Three-dimensional reconstructed magnetic resonance scans: Accuracy in identifying and defining knee meniscal tears

Contralateral MRI scan can be used reliably for three-dimensional meniscus sizing - Retrospective analysis of 160 healthy menisci

BACKGROUND

Meniscus allograft transplantation is a valuable surgical option for post-meniscectomy syndrome. For best results, the selected allograft should be as similar as possible to the original meniscus. Three-dimensional meniscus sizing could be a new approach to improve the accuracy of meniscus matching. The contralateral anatomy might therefore be a suitable reconstruction template. The purpose of this study was to compare the three-dimensional shape of the right and left menisci by bi-planar segmentation of magnetic resonance imaging (MRI) scans.

METHODS

Three-dimensional surface models of healthy menisci were created based on 40 bilateral MRI scans. Manual segmentation was performed on the MRI data in sagittal and coronal planes. For side-to-side comparison, each left meniscus model was mirrored and then superimposed to its corresponding right meniscus model. Differences between the meniscus pairs were assessed by width, length, height and surface distances. Inter-reader reliability, as well as accuracy of bi-planar segmentation was assessed by two different readers.

RESULTS

The meniscus pairs were not significantly different in terms of width, length and height (P = at least 0.138). Side difference of mean surface distances was 0.76 mm (±0.13 standard deviation (SD)) for medial and 0.78 mm (±0.15 SD) for lateral menisci. Inter-reader reliability was good to excellent (0.828-0.987).

CONCLUSION

The three-dimensional shapes of the left and right menisci are very similar. Therefore, the contralateral meniscus can be used as a template for three-dimensional meniscus allograft sizing. Three-dimensional meniscus segmentation and sizing can be performed accurately by combination of sagittal and coronal planes.

3D-MRI combined with signal-to-noise ratio measurement can improve the diagnostic accuracy and sensitivity in evaluating meniscal healing status after meniscal repair

PURPOSE

This study compared the diagnostic accuracy of second-look arthroscopy, clinical assessment, and magnetic resonance imaging (MRI) in evaluating meniscal healing following repair. To explore the application of 3D-MRI sequences with signal-to-noise ratio (SNR) measurements and the analysis of sensitivity and accuracy in evaluating meniscal healing status after meniscal repair.

METHODS

This study included 75 patients (75 menisci; mean age, 25.4 ± 6.6 years) who underwent arthroscopic meniscal repair (medial menisci, 45; and lateral menisci, 29). Follow-up evaluation included clinical assessment, magnetic resonance imaging (MRI), and second-look arthroscopy (mean follow-up, 10.2 ± 3.4 months). The criteria defined for unhealed menisci were: on arthroscopy, presence of cleft; on clinical assessment, presence of joint-line tenderness, swelling, locking, or positive McMurray test; on MRI, grade III signal intensity at the site of repair. Finally, the SNR on 3D-MRI of the healed and the unhealed groups were compared.

RESULTS

On second-look arthroscopy, 68 (90.7%) menisci were found to be completely healed. With the arthroscopic healing rate as the standard, the clinical healing rate was 69.3% (63 menisci) [sensitivity, 85.7%; specificity, 75.0%; accuracy, 76.0%; positive predictive value (PPV), 26.1%; negative predictive value (NPV), 98.1%]. Similarly, the sensitivity, specificity, accuracy, PPV and NPV, respectively, for MRI in four sequences were as follows: 2D-T2W-Sag: 71.4, 79.4, 78.7, 26.3, and 96.4%; 2D-T2W-Cor: 71.4, 82.4, 81.3, 29.4, and 96.6%; 3D-PDW-Sag: 100.0, 50.0, 54.7, 17.1, and 100.0%; 3D-PDW-Cor: 100.0, 58.8, 62.7, 20.0, and 100.0%. The mean SNRs of the healed group (3D-PDW-Sag, 66.7 ± 11.9; 3D-PDW-Cor, 63.9 ± 12.4) were significantly lesser than that of the unhealed group (3D-PDW-Sag, 89.2 ± 9.6; 3D-PDW-Cor, 82.2 ± 18.5) (P < 0.05). Identifying grade 3 meniscal damage on 3D-PDW images prior to applying the different SNR cut-off points revealed SNR values 80 in the sagittal plane and 70 in the coronal plane as having better diagnostic accuracy and sensitivity. The clinical relevance of the study was that 3D-MRI combined with SNR measurement may be a noninvasive and accurate method of assessment clinically, and a reliable alternative to second-look arthroscopy.

LEVEL OF EVIDENCE

III.