Healthy knee cartilage T2 mapping assessed with GRAPPATINI and multi-echo spin-echo (MESE) at 3T: regional values, longitudinal repeatability and method agreement

Introduction

Magnetic resonance imaging (MRI), radiological scoring systems, and quantitative image analysis have recently provided new insights into articular cartilage and related tissues under both healthy and pathological conditions. However, detecting early degenerative changes in cartilage, before actual tissue loss occurs, remains a challenge. Conventional MRI sequences show morphological changes, but cartilage loss is a nonspecific marker of osteoarthritis (OA) (1). Loss of cartilage is a well-established predictor of clinically relevant outcomes such as knee replacement; however, studies on its association with knee pain have produced inconsistent and sometimes contradictory findings (2).

Advances in MRI techniques have enhanced the ability to detect not only morphological but also macromolecular changes in cartilage, leading to the development of standardized scoring systems for OA evaluation (3). These techniques are based on articular cartilage extracellular matrix (ECM) composition and structural changes. In early disease stages, massive loss of proteoglycans (PGs) from the articular cartilage ECM increases the proportion and mobility of free water molecules; in later stages, progressive disorganization of collagen fibers, reduced water content, and glycosaminoglycan (GAG) depletion become more prominent (4).

MRI techniques such as T1 rho mapping, T2 mapping and delayed Gadolinium Enhanced MRI of Cartilage (dGEMRIC), were shown to be able to probe these changes. T1 rho relaxation time is sensitive to changes in cartilage biochemical composition, and in particular to variations in PG content. A decrease in PG concentration has been shown to correlate with altered T1 rho values. However, T1 rho contrast is not exclusively driven by PG depletion; it is also influenced by additional factors such as collagen fiber orientation and the concentration of other macromolecules. These microstructural and compositional changes are characteristic of the early stages of OA (5-7). T2 relaxation time has been shown to be associated with both cartilage water content and the organization of the collagen matrix. In particular, increased T2 relaxation times have been reported in regions of cartilage damage, reflecting disruption and loss of collagen matrix integrity (5). T2 mapping is widely used because it is sensitive to PG, collagen, and water content. T2 primarily reflects collagen fiber organization and tissue hydration. Numerous studies have shown that T2 relaxation times correlate with early biochemical changes in cartilage, particularly alterations in the collagen-PG matrix. Importantly, T2 values are more responsive to collagen fiber disorganization than to collagen content itself (8-11). T2 mapping can be easily implemented in clinical routine MRI. The duration of the sequences at 3T is approximately 6 min and does not require the use of contrast agent. However, T2 relaxation times may differ significantly among healthy individuals; in the context of cartilage repair, it is therefore important and well accepted to consider T2 values of adjacent healthy hyaline cartilage as a reference when comparing with repaired tissue. A further limitation is the variability in reproducibility across different MRI systems and sequences, which complicates its routine use in longitudinal monitoring (12).

Recently, more advanced techniques for T2 mapping have been developed. In particular, the GRAPPATINI sequence—which combines generalized autocalibrating partially parallel acquisition with model-based accelerated relaxometry using iterative nonlinear inversion (13) has been applied in several contexts, ranging from the brain to peripheral nerves (14-19). This sequence has been shown to provide similar reproducibility in T2 relaxation time estimates compared to multi-echo spin-echo (MESE), however, in shorter acquisition times. Moreover, it enables the generation of synthetic T2-weighted contrasts, eliminating the need for separate acquisitions and thereby saving additional scan time (20). However, the reliability of T2 relaxation time measurements across different cartilage areas has not yet been evaluated.

The aim of this study was to evaluate T2 values in different areas of healthy cartilage using both a conventional commercial T2 mapping sequence and the advanced accelerated GRAPPATINI sequence, and to assess their reliability in a test-retest design. Based on these results, we provide recommendations regarding the most suitable sequence for clinical and research applications. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2462/rc).

Methods

Patients

The patients included in this study were part of a prospective, multi-center national trial on autologous cartilage repair. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by Cantonal Research Ethics Commission for Human Research of Geneva (Basec-ID: 2023-02168) and informed consent was taken from all individual participants. For the present work, a single-site sub-analysis was performed: all data used in this study were collected exclusively at the Geneva University Hospitals.

All patients underwent a baseline MRI, followed by post-graft imaging at 3, 6, and 12 months as part of the national study protocol. At inclusion, patients reported clinical pain and functional impairment in the affected knee, as assessed by the KOOS (Knee Injury and Osteoarthritis Outcome Score), aside from the symptoms related to the treated cartilage lesion and standardized post-surgical rehabilitation. No new symptoms were detected in the healthy areas of the knees evaluated in this study.

Imaging protocol

Images were acquired using a 3T scanner (MAGNETOM Prisma Fit, Siemens Healthineers, Forchheim, Germany) with a 15-channel transmit/receive knee coil.

The MRI protocol included conventional anatomical 2D and 3D turbo spin echo (TSE) sequences as well as the T2 mapping MESE sequence and a research sequence for T2 mapping called GRAPPATINI. GRAPPATINI is a MESE sequence as well, however, it is 6-fold accelerated by combining parallel imaging and model-based image reconstruction (13). Notably, GRAPPATINI used a sampling strategy that combined block sampling with parallel imaging, in which every second k-space line was skipped. A detailed description and validation of this sampling pattern can be found in the original GRAPPATINI publication (13) as well as in the publication of its predecessor (21). The 6-fold acceleration provided by GRAPPATINI was not only used to reduce the acquisition time but also to increase the in-plane resolution from 0.8 to 0.4 mm and to extend the repetition time, which improved the signal-to-noise ratio. As a result, the present study compares protocols that reflect clinical practice rather than strictly harmonized sequences. Direct comparisons using identical sequence parameters, including simulations, phantom measurements, and in vivo experiments with both retrospective and prospective undersampling, were performed in the original publication (13). Both T2 mapping sequences were acquired in the same sagittal orientation, during the same MRI examination, without patient repositioning. Both sequences provide inline, vendor-integrated image reconstruction at the scanner console within a clinically acceptable time frame, allowing immediate image availability. Detailed imaging parameters are summarized in Table 1.

Image analysis

All data were anonymized and blinded prior to analysis. Both MESE and GRAPPATINI excluded the first echo from the T2-fitting procedure to mitigate stimulated-echo effects originating from B₁⁺ inhomogeneities and from the slice profile of the refocusing pulses.

T2 relaxation time measurements were performed using dedicated software (Cartilage assessment module, Advanced Visualisation Workspace v15.0, Philips Medical Systems, Best, The Netherlands). All analyses were conducted by a senior musculoskeletal radiologist with 20 years of experience. Regions of interest (ROIs) were manually drawn (as shown in Figure 1) in three anatomical locations: the medial femoral condyle, lateral femoral condyle, and patella. ROIs were placed in morphologically normal cartilage, as determined by conventional anatomical sequences, and specifically outside the repaired regions. Only these healthy areas were analyzed to establish reference T2 values and assess longitudinal repeatability. Specifically, the ROI on the lateral condyle was placed at the level of the first slice showing the tibiofibular joint; the ROI on the medial condyle was placed at the level of the recurrent semimembranosus tendon; and the patellar ROI was placed on the central axial slice.

Figure 1 Example of T2 mapping acquired with MESE and GRAPPATINI, along with the corresponding PD TSE sagittal slices for three different ROIs: lateral condyle, medial condyle, and patella. Zoomed views are provided for better ROI visualization. T2 maps are scaled from 1 ms (red) to 81 ms (dark blue). Note the presence of stripe artefacts in the trabecular bone on the MESE sequence, which do not appear to affect cartilage mapping. MESE, multi-echo spin-echo; PD TSE, proton density-weighted turbo spin echo; ROI, region of interest.

Each ROI was delineated by defining the bone-cartilage interface and the cartilage surface, after which the ROI was automatically segmented into three radial zones and three cartilage layers (deep, intermediate, and superficial). The values from the three radial zones were averaged to obtain a single mean T2 value for each cartilage layer within each anatomical area.

For the test-retest reliability analysis, the first two consecutive visits of patients were used as repeated measurements as no physiological change is expected in normal areas of cartilage during this time period (the interval between visit 1 and 2 ranged from 3 to 6 months).

The signal-to-noise ratio (SNR) of the resulting T2 maps was evaluated for both sequences. SNR was defined as the mean T2 value measured within each ROI divided by the corresponding standard deviation in the same ROI. A global SNR estimate was obtained by averaging these values across all areas and ROIs for MESE and GRAPPATINI, respectively.

Statistical analysis

All statistics were performed using R (version 4.5.0; R Foundation for Statistical Computing, Vienna, Austria). Given the hierarchical structure of the data—repeated measurements from the same patients with variable numbers of visits—a linear mixed-effects model (LMM) was employed to compare T2 values across cartilage areas (lateral condyle, medial condyle, and patella), cartilage depths (superficial, intermediate, deep), and acquisition methods (MESE vs. GRAPPATINI).

The model was specified as:

Model <- lmer(Mean ~ Method * Depth + (1|Patient), data)

Here, Mean refers to the measured T2 relaxation time, Method to the acquisition technique (MESE or GRAPPATINI), Depth to the cartilage layer, Method and Depth were included as fixed effects along with their interaction. Patient was included as a random intercept to account for within-subject correlation arising from repeated measurements. The analysis was performed separately for each anatomical area. The lmer function from the lme4 package was used to fit the model. Model parameters were estimated using restricted maximum likelihood (REML), which is the default estimation method in lme4, as well as all the other default settings of lmer function.

An analysis of variance (ANOVA) was conducted on the fitted model, followed by post hoc tests to compare T2 values between methods and across cartilage depths.

To assess agreement between the two T2 mapping techniques, Bland-Altman plots were generated. For test-retest reliability, Pearson correlation coefficients were calculated between visit 1 and visit 2, along with linear regression equations modelling the relationship for each method.

Additionally, the intraclass correlation coefficient (ICC) for absolute agreement based on single measurements [ICC (A,1)] was computed for each method and anatomical area. ICC values were interpreted according to the guidelines proposed by Cicchetti (22).

Results

Patient population

Nine patients, each undergoing between one and four MRI examinations, were included in the study between June 2023 and April 2025. Table 2 provides detailed information on the number of visits per patient and the time intervals between them.

Not all patients have both T2 mapping sequences at every time point, as the research prototype sequence (GRAPPATINI) was an optional addition to the standard imaging protocol, depending on remaining time left in the clinical exam slot, and was therefore not consistently acquired. Accordingly, timepoints for which the GRAPPATINI sequence was missing were omitted from the analysis to maintain consistency and avoid statistical bias arising from incomplete sequence pairs. In total, 22 MRI examinations were obtained, each acquired during the same session and including both MESE and GRAPPATINI sequences. From each examination, three ROIs were delineated, with three depth measurements recorded per ROI, resulting in a total of 198 distinct comparison points.

The mean age of the patients was 38 years (range, 30–48 years), and the mean body mass index (BMI) was 24 kg/m² (range, 21–28 kg/m²). The cohort included 5 left knees and 4 right knees, with a sex distribution of 6 males and 3 females.

T2 measurements in different areas

An example of images obtained with the MESE sequence as well as with the GRAPPATINI method, is shown in Figure 1. Three slices are presented, illustrating the three areas of interest, lateral condyle, medial condyle and patella, together with zoomed ROI to visualize the segmentation into three radial zones and three cartilage layers (deep, intermediate, and superficial). T2 relaxation times for each anatomical area (lateral condyle, medial condyle, and patella) are presented in Figure 2 for both acquisition methods (MESE and GRAPPATINI). Corresponding mean values and standard deviations are summarized in Table 3.

Figure 2 T2 relaxation time for each area and cartilage depth across all patients. MESE, multi-echo spin-echo.

A statistically significant difference was found between the deep cartilage layer and the intermediate and superficial layers, regardless of the imaging method used.

The SNR was assessed for both sequences to investigate its potential impact on measurement variability. The calculated SNR was 8.64 for MESE and 11.81 for GRAPPATINI, demonstrating that GRAPPATINI achieves a higher SNR.

Comparison of MESE and GRAPPATINI sequences

Overall, no significant difference in T2 values was observed between the MESE and GRAPPATINI sequences, except in the superficial layer of the lateral femoral condyle, where a significant difference was noted. Notably, greater variability (larger standard deviations) in T2 measurements was observed with the MESE method in the lateral condyle and patella, whereas this effect was not seen in the medial condyle. The Bland-Altman analysis in Figure 3 demonstrated no systematic bias between methods [mean difference:−0.26 ms; with limits of agreement of 15.7 ms (mean + 2SD) and −16.2 ms (mean − 2SD)].

Figure 3 Bland-Altman plot comparing MESE and GRAPPATINI methods for each measure. Each point represents a single measurement from one patient at a specific anatomical area and cartilage depth. MESE, multi-echo spin-echo; SD, standard deviation.

Repeatability study

Test-retest reliability was evaluated by comparing T2 values from the first and second visits of the same patients. A stronger correlation was observed with the GRAPPATINI method (R=0.73) compared to the MESE method (R=0.53), as shown in Figure 4. The slope of the regression line for GRAPPATINI was 0.91, indicating a high level of reproducibility between visits, whereas the MESE method yielded a lower slope of 0.56. Bland-Altman plots comparing the two visits for GRAPPATINI and MESE showed no systematic bias. The mean difference was −0.88 ms for MESE and −0.96 ms for GRAPPATINI, indicating good reproducibility between visits. The limits of agreement ranged from –15.3 to 13.6 ms for MESE and were slightly narrower for GRAPPATINI, from −13.6 to 11.7 ms, suggesting better agreement for this method.

Figure 4 Correlation between T2 measurements obtained for 2 successive visits of the same patient to test reproducibility. Corresponding Bland-Altman plots for MESE and GRAPPATINI methods. Each point represents a single measurement from one patient at a specific anatomical area and cartilage depth. MESE, multi-echo spin-echo; SD, standard deviation.

The ICC for absolute agreement between visits are shown in Table 4. It indicated good reliability for the MESE method (ICC =0.61) and excellent reliability for the GRAPPATINI method (ICC =0.75). When considering all anatomical areas and cartilage depths combined, ICC values were consistently higher for GRAPPATINI than for MESE.

When analysed by anatomical areas, GRAPPATINI demonstrated superior agreement in the 3 areas showing good reliability (ICC >0.60). While MESE showed slightly better agreement between visits in the lateral femoral condyle (ICC =0.67) compared to GRAPPATINI (ICC =0.62), reliability was poor for medial condyle (ICC <0.4) and only fair for patella (ICC between 0.4 and 0.59).

Discussion

T2 measurement in different areas

T2 mapping is a non-invasive imaging technique suitable for clinical applications that does not require contrast administration. It allows for the detection of alterations in the ECM, as well as changes in collagen content and orientation, before morphological cartilage abnormalities become apparent (23,24). Previous studies have reported good diagnostic accuracy in distinguishing healthy cartilage from early degenerative changes (25).

Normative T2 relaxation times were assessed in three cartilage areas: the lateral and medial femoral condyles and the patella. Mean T2 values ranged from 30.1 to 45.2 ms with MESE, and from 27.8 to 47.7 ms with GRAPPATINI. No significant differences in T2 values were observed between MESE and GRAPPATINI sequences, except in the superficial layer of the lateral femoral condyle, where the two methods yielded significantly different values. Bland-Altman analysis revealed no systematic bias between methods (mean difference: −0.26 ms) and relatively wide limits of agreement (±16 ms), suggesting that although the methods are generally comparable, measurement variability may influence precision in certain areas. Previous findings by Roux et al. demonstrated that GRAPPATINI method exhibited lower mean bias and smaller limits of agreement than MESE when compared to reference single-shot spin-echo sequence (20). Compared to MESE, GRAPPATINI offers the advantage of sampling more echo times (10 vs. 5), producing images with higher spatial resolution (0.4×0.4 vs. 0.8×0.8 mm) while reducing acquisition time (4 min 45 s vs. 7 min 24 s). Higher SNR was measured in GRAPPATINI sequence; this improvement in image quality may explain the smaller standard deviations observed for GRAPPATINI compared to MESE.

Our findings are consistent with prior work, demonstrating median T2 values in healthy knees of approximately 50 ms, in agreement with the observations of Soellner et al. (26) and Bae et al. (27). In a large cohort study of healthy volunteers (n=481, mean age 52 years, mean BMI 26 kg/m²), T2 relaxation times were shown to vary across different areas of the knee, with values comparable to our findings when taken for mean age 40 years: 34 ms in the lateral femoral condyle, 37 ms in the medial femoral condyle, and 31 ms in the patella. Notably, when measurements were repeated two years later, a small but significant increase in T2 was observed in the femoral and medial condyles (45 and 50 ms, respectively), while no significant change was detected in the patella (28).

In the study by Niu et al., no significant differences in T2 values were observed between early OA and healthy controls in the lateral compartment (femoral and tibial areas), in contrast to the medial and patellotrochlear compartments, where significant differences were detected (25). Median T2 values for healthy volunteers in that study were 47.0 ms for the medial femoral condyle, 54.2 ms for the lateral condyle, and 46.8 ms for the patella. These results are consistent with our findings, which also showed lower T2 values in the patella compared with the lateral and medial condyles. Additionally, an earlier investigation employing the Whole-Organ Magnetic Resonance Imaging Score (WORMS) system found T2 values ranging from 40.3 to 50.5 ms across ten subregions (29).

Knee malalignment has also been associated with T2 changes: Zhu et al. reported significantly higher T2 values in the lateral anterior femoral cartilage of healthy knees with varus or valgus alignment (approximately 52 ms) (30). Similarly, Zhao et al. reported elevated T2 values in the superficial lateral femoral cartilage (49–55 ms) and in the deep medial patellar cartilage (32–34 ms) in participants with mild OA symptoms (31). Although these values are associated with pathological conditions and are higher than our results in the medial and lateral femoral condyles, they fall within the same range as our patellar findings. This underscores the importance of establishing baseline healthy T2 values using identical MRI techniques to accurately detect pathological changes.

It is widely accepted that early changes in cartilage are induced by metabolic disturbances in chondrocytes and loss of PGs, leading to cartilage matrix degradation. These changes typically propagate from the superficial cartilage to the deeper layers (32,33). However, other researchers suggest that alterations originate from the osteochondral junction, which is the first site affected. This damage triggers a cascade of biological mechanisms, resulting in chondrocyte regression and ECM degradation, associated with increased bone porosity and vascularization (34). Recent studies using ultrashort echo time (UTE) sequences with T2* mapping have enabled improved visualization of the osteochondral junction (27,35). T2* values increase from the osteochondral junction to the superficial cartilage layers in both human and animal studies, and a strong correlation has been observed between T2* mapping and cartilage degeneration assessed by the MRI Osteoarthritis Knee Score (MOAKS) (36,37).

Cartilage degradation is typically characterized by increased T2 and decreased T2* values, with superficial cartilage showing higher values than deeper layers. T2 mapping values also differ across knee areas and are influenced by factors such as age (24,38). However, relatively few studies have examined variations across cartilage layers in morphologically normal cartilage. In our study, we observed higher T2 values in the intermediate and superficial layers compared with the deep layer in all three areas assessed. These findings are consistent with Tao et al., who reported higher T2 values in the superficial layer than in the deep layer of both the medial and lateral femoral condyles (39). Similarly, Chen et al. demonstrated significantly higher T2 values in the superficial layer of the patella compared with the deep layer (43 vs. 32 ms, P<0.001) (8).

Depth-dependent cartilage analysis was performed using a clinically implemented segmentation tool that provides a standardized three-layer stratification. This stratification is intended to support clinically relevant depth-dependent assessment rather than to represent histological cartilage zones. The same segmentation protocol and layer definitions were consistently applied to both MESE and GRAPPATINI acquisitions to ensure comparability. Although the coarser in-plane resolution of MESE may introduce partial-volume effects, both sequences demonstrated similar depth-dependent trends.

Comparison between MESE and GRAPPATINI sequences

It has already been shown that T2 values can vary substantially between centres using different vendors and sequences, although the repeatability of each method individually remains good (12). Previous work has also demonstrated that T2 mapping with GRAPPATINI is accurate compared with conventional T2 mapping in phantom studies, but this had not been confirmed in patient populations (20). In this study, we provide a direct comparison with the conventionally used MESE sequence for T2 mapping and show that GRAPPATINI yields less variability in T2 measurements without introducing systematic bias. Bland-Altman analysis demonstrated a small mean difference between MESE and GRAPPATINI T2 measurements (bias =0.26 ms). Although this indicates a measurable systematic offset, its magnitude is minimal relative to the physiological range of cartilage T2 values (approximately 20–60 ms) and is unlikely to be of clinical relevance. This small bias therefore does not meaningfully affect the agreement between the two methods.

Repeatability study

Overall, the agreement between T2 measurements acquired at two consecutive visits ranged from moderate to good. Among the evaluated areas, the lateral femoral condyle showed the highest intra-method consistency, indicating that it may be the most reliable site for robust T2 quantification. The GRAPPATINI technique demonstrated good agreement (ICC >0.60) across all three anatomical areas (lateral condyle, medial condyle, and patella), whereas the MESE method did not consistently reach this threshold in all areas.

Regarding longitudinal reproducibility, a previous study involving four healthy knees, phantom measurements, and data from multiple vendors reported excellent reproducibility for both mean T2 values and per-patient averages, supporting the use of T2 mapping for monitoring intra-subject changes over time (12).

Another study using the MESE sequence at 1.5T, demonstrated globally good reproducibility of T2 relaxation time over time in different areas of the knee with an ICC being globally higher in the patella (range, 0.61–0.98) than in lateral and medial condyle as we observed in our data (40).

Limitations

This study was designed to evaluate the feasibility, reproducibility, and quantitative agreement of the GRAPPATINI sequence compared with the reference MESE sequence under controlled conditions. Consequently, only cartilage regions without known pathological alterations were included, in order to minimize confounding factors and ensure robust assessment of measurement consistency. As a result, the diagnostic performance of GRAPPATINI for the detection and characterization of early cartilage lesions could not be assessed, which limits conclusions regarding its applicability in pathological scenarios. Further investigations including patients with varying degrees of cartilage degeneration are necessary to determine diagnostic sensitivity, specificity, and clinical utility. The delay between two consecutive visits was not the same for all patients (minimum 3 months, maximum 6 months). We hypothesize that normal cartilage would not undergo any change between visits, however, this variation, combined with differences in patient recovery from graft surgery, may have introduced an additional bias. Although age-related and BMI variations in T2 values have been reported, we did not include age and BMI in our model due to the restricted range (age range: 30 to 48 years, mean 38 years; BMI range: 21 to 28 kg/m², mean 24 kg/m²). These population characteristics represent a limitation of our study; future studies including larger and more diverse cohorts will be necessary to confirm these results and assess their generalizability. An additional limitation is that MESE and GRAPPATINI were implemented using different acquisition parameters, including spatial resolution and fat suppression, reflecting clinically realistic protocol designs for each technique rather than strictly matched settings; as a result, the observed agreement and performance may partly reflect parameter-dependent effects rather than intrinsic methodological differences alone. In particular, the absence of fat saturation in the MESE acquisition could lead to fat signal overlapping with cartilage, resulting in biased T2 values. This effect should be mitigated in the GRAPPATINI sequence, assuming effective fat saturation, which may explain some of the differences observed between the two tested protocols. T2 mapping is a technique sensitive to the magic angle effect, particularly for measurements acquired with short echo times (4). However, we believe that our results are representative of the clinical application of these methods, and that comparisons between methods were performed within the same ROIs, which should not compromise the conclusions. Finally, histological validation, which represents the reference gold standard, was not performed in this study. Such validation would require invasive tissue sampling and was therefore not feasible within the adopted population and study design. The aim of the present work was not to redefine T2 threshold values or to establish direct histopathological correlations, but rather to evaluate whether the GRAPPATINI sequence provides T2 measurements comparable to those obtained with the reference MESE sequence under identical acquisition conditions.

Conclusions

In conclusion, it is well established that T2 mapping is an effective tool for characterizing the knee, particularly for detecting early cartilage pathologies. Our study demonstrated that, although both the commercially available MESE sequence and the advanced research GRAPPATINI sequence showed no systematic bias, GRAPPATINI exhibited better reproducibility and narrower limits of agreement than MESE for T2 relaxation times in knee cartilage with a reduced acquisition time. Importantly, T2 values vary markedly across cartilage depth, highlighting the need to consider layer-specific measurements in quantitative assessments.

Acknowledgments

We acknowledge Vanarix SA, study sponsor (Dr. Vannary Tieng) and the clinical trial managers: Ms. Laura Bengtsson and Ms. Ilse Jonker for the data used in this study as a part of a clinical trial with Advanced Therapeutic Medicinal Product (ATMP) - Allogeneic Cartibeads clinical trial (Swissmedic ID: 701779_Allogeneic_Cartibead, BASEC ID: 2023-02168). We also acknowledge the CIBM Center for Biomedical Imaging for providing resources to conduct this study, in particular Sébastien Courvoisier and Frédéric Grouiller for the protocol setup at the console and for providing support during image acquisitions.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2462/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2462/dss

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2462/coif). T.H. is a current employee of Siemens Healthineers. J.M. is a current employee of Hirslanden Clinic La Colline. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by Cantonal Research Ethics Commission for Human Research of Geneva (Basec-ID: 2023-02168) and informed consent was taken from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Davis KW. MR imaging of the knee. Magn Reson Imaging Clin N Am 2014;22:xv. [Crossref] [PubMed]
2. Pan F, Laslett L, Tian J, Cicuttini F, Winzenberg T, Ding C, Jones G. Association Between Pain at Sites Outside the Knee and Knee Cartilage Volume Loss in Elderly People Without Knee Osteoarthritis: A Prospective Study. Arthritis Care Res (Hoboken) 2017;69:659-66. [Crossref] [PubMed]
3. Bruno MA, Mosher TJ, Gold GE. Arthritis in Color: Advanced Imaging of Arthritis. AJNR Am J Neuroradiol 2010;31:R1.
4. Matzat SJ, van Tiel J, Gold GE, Oei EH. Quantitative MRI techniques of cartilage composition. Quant Imaging Med Surg 2013;3:162-74. [Crossref] [PubMed]
5. Banjar M, Horiuchi S, Gedeon DN, Yoshioka H. Review of Quantitative Knee Articular Cartilage MR Imaging. Magn Reson Med Sci 2022;21:29-40. [Crossref] [PubMed]
6. Piccolo CL, Mallio CA, Vaccarino F, Grasso RF, Zobel BB. Imaging of knee osteoarthritis: a review of multimodal diagnostic approach. Quant Imaging Med Surg 2023;13:7582-95. [Crossref] [PubMed]
7. Mosher TJ. Quantitative Cartilage T2 and T1rho Mapping: Is There a Clinical Role? From the AJR Special Series on Quantitative Imaging. AJR Am J Roentgenol 2025;224:e2431655. [Crossref] [PubMed]
8. Chen M, Qiu L, Shen S, Wang F, Zhang J, Zhang C, Liu S. The influences of walking, running and stair activity on knee articular cartilage: Quantitative MRI using T1 rho and T2 mapping. PLoS One 2017;12:e0187008. [Crossref] [PubMed]
9. Kester BS, Carpenter PM, Yu HJ, Nozaki T, Kaneko Y, Yoshioka H, Schwarzkopf R. T1ρ/T2 mapping and histopathology of degenerative cartilage in advanced knee osteoarthritis. World J Orthop 2017;8:350-6. [Crossref] [PubMed]
10. Årøen A, Brøgger H, Røtterud JH, Sivertsen EA, Engebretsen L, Risberg MA. Evaluation of focal cartilage lesions of the knee using MRI T2 mapping and delayed Gadolinium Enhanced MRI of Cartilage (dGEMRIC). BMC Musculoskelet Disord 2016;17:73. [Crossref] [PubMed]
11. Trattnig S, Burstein D, Szomolanyi P, Pinker K, Welsch GH, Mamisch TC. T1(Gd) gives comparable information as Delta T1 relaxation rate in dGEMRIC evaluation of cartilage repair tissue. Invest Radiol 2009;44:598-602. [Crossref] [PubMed]
12. Verschueren J, Eijgenraam SM, Klein S, Poot DHJ, Bierma-Zeinstra SMA, Hernandez Tamames JA, Wielopolski PA, Reijman M, Oei EHG T. (2) mapping of healthy knee cartilage: multicenter multivendor reproducibility. Quant Imaging Med Surg 2021;11:1247-55. [Crossref] [PubMed]
13. Hilbert T, Sumpf TJ, Weiland E, Frahm J, Thiran JP, Meuli R, Kober T, Krueger G. Accelerated T2 mapping combining parallel MRI and model-based reconstruction: GRAPPATINI. J Magn Reson Imaging 2018;48:359-368. [Crossref] [PubMed]
14. Xie Y, Cai J, Li Q, Kober T, Hilbert T, Chen S, Tao H. To evaluate repaired rotator cuff morphologically and quantitatively in one GRAPPATINI MRI examination. Acta Radiol 2025;66:351-8. [Crossref] [PubMed]
15. Wang F, Zhang H, Wu C, Wang Q, Hou B, Sun Y, Kober T, Hilbert T, Zhang Y, Zeng X, Jin Z. Quantitative T2 mapping accelerated by GRAPPATINI for evaluation of muscles in patients with myositis. Br J Radiol 2019;92:20190109. [Crossref] [PubMed]
16. Raudner M, Schreiner M, Hilbert T, Kober T, Weber M, Windhager R, Trattnig S, Juras V. Accelerated T2 Mapping of the Lumbar Intervertebral Disc: Highly Undersampled K-Space Data for Robust T2 Relaxation Time Measurement in Clinically Feasible Acquisition Times. Invest Radiol 2020;55:695-701. [Crossref] [PubMed]
17. Gruenebach N, Abello Mercado MA, Grauhan NF, Sanner A, Kronfeld A, Groppa S, Schoeffling VI, Hilbert T, Brockmann MA, Othman AE. Clinical feasibility and validation of the accelerated T2 mapping sequence GRAPPATINI in brain imaging. Heliyon 2023;9:e15064. [Crossref] [PubMed]
18. Draveny R, Ambarki K, Han F, Hilbert T, Laurent V, Morel O, Bertholdt C, Beaumont M. Comparison of T2 Quantification Strategies in the Abdominal-Pelvic Region for Clinical Use. Invest Radiol 2022;57:412-21. [Crossref] [PubMed]
19. Lu AZ, Lin Y, Hilbert T, Igbinoba Z, Shakoor D, Cai X, Sneag DB, Tan ET. Feasibility of Qualitative Evaluation and Quantitative T2 Mapping of Peripheral Nerves and Muscles Using GRAPPATINI. J Magn Reson Imaging 2025;62:781-9. [Crossref] [PubMed]
20. Roux M, Hilbert T, Hussami M, Becce F, Kober T, Omoumi P. MRI T2 Mapping of the Knee Providing Synthetic Morphologic Images: Comparison to Conventional Turbo Spin-Echo MRI. Radiology 2019;293:620-30. [Crossref] [PubMed]
21. Sumpf TJ, Uecker M, Boretius S, Frahm J. Model-based nonlinear inverse reconstruction for T2 mapping using highly undersampled spin-echo MRI. J Magn Reson Imaging 2011;34:420-8. [Crossref] [PubMed]
22. Cicchetti DV. Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychol Assess 1994;6:284-90.
23. Wu Z, Zaylor W, Sommer S, Xie D, Zhong X, Liu K, Kim J, Beveridge JE, Zhang X, Li X. Assessment of ultrashort echo time (UTE) T(2)* mapping at 3T for the whole knee: repeatability, the effects of fat suppression, and knee position. Quant Imaging Med Surg 2023;13:7893-909. [Crossref] [PubMed]
24. Pan J, Pialat JB, Joseph T, Kuo D, Joseph GB, Nevitt MC, Link TM. Knee cartilage T2 characteristics and evolution in relation to morphologic abnormalities detected at 3-T MR imaging: a longitudinal study of the normal control cohort from the Osteoarthritis Initiative. Radiology 2011;261:507-15. [Crossref] [PubMed]
25. Niu F, Meng X, Zhu N, Sun M, Shen Z, Han J, Zhao T, Hu Y, Wang Z. Exploring the asynchronous changes of articular cartilage and osteochondral junction in knee osteoarthritis based on T2 mapping and ultrashort echo time imaging: a pilot study. Quant Imaging Med Surg 2025;15:4414-30. [Crossref] [PubMed]
26. Soellner ST, Goldmann A, Muelheims D, Welsch GH, Pachowsky ML. Intraoperative validation of quantitative T2 mapping in patients with articular cartilage lesions of the knee. Osteoarthritis Cartilage 2017;25:1841-9. [Crossref] [PubMed]
27. Bae WC, Biswas R, Chen K, Chang EY, Chung CB. UTE MRI of the Osteochondral Junction. Curr Radiol Rep 2014;2:35. [Crossref] [PubMed]
28. Joseph GB, McCulloch CE, Nevitt MC, Heilmeier U, Nardo L, Lynch JA, Liu F, Baum T, Link TM. A reference database of cartilage 3 T MRI T2 values in knees without diagnostic evidence of cartilage degeneration: data from the osteoarthritis initiative. Osteoarthritis Cartilage 2015;23:897-905. [Crossref] [PubMed]
29. Na H, Yan Z, Min C, Zhi-hong W, Wei M. T2 mapping of articular cartilage in the healthy knee using a magnetic resonance segmentation. Zhongguo Yi Xue Ke Xue Yuan Xue Bao 2011;33:66-70. [Crossref] [PubMed]
30. Zhu J, Hu N, Liang X, Li X, Guan J, Wang Y, Wang L. T2 mapping of cartilage and menisci at 3T in healthy subjects with knee malalignment: initial experience. Skeletal Radiol 2019;48:753-63. [Crossref] [PubMed]
31. Zhao H, Li H, Liang S, Wang X, Yang F. T2 mapping for knee cartilage degeneration in young patients with mild symptoms. BMC Med Imaging 2022;22:72. [Crossref] [PubMed]
32. Arokoski JP, Jurvelin JS, Väätäinen U, Helminen HJ. Normal and pathological adaptations of articular cartilage to joint loading. Scand J Med Sci Sports 2000;10:186-98. [Crossref] [PubMed]
33. Hollander AP, Pidoux I, Reiner A, Rorabeck C, Bourne R, Poole AR. Damage to type II collagen in aging and osteoarthritis starts at the articular surface, originates around chondrocytes, and extends into the cartilage with progressive degeneration. J Clin Invest 1995;96:2859-69. [Crossref] [PubMed]
34. Radin EL, Rose RM. Role of subchondral bone in the initiation and progression of cartilage damage. Clin Orthop Relat Res 1986;34-40.
35. Mackay JW, Low SB, Houston GC, Toms AP. Ultrashort TE evaluation of the osteochondral junction in vivo: a feasibility study. Br J Radiol 2016;89:20150493. [Crossref] [PubMed]
36. Mahar R, Batool S, Badar F, Xia Y. Quantitative measurement of T2, T1ρ and T1 relaxation times in articular cartilage and cartilage-bone interface by SE and UTE imaging at microscopic resolution. J Magn Reson 2018;297:76-85. [Crossref] [PubMed]
37. Shao H, Chang EY, Pauli C, Zanganeh S, Bae W, Chung CB, Tang G, Du J. UTE bi-component analysis of T2* relaxation in articular cartilage. Osteoarthritis Cartilage 2016;24:364-73. [Crossref] [PubMed]
38. Mosher TJ, Dardzinski BJ. Cartilage MRI T2 relaxation time mapping: overview and applications. Semin Musculoskelet Radiol 2004;8:355-68. [Crossref] [PubMed]
39. Tao H, Qiao Y, Hu Y, Xie Y, Lu R, Yan X, Chen S. Quantitative T2-Mapping and T2(*)-Mapping Evaluation of Changes in Cartilage Matrix after Acute Anterior Cruciate Ligament Rupture and the Correlation between the Results of Both Methods. Biomed Res Int 2018;2018:7985672. [Crossref] [PubMed]
40. Hannila I, Lammentausta E, Tervonen O, Nieminen MT. The repeatability of T2 relaxation time measurement of human knee articular cartilage. MAGMA 2015;28:547-53. [Crossref] [PubMed]