The misdiagnosis rate of anterior cruciate ligament (ACL) tear of the knee joint is as high as 47% due to the confusion of high signal generated by synovial hyperplasia and partial volume effect on the standard MRI sequences [8]. A study revealed the evaluation of ACL tear by standard MRI sequences with additional oblique-sagittal plane showed high sensitivity and specificity [18, 15]. Furthermore, another study using two additional oblique sequences (sagittal and coronal) to evaluate the ACL tear concluded that the use of oblique coronal and oblique sagittal MRI of the knee improved the diagnostic accuracy of the ACL tear [19]. Therefore, we performed this study to analyze the diagnostic accuracy by adding a new modified dual precision positioning of thin-slice oblique sagittal fat suppression proton-density weighted imaging (DPP-TSO-Sag-FS-PDWI) to standard MRI sequences. The diagnostic consistency of the new modified DPP-TSO-Sag-FS-PDWI sequence and arthroscopy was better than standard MRI sequences for both reader 1 (K = 0.876 vs. 0.620) and reader 2 (K = 0.833 vs. 0.683) with good diagnostic repeatability (K = 0.794 vs. 0.598). Compared to standard sequences, the new modified DPP-TSO-Sag-FS-PDWI sequence can accurately classify and correctly diagnose the grades of ACL injury, whose sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV) were all greater than 84%, especially for grade II injury (partial tear) as the PPV value was superior for the reader 1 (92.3% vs. 53.9%) and reader 2 (84.6% vs. 69.2%).
Previous studies have used standard MRI sequences with or without the additional sequence to diagnose an ACL tear. A previous study using 3.0-T MRI with standard sequences revealed that the specificity and accuracy of the partial injury (97%, 95%) and complete injury (99%, 97%), which are similar to present study. However, the sensitivity of the partial injury (77%) was lower than our study (85.7% or 84.6%) [20]. This indicates that although 3.0-T MRI is a highly accurate method for the diagnosis of ACL injuries, the new modified DPP-TSO-Sag-FS-PDWI sequence on a 1.5-T scanner with a single channel knee coil can achieve similar effects and further improve the diagnostic sensitivity to partial tears. There was a study of 149 patients with three different methods, including standard MRI sequences, patient knee in 17-degree flexion MRI technique, and additional sagittal-oblique sequence, and it revealed the sagittal-oblique sequence more sensitive (p < 0.001) than standard MRI sequences. The author mentioned the oblique-sagittal technique can display the partial tear because of the double angulation which is following the course of the patient’s ligament due to the approximate orientation of the external rotation of the foot [15]. The study using two additional oblique sequences (sagittal and coronal) to evaluate the ACL tear concluded that the usage of oblique coronal and oblique sagittal MRI of the knee improved the diagnostic accuracy of the ACL tear [19]. Furthermore, another study revealed the evaluation of ACL tear by standard MRI sequences with additional oblique-sagittal plane showed high sensitivity and specificity compared to a standard MRI protocol [18]. The results in these studies are similar to ours and we further found that our new modified DPP-TSO-Sag-FS-PDWI sequence can maximize the acquisition of ACL complete and true anatomical images and exclude the influence of anatomical differences between individuals, which is helpful to improve the authenticity and accuracy of ACL scanning images and to improve the diagnostic accuracy of ACL injury with its three grades.
To our knowledge, this is the first study to display and diagnose the ACL injury with its three grades by using a dual precision positioning based on axial and coronal localization. Our study not only validated the diagnostic accuracy of standard sequences used to diagnose ACL injury and its grades but also added a new modified dual precision positioning of thin-slice oblique sagittal sequence to improve the diagnostic accuracy. Therefore, the unnecessary arthroscopic surgery can be prevented and the financial burden can be decreased [9]. Additionally, our scanning method is simple, the scanning time is moderate, and there is no special requirement for the MR scanner and the patient's position. Therefore, it is very suitable for popularization and application in clinical work.
The present study has several limitations. Firstly, the sample size was small which may be caused bias. Secondly, the new modified DPP-TSO-Sag-FS-PDWI sequence cannot identify meniscus injury simultaneously, which is caused by a narrow scanning range. Lastly, our MR scanner and knee coils were relatively backward, and given the time required to scan and patient tolerance, our sequence was compared only with standard sequences, not with 3D isotropic or other sequences. In the future, we will also report the related study by using this sequence to distinguish the anteromedial bundle and posterolateral bundle of the ACL.