Multi-vendor multi-site quantitative MRI analysis of cartilage degeneration 10 Years after anterior cruciate ligament reconstruction: MOON-MRI protocol and preliminary results

Whole knee joint mapping using a phase modulated UTE adiabatic T1ρ (PM-UTE-AdiabT1ρ ) sequence

PURPOSE

To develop a 3D phase modulated UTE adiabatic T1ρ (PM-UTE-AdiabT1ρ ) sequence for whole knee joint mapping on a clinical 3 T scanner.

METHODS

This new sequence includes six major features: (1) a magnetization reset module, (2) a train of adiabatic full passage pulses for spin locking, (3) a phase modulation scheme (i.e., RF cycling pair), (4) a fat saturation module, (5) a variable flip angle scheme, and (6) a 3D UTE Cones sequence for data acquisition. A simple exponential fitting was used for T1ρ quantification. Phantom studies were performed to investigate PM-UTE-AdiabT1ρ 's sensitivity to compositional changes and reproducibility as well as its correlation with continuous wave-T1ρ measurement. The PM-UTE-AdiabT1ρ technique was then applied to five ex vivo and five in vivo normal knees to measure T1ρ values of femoral cartilage, meniscus, posterior cruciate ligament, anterior cruciate ligament, patellar tendon, and muscle.

RESULTS

The phantom study demonstrated PM-UTE-AdiabT1ρ 's high sensitivity to compositional changes, its high reproducibility, and its strong linear correlation with continuous wave-T1ρ measurement. The ex vivo and in vivo knee studies demonstrated average T1ρ values of 105.6 ± 8.4 and 77.9 ± 3.9 ms for the femoral cartilage, 39.2 ± 5.1 and 30.1 ± 2.2 ms for the meniscus, 51.6 ± 5.3 and 29.2 ± 2.4 ms for the posterior cruciate ligament, 79.0 ± 9.3 and 52.0 ± 3.1 ms for the anterior cruciate ligament, 19.8 ± 4.5 and 17.0 ± 1.8 ms for the patellar tendon, and 91.1 ± 8.8 and 57.6 ± 2.8 ms for the muscle, respectively.

CONCLUSION

The 3D PM-UTE-AdiabT1ρ sequence allows volumetric T1ρ assessment for both short and long T2 tissues in the knee joint on a clinical 3 T scanner.

Deep learning-based automatic pipeline for quantitative assessment of thigh muscle morphology and fatty infiltration

PURPOSE

Fast and accurate thigh muscle segmentation from MRI is important for quantitative assessment of thigh muscle morphology and composition. A novel deep learning (DL) based thigh muscle and surrounding tissues segmentation model was developed for fully automatic and reproducible cross-sectional area (CSA) and fat fraction (FF) quantification and tested in patients at 10 years after anterior cruciate ligament reconstructions.

METHODS

A DL model combining UNet and DenseNet was trained and tested using manually segmented thighs from 16 patients (32 legs). Segmentation accuracy was evaluated using Dice similarity coefficients (DSC) and average symmetric surface distance (ASSD). A UNet model was trained for comparison. These segmentations were used to obtain CSA and FF quantification. Reproducibility of CSA and FF quantification was tested with scan and rescan of six healthy subjects.

RESULTS

The proposed UNet and DenseNet had high agreement with manual segmentation (DSC >0.97, ASSD < 0.24) and improved performance compared with UNet. For hamstrings of the operated knee, the automated pipeline had largest absolute difference of 6.01% for CSA and 0.47% for FF as compared to manual segmentation. In reproducibility analysis, the average difference (absolute) in CSA quantification between scan and rescan was better for the automatic method as compared with manual segmentation (2.27% vs. 3.34%), whereas the average difference (absolute) in FF quantification were similar.

CONCLUSIONS

The proposed method exhibits excellent accuracy and reproducibility in CSA and FF quantification compared with manual segmentation and can be used in large-scale patient studies.