Inter- and intra-observer reliability of patellar height measurements in patients with and without patellar instability on plain radiographs and magnetic resonance imaging

MRI analysis of the physiological patellofemoral joint morphology of adult knees

OBJECTIVE

The aim of the present study was to determine physiological reference values for the morphology of the patella and to analyse these parameters according to patella position in healthy knee joints.

MATERIAL AND METHODS

Healthy knee joints of 409 patients (mean age, 52.3 years [± 16.8]) were analysed retrospectively on MRI images for Insall-Salvati index (ISI), sagittal patella thickness (PTS) and patella length (PLS) as well as axial patella thickness (PTA) and patella width (PWA). Differences between patellar diameters were analysed depending on ISI, side, age and gender.

RESULTS

Mean PTS was 20.1 mm (± 2.4), PLS 44.0 mm (± 4.4), PTA 21.8 mm (± 2.4) and PWA 44.5 mm (± 4.7). Depending on the vertical patellar position (ISI), all patellar parameters (p < 0.01) showed significant differences between patients with a patella alta, norma and baja. In general, a smaller ISI showed higher measured values for the patellar parameters. There were no significant differences for the laterality. Only PTS showed a significant age difference (p = 0.031). All parameters were significantly larger in male compared to female knees (p < 0.001).

CONCLUSION

Reference parameters for the patella morphology are reported. Concluding from the results, a relationship between vertical patellar position and patellar morphology seems to exist. This finding should be taken into account in diagnostics and therapy of patella disorders.

An evaluation of a novel method for the MRI-based assessment of Caton-Deschamps index in the Knee

INTRODUCTION

The radiographical assessment of patella height has historically been performed using X-Ray. The aim of this study was to evaluate a new method for the assessment of patella height using MRI and to assess the correlation with the X-Ray based assessment.

MATERIALS AND METHODS

159 patients who had both lateral radiographs and MRI images were included. Parameters measured included traditional radiographical CDI, MRI-based CDI, and TT-TG distance. On the basis of the TT-TG, the patients were divided into 2 groups. Two different methods were used to assess CDI using MRI: using a single slice image, and an alternative technique using two different cross-sectional images. The correlation of the two measurement methods was assessed using Pearson's correlation coefficient. The intraclass correlation coefficient (ICC) was determined from the measurements of the two investigators.

RESULTS

The average TT-TG distance was 11.6 mm (± 4.6). In patients with a TT-TG < 15 mm, both measurement methods showed comparable correlation with measurements on X-Ray. In patients with a TT-TG of > 15 the the new cross-sectional imaging method showed higher correlation with traditional X-Ray assessment compared to CDI assessment using the traditional single slice method (r = 0.594, p < 0.001 vs. r = 0.302, p = 0.055).

CONCLUSIONS

The assessment of CDI on MRI using a cross-sectional imaging method has a better correlation with traditional X-Ray assessment of CDI than single-slice assessment. This is particularly true in patients with elevated TT-TG and as such should be preferentially used in the assessment of Patellar height in this cohort.

Patella height measurements in patients with patellofemoral instability: the power to predict patella height indices

INTRODUCTION

Numerous indices for patella height measurement are described in the literature; however, there is not a universally accepted 'gold standard'. Choice of indices varies depending on personal preference and previous experience, making it difficult to compare results between studies. The aims of this study were to assess the correlation between different patella height indices and generate statistically derived formulae that can be used to predict the value of one patella height ratio when another is known.

MATERIALS AND METHODS

Patients with patellofemoral instability were radiologically evaluated using both X-ray and Magnetic Resonance Imaging (MRI) of the knee. Two observers measured six validated patella height indices at two separate time intervals. All measurements were aggregated together and a mean average for each patella height ratio was calculated. These results were then statistically analysed using Pearson product-moment correlation and linear regression tests.

RESULTS

Forty-four patients were included in the study. A statistically significant correlation was found between the majority of patella height ratios. On X-ray, the strongest correlation (r = 0.92, p < 0.001) and regression (R2 = 0.85, p < 0.001) coefficients were between the Blackburne-Peel and Caton-Dechamps ratios, with the formula; CD = 0.20 + (0.95xBP). On MRI, the strongest correlation (r = - 0.84, p < 0.001) and regression (R2 = 0.71, p < 0.001) coefficients were between the Patellophyseal index and Patellotrochlear index, with the formula; PP = 1.02-(1.10xPT).

CONCLUSION

The formulae produced in this study can predict the value of one patella height ratio when the results of another are known. This could facilitate systematic reviews and meta-analyses by allowing the pooling together of data from many studies.

Automated patellar height assessment on high-resolution radiographs with a novel deep learning-based approach

BACKGROUND

Artificial intelligence and deep learning have shown promising results in medical imaging and interpreting radiographs. Moreover, medical community shows a gaining interest in automating routine diagnostics issues and orthopedic measurements.

AIM

To verify the accuracy of automated patellar height assessment using deep learning-based bone segmentation and detection approach on high resolution radiographs.

METHODS

218 Lateral knee radiographs were included in the analysis. 82 radiographs were utilized for training and 10 other radiographs for validation of a U-Net neural network to achieve required Dice score. 92 other radiographs were used for automatic (U-Net) and manual measurements of the patellar height, quantified by Caton-Deschamps (CD) and Blackburne-Peel (BP) indexes. The detection of required bones regions on high-resolution images was done using a You Only Look Once (YOLO) neural network. The agreement between manual and automatic measurements was calculated using the interclass correlation coefficient (ICC) and the standard error for single measurement (SEM). To check U-Net's generalization the segmentation accuracy on the test set was also calculated.

RESULTS

Proximal tibia and patella was segmented with accuracy 95.9% (Dice score) by U-Net neural network on lateral knee subimages automatically detected by the YOLO network (mean Average Precision mAP greater than 0.96). The mean values of CD and BP indexes calculated by orthopedic surgeons (R#1 and R#2) was 0.93 (± 0.19) and 0.89 (± 0.19) for CD and 0.80 (± 0.17) and 0.78 (± 0.17) for BP. Automatic measurements performed by our algorithm for CD and BP indexes were 0.92 (± 0.21) and 0.75 (± 0.19), respectively. Excellent agreement between the orthopedic surgeons’ measurements and results of the algorithm has been achieved (ICC > 0.75, SEM < 0.014).

CONCLUSION

Automatic patellar height assessment can be achieved on high-resolution radiographs with the required accuracy. Determining patellar end-points and the joint line-fitting to the proximal tibia joint surface allows for accurate CD and BP index calculations. The obtained results indicate that this approach can be valuable tool in a medical practice.

Magnetic resonance imaging overestimates patellar height compared with radiographs

PURPOSE

To evaluate the inter-observer and inter-method reliability for patellar height measurements between conventional radiographs (CR) and magnetic resonance imaging (MRI) using one or two slices.

METHODS

This was a reliability study, with 60 patients divided in two groups: 30 patients with patellar instability (patella group) and 30 patients with anterior cruciate ligament or meniscus injury (control group). CR and MRI were evaluated by two independent observers. Insall-Salvati index (IS) and Caton-Deschamps index (CD) were measured using three different methods: CR, one-slice MRI or two-slice MRI. Intra-class correlation coefficients (ICC) were calculated for inter-observer reliability and inter-method reliability. Bland-Altman agreement was also calculated.

RESULTS

The inter-observer reliability was very good for the IS with ICCs of 0.93, 0.84 and 0.82, for the CR, one-slice MRI and two-slice MRI, respectively. Similarly, for the CD the ICCs were good, 0.76, 0.80 and 0.75 for the CR, one-slice MRI and two-slice MRI, respectively. No differences were found between the patella and the control group. The inter-method analysis results were: ICCs for IS (0.83, 0.86, 0.93) and CD (0.72, 0.82, 0.83), for the comparisons of CR/one-slice MR, CR/two-slice MRI and one-slice MRI/two-slice MRI, respectively. The Bland-Altman mean differences showed an 8% and a 7% increase on IS values with one-slice MRI and two-slice MRI compared to CR results, while the increase was of 9% and 1% in CD for the respective comparisons with CR.

CONCLUSION

MRI can overestimate patellar height compared to CR, as much as an 8% increase in Insall-Salvati values when using one- or two-slice MRI measurements, and up to a 9% increase in Caton-Deschamps value when using the one-slice MRI method. It is recommended to use the CR as the preferred method when measuring patellar height.

LEVEL OF EVIDENCE

III.