300 to 500 N, and then decreased to 19.1% after unloading.

Figure 3.

The UTE-MT modeling analyses in posterior meniscus (M shown in Fig. 2 ) of a representative cadaveric knee joint (54-year-old male) during 3 different loading conditions; (a) Load1, (b) Load2, and (c) Unload. Modeling was performed using 3 pulse power levels (500° in blue, 1,000° in green, and 1,500° in red) and 5 frequency offsets (2, 5, 10, 20, 50 kHz). MMF and T2mm refer to macromolecular fraction and macromolecular T2, respectively. MMF has increased by loading and eventually decreased by unloading.

Figure 4 shows MMF pixel maps in tibiofemoral cartilage and meniscus of the same cadaveric knee joint (lateral joint compartment) under the 3 loading conditions (Load1, Load2, and Unload). MMF demonstrated an obvious increasing pattern by loading, and a decreasing pattern by unloading.

Figure 4.

MMF pixel maps on a sagittal slice at lateral compartment of a representative knee joint (54-year-old male) under 3 different loading conditions: (a) Load1, (b) Load2, and (c) Unload. Obviously MMF has increased by increasing the applied load from 300 to 500 N approximately and eventually decreased to lower values after unloading.

Average and standard deviation values of T1, MMF, and T2mm within the 5 different sets of ROIs (M, FAC, TAC, AC-MC, and AC-UC), averaged over lateral and medical joint compartments, are presented in Table 1 for young and old joint groups.

Table 1.

Average T1, MMF, and T2mm within Different ROIs for Joints from Young and Elderly Donors^a.

		T1 (ms)			MMF (%)			T2mm (µs)
		Load1	Load2	Unload	Load1	Load2	Unload	Load1	Load2	Unload
Young	M	702 ± 219	627 ± 76	651 ± 82	17.7 ± 3.6	20.0 ± 2.7	18.7 ± 3.2	7.8 ± 1.2	7.5 ± 0.2	7.3 ± 0.5
	FAC	665 ± 141	679 ± 73	662 ± 120	12.7 ± 3.1	13.9 ± 2.8	13.0 ± 2.1	6.9 ± 0.6	6.8 ± 0.5	6.8 ± 0.5
	TAC	633 ± 100	646 ± 81	763 ± 221	14.3 ± 3.1	15.4 ± 3.0	13.8 ± 2.5	7.1 ± 0.4	7.0 ± 0.4	6.9 ± 0.4
	AC-UC	644 ± 99	669 ± 84	755 ± 260	14.1 ± 3.1	15.0 ± 3.3	13.8 ± 2.5	6.8 ± 0.5	6.7 ± 0.4	6.8 ± 0.3
	AC-MC	651 ± 133	659 ± 76	692 ± 201	13.3 ± 3.2	14.5 ± 2.9	13.2 ± 2.2	7.1 ± 0.5	7.0 ± 0.5	6.9 ± 0.5
Elderly	M	686 ± 85	756 ± 143	791 ± 210	16.0 ± 1.7	15.5 ± 3.3	16.2 ± 2.9	7.6 ± 0.3	7.8 ± 0.3	7.7 ± 0.3
	FAC	665 ± 76	700 ± 84	790 ± 179	12.9 ± 1.9	13.3 ± 3.1	12.0 ± 3.2	7.1 ± 0.3	7.2 ± 0.3	7.3 ± 0.3
	TAC	621 ± 90	689 ± 85	740 ± 95	14.5 ± 1.6	14.8 ± 2.0	14.0 ± 2.1	7.2 ± 0.3	7.3 ± 0.2	7.3 ± 0.3
	AC-UC	615 ± 90	650 ± 60	750 ± 169	14.2 ± 2.3	14.5 ± 2.4	13.6 ± 3.0	7.0 ± 0.2	7.2 ± 0.3	7.2 ± 0.2
	AC-MC	657 ± 80	717 ± 86	773 ± 132	13.4 ± 1.7	13.8 ± 2.8	12.7 ± 2.8	7.2 ± 0.3	7.3 ± 0.2	7.3 ± 0.3

ROI = region of interest; M = meniscus; FAC = femoral articular cartilage; TAC = tibial articular cartilage; AC-UC = articular cartilage uncovered by meniscus; AC-MC = articular cartilage covered by meniscus.

All values are presented as mean ± SD.

Average percentage differences of T1, MMF, and T2mm between Load2 and Load1, as well as between Unload and Load2, in knee joints from young and old donors are presented in Table 2 . For joints from young donors, average MMF increased in all grouped ROIs by increasing the load from 300 to 500 N (13.3%, 9.0%, 7.3%, 6.1%, and 9.2% for M, FAC, TAC, AC-UC, and AC-MC, respectively). The observed MMF increases by loading was significant only for meniscus (M, 13.3%, P < 0.01) and cartilage covered by meniscus (AC-MC, 9.2%, P = 0.04). Average MMF decreased in all studied regions after unloading (−6.7%, −6.3%, −10.4%, −7.5%, and −8.9% for M, FAC, TAC, AC-UC, and AC-MC, respectively), but within a lower range compared with average MMF increases induced by loading. The MMF decrease by unloading was significant only for meniscal covered cartilage (AC-MC, −8.9%, P = 0.01). T1 and T2mm did not show any significant changes by loading or unloading in joints from young donors. For joints from elderly donors, none of the studied MRI parameters showed significant changes by loading or unloading.

Table 2.

Average Percentage Differences of T1, MMF, and T2mm between the Load2 and Load1 as Well as between Load2 and Load in Knee Joints from Young and Old Donors^a.

		T1		MMF		T2mm
		L1-L2 Diff^b (%)	L2-UnL Diff^c (%)	L1-L2 Diff (%)	L2-UnL Diff (%)	L1-L2 Diff (%)	L2-UnL Diff (%)
Young	M	−10.7 (P = 0.45)	3.9 (P = 0.33)	13.3 (P < 0.01)	−6.7 (P = 0.11)	−3.3 (P = 0.52)	−2.7 (P = 0.28)
	FAC	2.2 (P = 0.07)	−2.5 (P = 0.63)	9.0 (P = 0.09)	−6.3 (P = 0.14)	−2.4 (P = 0.12)	1.2 (P = 0.44)
	TAC	2.2 (P = 0.28)	18.1 (P = 0.15)	7.3 (P = 0.16)	−10.4 (P = 0.01)	−1.0 (P = 0.39)	−1.8 (P = 0.16)
	AC-UC	4.0 (P = 0.20)	12.7 (P = 0.77)	6.1 (P = 0.40)	−7.5 (P = 0.21)	−1.8 (P = 0.13)	1.5 (P = 0.52)
	AC-MC	1.2 (P = 0.08)	4.9 (P = 0.51)	9.2 (P = 0.04)	−8.9 (P < 0.01)	−1.6 (P = 0.26)	−1.3 (P = 0.39)
Elderly	M	10.1 (P = 0.16)	4.6 (P = 0.85)	−3.0 (P = 0.99)	4.0 (P = 0.46)	1.5 (P = 0.29)	−1.2 (P = 0.3)
	FAC	5.4 (P = 0.18)	12.8 (P = 0.09)	3.2 (P = 0.51)	−9.9 (P = 0.23)	1.4 (P = 0.07)	1.5 (P = 0.5)
	TAC	10.8 (P = 0.07)	7.5 (P = 0.06)	2.4 (P = 0.50)	−5.2 (P = 0.12)	0.5 (P = 0.59)	−0.2 (P = 0.47)
	AC-UC	5.7 (P = 0.16)	15.4 (P = 0.01)	1.8 (P = 0.70)	−5.8 (P = 0.46)	2.1 (P = 0.08)	0.9 (P = 0.47)
	AC-MC	9.1 (P = 0.08)	7.8 (P = 0.08)	3.2 (P = 0.48)	−8.3 (P = 0.09)	0.4 (P = 0.48)	0.5 (P = 0.96)

M = meniscus; FAC = femoral articular cartilage; TAC = tibial articular cartilage; AC-UC = articular cartilage uncovered with meniscus; AC-MC = articular cartilage covered with meniscus; MRI = magnetic resonance imaging.

Significant P values were calculated using Wilcoxon rank sum test.

L1-L2 Diff (%) corresponds to MRI measure difference between Load2 and Load1 conditions.

L2-UnL Diff (%) corresponds to MRI measure difference between Unload and Load2 conditions.

Figure 5 shows the bar plots for average MMF in meniscus and meniscal covered cartilage for joints from young donors under 3 different loading conditions (Load1, Load2, and Unload).

Figure 5.

Average MMF in (a) meniscus (M) and (b) meniscal covered cartilage (AC-MC) for joints from young donors under 3 different loading conditions (Load1, Load2, and Unload). The central mark in boxplots indicates the median, while the bottom and top edges of the boxes indicate the 25th and 75th percentiles, respectively. “+” symbol refers to outliers. Average MMF significantly increased by load increase from 300 to 500 N for both meniscus and the meniscal covered cartilage (P < 0.01). MMF decreased by unloading for both studied regions yet only significant for meniscal covered cartilage.

Average percentage differences of MRI measures between the Unload and Load0 (preloading scans) in the last 4 knee joints with extra scans are presented in Table 3 . MMF values in most of regions were significantly different between Load0 and Unload condition (P < 0.05).

Table 3.

Average Percentage Differences (%) of T1, MMF, and T2mm between the Unload and Load0 in 4 Knee Joints with Extra Scans^a. Significance p values were calculated using Wilcoxon rank sum test.

	T1	MMF	T2mm
M	12.7 (P < 0.01)	12.9 (P = 0.01)	2.1 (P = 0.12)
FAC	11.3 (P = 0.01)	13.3 (P = 0.01)	−0.3 (P = 0.56)
TAC	6.5 (P = 0.15)	13.8 (P = 0.06)	2.0 (P = 0.48)
AC-UC	−0.1 (P = 0.98)	19.2 (P = 0.01)	−0.8 (P = 0.46)
AC-MC	1.6 (P = 0.86)	20.4 (P = 0.02)	−0.7 (P = 0.70)

Significant P values were calculated using Wilcoxon rank sum test.

M = meniscus; FAC = femoral articular cartilage; TAC = tibial articular cartilage; AC-UC = articular cartilage uncovered with meniscus; AC-MC = articular cartilage covered with meniscus.

Discussion

The feasibility of using the UTE-MT modeling biomarkers paired with mechanical load application was investigated in this study on human cadaveric knee joints. UTE-MT modeling is very appealing for knee imaging. It has been developed for MSK tissues with both short and long T2 values. Furthermore, UTE-MT modeling has demonstrated negligible sensitivity to tissue orientation angle in the scanner magnet.^13,14 Therefore, this technique is expected to be more robust than conventional T2 and T1ρ techniques which can only evaluate long T2 tissues in the knee joint (e.g., the superficial layers of articular cartilage), and are sensitive to the magic angle.^13,19,20,44

MMF, as the main parameter calculated by UTE-MT modeling, indicates an estimation of the macromolecular matrix density (i.e., collagen and PG) in cartilage and meniscus. Compressing the cartilage and meniscus tissues results is an outward water flux. The effect of such a deformation and water reduction in the tissue can be detected by UTE-MT modeling without high-resolution requirements and can be demonstrated with an MMF increase. The variations of MMF with loading demonstrated different patterns between joints from young and elderly donors. Significant MMF changes were detected in meniscus and meniscal covered cartilage of young donors when increasing the mechanical load ( Table 2 , Fig. 5 ). The studied MRI measures did not demonstrate significant changes in joints from elderly donors. Tissue deformations and its detectable impact by UTE-MT modeling under a certain load were likely higher in young donors compared with elderly donors with potentially degenerated tissues. The main portion of mechanical load was likely transferred through the meniscus from the femur to the tibia, where larger deformation occurs in young and normal knee joints. Mechanical load in elderly joints might be distributed differently from young joints, such that the meniscus bears less of the load distribution. Specifically, mechanical load increase might lead to meniscus extrusion, which is common in elderly patients.⁴⁵ Moreover, loading the older joints may lead to joint misalignment and inconsistent or even reversed MRI variation patterns in the lateral, medial, posterior, and anterior planes of the joints.

MMF did not decrease significantly after unloading, except for meniscus and tibial cartilage in joints from young donors. Since this was a cadaveric study, cartilage and menisci could not restore their original shapes, particularly with the limited rest time (i.e., 20 minutes). MMF values in most regions were significantly higher in the Unload dataset compared with Load0 (before loading scans) in the last 4 knee joints with extra scans ( Table 3 ). These results undermined our initial assumptions for the required rest time for cadaveric tissues after loading.

Other quantitative MRI biomarkers, such as delayed gadolinium-enhanced T1, T2, and T1ρ, have been used in previous knee loading studies.^33-36,46 MRI-based knee loading studies have been summarized recently in a review paper.³¹ Nevertheless, such quantitative imaging techniques cannot reliably evaluate the loading effects in some knee joint tissues with short T2 (e.g., meniscus and deep layer cartilage) because of 2 main limitations: first, these MRI techniques are incapable of proper imaging of short T2 tissues, and second, they are orientation-sensitive techniques as explained by the magic angle effect.^20,47-49

MT modeling is an emerging technique designed for short T2 tissues in the knee joint, demonstrating negligible sensitivity to tissue orientation in the magnet and, importantly, a remarkable sensitivity to cartilage and meniscus deformation under mechanical loading. The different patterns of MMF variations in joints from young and elderly donors under mechanical loading indicate that UTE-MT paired with mechanical loading can serve as a potential tool to differentiate between normal and abnormal knees. However, more investigations, particularly in vivo studies, are required to understand the potential of this combined MRI-mechanical technique.

This study was limited in several aspects. First, a small number of cadaveric knee joints were studied. As is the nature of pilot studies, our results require a validation with larger sampling sizes. Increasing the number of samples can improve the statistical significances. Second, the knee joints were grouped based on the age difference, not clinical OA scoring. Since no significant cartilage or meniscus defect was observable in the MRI images, the classification of the joints into 2 balanced groups of normal and OA groups was not possible in this study. Nevertheless, the age ranges in the young and old donor groups were significantly different; therefore, the unwanted mixing of various conditions in one group was not likely to happen. Third, this study was performed ex vivo on cadaveric knee joints which exhibit several major differences compared with in vivo joints, including the initiated denaturing process in tissues, lack of muscular support for joint alignment, lack of blood and synovial fluid support, and temperature difference.⁵⁰ A well-designed in vivo study should be performed to investigate the variation patterns of UTE-MT measures in joints under mechanical loading for healthy, mild OA, and severe OA subjects. Fourth, this study required a long scan time, which would pose an eventual challenge for subjects to remain still, particularly under mechanical load. Employing different accelerating techniques, such as stretching the cones readout trajectory, could be used to accelerate and optimize the UTE sequences with negligible errors.⁵¹ Advanced image reconstruction techniques such as parallel imaging and compressed sensing may help reducing the total scan time for translational clinical applications.⁵² Fifth, the temporal differences in cartilage and meniscus under loading was not considered. Specifically, retaining constant mechanical load on the specimens is challenging due to the mechanical stress relaxations in the tissues (viscoelastic materials) and in the device parts.

Conclusions

Variations in MMF, an orientation-insensitive measure from UTE-MT modeling, was investigated in limited number of cadaveric knee joints during mechanical loading. MMF variations with loading demonstrated different patterns in joints from young and elderly donors. Significant MMF variations were detected in meniscus and meniscal covered cartilage of young donors by increasing the mechanical load. However, MRI measures did not demonstrate significant variations in joints from elderly donors. The effect of the tissue deformations under a certain load were likely higher in young donors compared with degenerated tissues in elderly donors. Different patterns of MMF variations in joints from young and elderly donors under mechanical loading indicate that UTE-MT paired with mechanical loading can potentially serve as a new tool to differentiate between normal and abnormal knees.

Footnotes

Acknowledgments and Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors acknowledge grant support from National Institute of Health, NIH (R21AR075851, R01AR075825, 1R01NS092650, R01AR062581-06) and VA Clinical Science and Rehabilitation R&D Awards (I01CX001388 and I01RX002604). The authors also acknowledge some free scan hours from GE Healthcare. We thank Jonathan Lee and Lena Kakos for their help on initial data analysis.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical Approval

Ethical approval was not sought for the present study because this was an ex vivo study performed on tissues donated for research purposes.

Informed Consent

Informed consent was not sought for the present study because this was an ex vivo study performed on tissues donated for research purposes provided by United Tissues Network, a nonprofit donation company.

ORCID iD

Saeed Jerban

References

Mansour

. Biomechanics of cartilage. In: Kinesiol Mech Pathomechanics Hum Mov. 2009:66-79.

Fox

AJS

Bedi

Rodeo

. The basic science of articular cartilage: structure, composition, and function. Sports Health. 2009;1(6):461-8.

Makris

Hadidi

Athanasiou

KA.

The knee meniscus: structure-function, pathophysiology, current repair techniques, and prospects for regeneration. Biomaterials. 2011;32(30):7411-31.

Burstein

Bashir

Gray

ML.

MRI techniques in early stages of cartilage disease. Invest Radiol. 2000;35(10):622-38. doi:10.1097/00004424-200010000-00008

Eckstein

Burstein

Link

TM.

Quantitative MRI of cartilage and bone: degenerative changes in osteoarthritis. NMR Biomed. 2006;19(7):822-54. doi:10.1002/nbm.1063

Gold

Thedens

Pauly

Fechner

Bergman

Beaulieu

, et al. MR imaging of articular cartilage of the knee: new methods using ultrashort TEs. AJR Am J Roentgenol. 1998;170(5):1223-6. doi:10.2214/ajr.170.5.9574589

Gatehouse

Bydder

GM.

Magnetic resonance imaging of short T2 components in tissue. Clin Radiol. 2003;58(1_suppl):1-19. doi:10.1053/crad.2003.1157

Robson

Gatehouse

Bydder

GM.

Magnetic resonance: an introduction to ultrashort TE (UTE) imaging. J Comput Assist Tomogr. 2003;27(6):825-46. doi:10.1097/00004728-200311000-00001

Chang

Chung

CB.

UTE imaging in the musculoskeletal system. J Magn Reson Imaging. 2015;41(4):870-83. doi:10.1002/jmri.24713

10.

Bydder

GM.

Qualitative and quantitative ultrashort-TE MRI of cortical bone. NMR Biomed. 2013;26(5):489-506. doi:10.1002/nbm.2906

11.

Jerban

Nazaran

Dorthe

Cory

Carl

, et al. Detecting stress injury (fatigue fracture) in fibular cortical bone using quantitative ultrashort echo time-magnetization transfer (UTE-MT): an ex vivo study. NMR Biomed. 2018;31(11):e3994. doi:10.1002/nbm.3994

12.

Carl

Searleman

Chang

3D adiabatic T_1ρ prepared ultrashort echo time cones sequence for whole knee imaging. Magn Reson Med. 2018; 80(4):1429-39.

13.

Zhu

Cheng

Wong

Xie

, et al. Rotator cuff tendon assessment using magic-angle insensitive 3D ultrashort echo time cones magnetization transfer (UTE-Cones-MT) imaging and modeling with histological correlation. J Magn Reson Imaging. 2018;48(1_suppl):160-8.

14.

Shao

Chang

EY.

Ultrashort echo time magnetization transfer (UTE-MT) imaging and modeling: magic angle independent biomarkers of tissue properties. NMR Biomed. 2016;29(11):1546-52.

15.

Namiranian

Jerban

Dorthe

Masoud-Afsahi

Wong

, et al. Assessment of mechanical properties of articular cartilage with quantitative three-dimensional ultrashort echo time (UTE) cones magnetic resonance imaging. J Biomech. 2020;113:110085.

16.

Jerban

Wei

Jang

Chang

Quantitative magnetic resonance imaging of cortical and trabecular bone. Semin Musculoskelet Radiol. 2020;24(4):386-401.

17.

Jerban

Jang

Chang

Quantitative ultrashort echo time (UTE) magnetic resonance imaging of bone: an update. Front Endocrinol (Lausanne). 2020;11:567417.

18.

Jerban

Chang

Jang

Chang

An update in qualitative imaging of bone using ultrashort echo time magnetic resonance. Front Endocrinol (Lausanne). 2020;11:555756.

19.

Pauli

Bae

Lee

Lotz

Bydder

D’Lima

, et al. Ultrashort-echo time MR imaging of the patella with bicomponent analysis: correlation with histopathologic and polarized light microscopic findings. Radiology. 2012;264(2):484-93. doi:10.1148/radiol.12111883

20.

Shao

Pauli

Tadros

Kavanaugh

, et al. Magic angle effect plays a major role in both T1rho and T2 relaxation in articular cartilage. Osteoarthritis Cartilage. 2017;25(12):2022-30. doi:10.1016/j.joca.2017.01.013

21.

Chang

MR parametric mapping as a biomarker of early joint degeneration. Sports Health. 2016; 8(5):405-11. doi:10.1177/1941738116661975

22.

Carl

Shao

Tadros

Chang

Three-dimensional ultrashort echo time cones T_1ρ (3D UTE-cones-T_1ρ) imaging. NMR Biomed. 2017;30(6):e3709. doi:10.1002/nbm.3709

23.

Jerban

Dorthe

Kakos

Alenezi

, et al. Assessing cortical bone mechanical properties using collagen proton fraction from ultrashort echo time magnetization transfer (UTE-MT) MRI modeling. Bone Rep. 2019;11:100220.

24.

Jerban

Namiranian

Ashir

Shirazian

Wei

, et al. Age-related decrease in collagen proton fraction in tibial tendons estimated by magnetization transfer modeling of ultrashort echo time magnetic resonance imaging (UTE-MRI). Sci Rep. 2019;9(1_suppl):17974.

25.

Chang

Carl

Quantitative magnetization transfer ultrashort echo time imaging using a time-efficient 3D multispoke Cones sequence. Magn Reson Med. 2017;79(2):692-700.

26.

Jerban

Wong

Nazaran

Searleman

Wan

, et al. Ultrashort echo time magnetic resonance imaging (UTE-MRI) of cortical bone correlates well with histomorphometric assessment of bone microstructure. Bone. 2019;123:8-17.

27.

Jerban

Wan

Searleman

Jang

Sah

, et al. Collagen proton fraction from ultrashort echo time magnetization transfer (UTE-MT) MRI modelling correlates significantly with cortical bone porosity measured with micro-computed tomography (µCT). NMR Biomed. 2019;32(2): e4045. doi:10.1002/nbm.4045

28.

Diaz

Carl

Bae

Chung

Bydder

GM.

Ultrashort echo time imaging with bicomponent analysis. Magn Reson Med. 2012;67(3):645-9. doi:10.1002/mrm.23047

29.

Juras

Bittsansky

Majdisova

Szomolanyi

Sulzbacher

Gäbler

, et al. In vitro determination of biomechanical properties of human articular cartilage in osteoarthritis using multi-parametric MRI. J Magn Reson. 2009;197(1_suppl):40-7. doi:10.1016/j.jmr.2008.11.019

30.

Desrochers

Amrein

Matyas

. Viscoelasticity of the articular cartilage surface in early osteoarthritis. Osteoarthritis Cartilage. 2012;20(5):413-21. doi:10.1016/j.joca.2012.01.011

31.

Jerban

Chang

Magnetic resonance imaging (MRI) studies of knee joint under mechanical loading: review. Magn Reson Imaging. 2020;65(1_suppl):27-36.

32.

Mayerhoefer

Welsch

Mamisch

Kainberger

Weber

Nemec

, et al. The in vivo effects of unloading and compression on T1-Gd (dGEMRIC) relaxation times in healthy articular knee cartilage at 3.0 Tesla. Eur Radiol. 2010;20(2):443-9.

33.

Nishii

Kuroda

Matsuoka

Sahara

Yoshikawa

Change in knee cartilage T2 in response to mechanical loading. J Magn Reson Imaging. 2008;28(1_suppl):175-80.

34.

Zevenbergen

Gsell

Chan

Sloten

Himmelreich

Neu

, et al. Functional assessment of strains around a full-thickness and critical sized articular cartilage defect under compressive loading using MRI. Osteoarthritis Cartilage. 2018;26(12):1710-21. doi:10.1016/j.joca.2018.08.013

35.

Souza

Stehling

Wyman

Le Graverand

MPH

Link

et al. The effects of acute loading on T1rho and T2 relaxation times of tibiofemoral articular cartilage. Osteoarthritis Cartilage. 2010;18(12):1557-63.

36.

MacLeod

Subburaj

Kumar

Wyatt

Souza

RB.

Magnetic resonance analysis of loaded meniscus deformation: a novel technique comparing participants with and without radiographic knee osteoarthritis. Skeletal Radiol. 2015;44(1_suppl):125-35.

37.

Nebelung

Sondern

Oehrl

Tingart

Rath

Pufe

, et al. Functional MR imaging mapping of human articular cartilage response to loading. Radiology. 2017;282(2): 464-74.

38.

Wang

Koff

Potter

Warren

Rodeo

Maher

SA.

An MRI-compatible loading device to assess knee joint cartilage deformation: effect of preloading and inter-test repeatability. J Biomech. 2015;48(12):2934-40. doi:10.1016/j.jbiomech.2015.08.006

39.

Chan

Cai

Butz

Trippel

Nauman

Neu

CP.

In vivo articular cartilage deformation: noninvasive quantification of intratissue strain during joint contact in the human knee. Sci Rep. 2016;6:19220.

40.

Neu

CP.

Functional imaging in OA: role of imaging in the evaluation of tissue biomechanics. Osteoarthritis Cartilage. 2014;22(10):1349-59. doi:10.1016/j.joca.2014.05.016

41.

Neu

Walton

JH.

Displacement encoding for the measurement of cartilage deformation. Magn Reson Med. 2008; 59(1_suppl):149-55.

42.

Martin

Neu

Hull

ML.

Quasi-steady-state displacement response of whole human cadaveric knees in a MRI scanner. J Biomech Eng. 2009;131:081004.

43.

Carl

Zhu

Szeverenyi

Bydder

, et al. Accurate T₁ mapping of short T₂ tissues using a three-dimensional ultrashort echo time cones actual flip angle imaging-variable repetition time (3D UTE-Cones AFI-VTR) method. Magn Reson Med. 2018;80(2):598-608. doi:10.1002/mrm.27066

44.

Takahashi

Chung

CB.

Ultrashort TE spectroscopic imaging (UTESI): application to the imaging of short T2 relaxation tissues in the musculoskeletal system. J Magn Reson Imaging. 2009;29(2):412-21.

45.

Svensson

Felson

Zhang

Guermazi

Roemer

Niu

, et al. Meniscal body extrusion and cartilage coverage in middle-aged and elderly without radiographic knee osteoarthritis. Eur Radiol. 2019;29(4):1848-54.

46.

Souza

Baum

Feeley

Kadel

, et al. Effects of unloading on knee articular cartilage T1rho and T2 magnetic resonance imaging relaxation times: a case series. J Orthop Sports Phys Ther. 2012;42(6):511-20.

47.

Fullerton

Cameron

Ord

VA.

Orientation of tendons in the magnetic field and its effect on T2 relaxation times. Radiology. 1985;155(2):433-35. doi:10.1148/radiology.155.2.3983395

48.

Erickson

Prost

Timins

ME.

The “magic angle” effect: background physics and clinical relevance. Radiology. 1993;188(1_suppl):23-5. doi:10.1148/radiology.188.1.7685531

49.

Pak

Znamirowski

Statum

Takahashi

Chung

, et al. Magic angle effect in magnetic resonance imaging of the Achilles tendon and enthesis. Magn Reson Imaging. 2009;27(4):557-64. doi:10.1016/j.mri.2008.09.003

50.

Jerban

Szeverenyi

Guo

Namiranian

, et al. Ultrashort echo time MRI (UTE-MRI) quantifications of cortical bone varied significantly at body temperature compared with room temperature. Investig Magn Reson Imaging. 2019;23(3):202-9.

51.

Wan

Zhao

Jerban

Searleman

Carl

, et al. Fast quantitative 3D ultrashort echo time MRI of cortical bone using extended cones sampling. Magn Reson Med. 2019;82(1_suppl):225-36.

52.

Otazo

Kim

Axel

Sodickson

DK.

Combination of compressed sensing and parallel imaging for highly accelerated first-pass cardiac perfusion MRI. Magn Reson Med. 2010;64(3):767-76.