MR Imaging of the Knee at 3T - Diagnostic Performance and Comparison with 1.5T

Datum: 19 november 2013

Locatie: Universitair Ziekenhuis Antwerpen - Auditorium Kinsbergen - Wilrijkstraat 10 - 2650 Edegem

Tijdstip: 16.30 uur

Organisatie / co-organisatie: Faculty of Medicine and Health Sciences

Promovendus: Pieter Van Dyck

Promotor: Prof. P. Parizel, Prof. J. Gielen

Korte beschrijving: PhD defence Pieter Van Dyck - Faculty of Medicine and Health Sciences

Abstract: Until recently, the standard magnet strength for routine knee MRI has been 1.5T. It is expected that, with the advent of 3T systems, MR diagnoses will be more accurate compared to 1.5T. However, this expectation is not evidence-based. This thesis bundles several papers regarding the diagnostic performance of knee MRI obtained at 1.5 and 3T.

Firstly, we determined the accuracy of MRI in the diagnosis of meniscal and anterior cruciate ligament (ACL) tears at 1.5 and 3-T, in different patient groups, and analyzed the causes of diagnostic error, compared to arthroscopy. In our studies, we found comparable accuracy rates with similar causes of errors for 1.5 and 3T MRI.
Secondly, we have provided a prospective comparison between 1.5 and 3T MRI of the knee, in the same individuals, to assess the theoretical improvement in performance of 3T MRI. In our study, we could not demonstrate a significant improvement in accuracy of 3T compared to 1.5T MRI. Our study findings disagree with the general assumption that higher field strength automatically increases the reliability of the MR report in the detection of knee pathologies.
Thirdly, we investigated the value of the newly developed 3D turbo spin-echo (TSE) acquisition as compared to conventional 2D TSE acquisition for knee joint assessment at 3T. In our study, we found no clear advantage of the 3D TSE acquisition.

Finally, we examined whether different clinically and arthroscopically confirmed ACL pathologies have distinctive preoperative findings on MRI. We found that the use of MRI is insufficient to assess the functional status of the ACL, mainly because of the many patterns of partial tears and the frequent similarity of partial tears to complete tears or even to mucoid degeneration of the ACL.
In general, we conclude that a routine 3T protocol does not significantly improve diagnostic performance of knee MRI compared to a 1.5T protocol. Apparently, many other factors affect diagnostic performance of knee MRI, including study design and population characteristics, and definition of disease criteria. Further studies are needed to determine which field strength is optimal for routine knee imaging.