Abstract
Several studies have investigated the relationship between the thickness of the posterior meniscofemoral ligament (pMFL) and the presence of a discoid meniscus. We investigated the correlation between meniscal pathology and anatomic features of pMFL such as attachment type, thickness, and volume. We retrospectively evaluated 191 patients who underwent knee MRI. MR images were reviewed to assess the attachment type of the pMFL on the medial femoral condyle (high vs. low), the thickness of the pMFL, and the presence of a meniscal tear or a discoid meniscus. The pMFL volume was quantified by using three-dimensional (3D) segmentation software. The relationship between the frequency of medial or lateral meniscal tear and anatomic features of pMFL were analyzed using Chi-square, Fisher’s exact, or Mann–Whitney U test. High type pMFLs had significantly greater thickness and volume than low type pMFLs (p < 0.001). Patients with degenerative lateral meniscal tear had significantly higher thickness and volume of the pMFL than patients with intact lateral meniscus (p < 0.05). The pMFL thickness and volume were not significantly related to traumatic lateral meniscal tear, medial meniscal tear, and discoid meniscus. High type pMFLs tended to be thicker and larger than low type pMFLs and higher thickness and volume of the pMFL was significantly related to the degenerative lateral meniscal tear. However, the attachment type of the pMFL itself was not significantly related to the lateral meniscal tear as well as the medial meniscal tear.
Supplementary Information
The online version contains supplementary material available at 10.1007/s10278-022-00651-6.
Keywords: Wrisberg ligament, High location, Meniscofemoral ligament, Discoid lateral meniscus
Introduction
The meniscofemoral ligament (MFL) extends from the posterior horn of the lateral meniscus to the lateral aspect of the medial femoral condyle [1]. It is divided into two ligaments according to the relationship with the posterior cruciate ligament (PCL). The anterior meniscofemoral ligament (aMFL) runs anterior to the PCL and is also known as the ligament of Humphrey, while the posterior meniscofemoral ligament (pMFL) runs posterior to the PCL and is known as the ligament of Wrisberg [2–4]. MFL plays some role in posterior movement of the lateral meniscus during tibial flexion and inversion [5]. Among these two ligaments, it has also been suggested that the pMFL somewhat protects the dorsal horn of the lateral meniscus [6]. Years of knee arthroscopy experience at our institution have led to the orthopedic surgeons realizing that the pMFL can be classified into two different types according to where it attaches to the medial femoral condyle, which can be determined by either arthroscopy, knee MRI, or both. If the pMFL attaches to the lower half surface of the medial femoral condyle, we refer to it as a ‘low type’ pMFL. If it attaches to the upper half surface of the medial femoral condyle, we refer to it as a ‘high type’ pMFL. Radiologists can also use this classification if these attachments are noticeable on knee MRI. Several studies have investigated the relationship between the thickness of the pMFL and the presence of a discoid meniscus. Kim et al. reported that patients with complete discoid lateral meniscus had thicker and high-riding pMFLs than patients with non-discoid lateral meniscus [7]. On the other hand, Lee et al. examined the knee MRI data of patients with arthroscopically confirmed lateral meniscal tears, and found no relationship between the presence of the pMFL and lateral meniscal tears [4]. More recently, Ahn et al. reported that a high and thick pMFL is related to the posterocentral shift type discoid lateral meniscus based on peripheral detachment [8]. Our researchers hypothesized, based on experience, that there is a difference in the incidence of meniscal tears and attachment type of the pMFL, and suggested that there may also be a relationship between the presence of a discoid meniscus and the attachment of the pMFL. In addition, the anatomical differences in 3-dimensional (3D) volume as well as thickness of the pMFL between healthy subjects and those with meniscal pathology remain unclear. Our purpose in this study was to investigate the correlation between meniscal pathology and anatomic features of pMFL such as attachment type, thickness, and volume.
Methods
Case Selection
This study was approved by the institutional review board of our hospital, and because the study was retrospective, requirements for informed consent were waived. We retrospectively evaluated 344 consecutive patients who had undergone knee MRI because of knee pain from March 2016 to May 2016. Patients who had undergone previous surgical interventions such as anterior or posterior cruciate ligament reconstruction (51 cases), acute trauma including bone fracture (51 cases), or pyogenic infection (13 cases) were excluded. In addition, patients whose MR images did not show or had indistinct pMFLs (38 cases) were excluded. A total of 191 patients were therefore ultimately included in this study. The mean age of the study population was 45.3 ± 15.6 years (range, 4–80 years). Eighty-four (44%) patients were male and 107 (56%) patients were female.
Image Acquisition
MR images were acquired using a 3.0 T MR system (Achieva, Philips Medical Systems, Best, The Netherlands). We used an 8-channel coil dedicated to the knee joint. Detailed sequences and MR parameters are described in Table 1.
Table 1.
Parameters of the knee joint MR sequences
| Imaging Parameter | Coronal T1 1FSE | Coronal 2FS T2 FSE | Sagittal FS 3PD FSE | Sagittal T2 FSE | Axial FS T2 FSE |
|---|---|---|---|---|---|
| TR (msec) | 500–800 | 3500–4300 | 3000–4500 | 3500–4000 | 4000–5000 |
| TE (msec) | 15 | 60–100 | 30–40 | 60–100 | 60–100 |
| Flip Angle ( °) | 90 | 90 | 90 | 90 | 90 |
| Matrix Size | 350 × 350 | 350 × 300 | 320 × 240 | 350 × 280 | 320 × 230 |
| Field of view (cm) | 16 | 16 | 16 | 16 | 16 |
| Slice thickness (mm) | 3 | 3 | 3 | 3 | 3 |
| Inter-slice gap (mm) | 0.3 | 0.3 | 0.3 | 0.3 | 0.3 |
| Bandwidth (kHz) | 270 | 250 | 250 | 240 | 290 |
| Echo train length | 5 | 14 | 12 | 14 | 15 |
| Signal average | 1 | 1 | 1 | 1 | 1 |
| Scan time (min:s) | 2:40 | 2:20 | 2:30 | 2:30 | 2:30 |
FSE fast spin-echo, FS fat saturation, PD proton density
Image Analysis
MR images were analyzed by two trained and qualified radiologists (HJP and JNK) with 14 and 8 years of experience in musculoskeletal MR imaging, respectively. Both readers were unaware of the interpretations of the other, clinical findings, and arthroscopic results. They classified the attachment type of the pMFL (high vs. low) and measured the thickness of the pMFL independently. The evaluation of the aMFL was excluded from the study. First, they classified the site of attachment of the pMFL to the lateral aspect of the medial femoral condyle. They drew two horizontal lines starting from low and high portion of the lateral intercondylar surface of the medial femoral condyle on the coronal T1-weighted image of the knee joint at the slice of the femoral attach site of the pMFL (Fig. 1a). Another mid-line was drawn between two lines. If the pMFL was attached to the upper half surface of the medial femoral condyle (above this line), the pMFL was classified as a high type pMFL (Fig. 1a, b). If the pMFL was attached to the lower half surface of the medial femoral condyle, it was classified as a low type pMFL (Fig. 2a, b). Second, they independently measured the thickness (diameter) of the pMFL on the MR image. They used the measuring tool from the PACS module, and thickness was measured at the mid portion of pMFL from the sagittal T2-weighted image (Figs. 1c and 2c). Then they evaluated the presence of meniscal tear or discoid meniscus. Longitudinal-vertical tear, radial tear, or root tear was considered to be of traumatic origin, whereas horizontal tear was considered degenerative. Discoid meniscus was defined if the transverse width of lateral meniscus is greater than 14 mm on the coronal MR image at the mid-portion of a meniscus [9]. Final diagnosis of a meniscal tear was made based on radiologic findings and arthroscopic findings by consensus of the two radiologists. Arthroscopic confirmations were made on 50 cases (26%).
Fig. 1.
A high type pMFL based on the attachment site of the pMFL to the lateral aspect of the medial femoral condyle in a 36-year-old female with knee pain without trauma. a, b Coronal T1-weighted image of the right knee (TR/TE = 709/10). The pMFL is attached to the upper half of the medial femoral condyle, so it was classified as high type. c We measured the thickness of the pMFL on the MR image at the mid portion of the pMFL from the sagittal T2-weighted image (TR/TE = 3600/70)
Fig. 2.
Low type pMFL according to the attachment site of pMFL to the lateral aspect of the medial femoral condyle in a 66-year-old male with knee pain without trauma. a, b Coronal T1-weighted image of the right knee (TR/TE = 630/10). The pMFL is attached to the lower half of the medial femoral condyle, so it was classified as low type. c We measured the thickness of the pMFL on the MR image at the mid portion of the pMFL from the sagittal T2-weighted image (TR/TE = 4500/70)
Volumetric Segmentation
The open source software ITK-SNAP version 3.8.0 (http://www.itksnap.org) was used for segmentation of the pMFL volume. The MRI DICOM datasets were imported into ITK-SNAP and the pMFL volume was segmented by manually outlining the ligament in the sagittal T2-weighted image, using the polygon and paintbrush tools in ITK-SNAP. Segmentation was performed by a radiologist (MSK) with 4 years of experience in musculoskeletal imaging and familiar with the software. All contoured sagittal slices were utilized in the calculation of pMFL volume. An example of 3D-rendering made from a segmentation in ITK-SNAP is shown in Fig. 3.
Fig. 3.
Representative 3D rendering of pMFL (red), lateral meniscus (green), medial meniscus (blue), and bone (yellow) segmentations from sagittal images
Statistical Analysis
We evaluated if there was a significant difference in the incidence of meniscal tears and discoid menisci according to the type of pMFL using Chi-square test or Fisher’s exact test. Numerical data were tested for normality with Shapiro–Wilk test. We evaluated if there was a significant association between the presence of a meniscal tear or discoid meniscus and the size of the pMFL using the Mann–Whitney U test. Spearman’s correlation coefficients (R) were used to measure the association between thickness and volume of the pMFL. R between 0.5 and 0.7 was considered as moderate correlation, between 0.7 and 0.9 was considered as relatively high correlation, and > 0.9 was considered as very high correlation. Statistical analyses were performed using the statistical software package PASW version 18.0 (IBM, Armonk, NY, USA). A p value less than 0.05 was considered statistically significant.
Results
Classification of the pMFL Based on Its Attachment Site to the Femoral Condyle
Reader 1 classified 135 (70.7%) of the 191 patients as having a high type pMFL and 56 (29.3%) of the 191 patients as having a low type pMFL. Reader 2 classified 130 (68.1%) of the 191 patients as having a high type pMFL and 61 (31.9%) of the 191 patients as having a low type pMFL. Discrepancies between the two readers occurred for five cases. In these five cases, the femoral attachment sites of the pMFL were about the midpoint of the lateral aspect of the medial femoral condyle. For both readers, more pMFLs were classified as high type than as low type. Inter-observer agreement for classifying pMFLs as high or low type was almost perfect (κ = 0.941).
Correlation of Thickness and Volume of the pMFL
A relatively high correlation was found between the pMFL thickness for reader 1 and reader 2 and the pMFL volume with a Spearman’s correlation coefficient of R = 0.812 (p < 0.001) and R = 0.802 (p < 0.001), respectively. Two scatter plots for the measures were presented in Fig. 4.
Fig. 4.
Scatter plots show the presence of a relatively high correlation between the pMFL thickness and the pMFL volume: a pMFL thickness for reader 1 (R = 0.812, p < 0.001) and b pMFL thickness for reader 2 (R = 0.802, p < 0.001)
The Thickness and Volume of the pMFL and Its Relationship to Attachment Type
The median thickness of the pMFL was 2.69 mm (interquartile range 2.21–3.52) for reader 1 and 2.75 mm (IQR 2.22–3.48) for reader 2. The median thicknesses of each type of pMFL are shown in Table 2. The median thickness of the pMFL in the high type pMFL group was 2.95 mm (IQR 2.44–3.67) for reader 1 and 3.03 mm (IQR 2.38–3.65) for reader 2. In the low type pMFL group, the median thickness of the pMFL was 2.27 mm (IQR 2.00–2.69) for reader 1 and 2.35 mm (IQR 2.01–2.80) for reader 2. The high type pMFL was significantly thicker than the low type pMFL (p < 0.001). Similarly, the median volume of the high type pMFL was 177.9 mm3 (IQR 140.4–222.5) for reader 1 and 175.1 mm3 (IQR 140.3–224.7) for reader 2, which showed a significantly larger volume than the low type pMFL with median volume of 138.3 mm3 (IQR 114.9–171.1) for reader 1 and 142.4 mm3 (IQR 115.7–176) for reader 2 (p < 0.001, Table 2).
Table 2.
Median thickness, incidence of meniscal tears, and incidence of discoid meniscus according to the type of the pMFL
| Reader 1 | p | Reader 2 | p | |||
|---|---|---|---|---|---|---|
| High type (n = 135) | Low type (n = 56) | High type (n = 130) | Low type (n = 61) | |||
| Thickness of the pMFL (mm) |
2.95 (2.44–3.67) |
2.27 (2.00–2.69) |
< 0.001 |
3.03 (2.38–3.65) |
2.35 (2.01–2.80) |
< 0.001 |
| Volume of the pMFL (mm3) |
177.9 (140.4–222.5) |
138.3 (114.9–171.1) |
< 0.001 |
175.1 (140.3–224.7) |
142.4 (115.7–176) |
< 0.001 |
| Medial meniscus | ||||||
| Intact | 86 (63.7%) | 34 (60.7%) | 0.510 | 85 (65.4%) | 35 (57.4%) | 0.210 |
| Degenerative tear | 42 (31.1%) | 21 (37.5%) | 38 (29.2%) | 25 (41%) | ||
| Traumatic tear | 7 (5.2%) | 1 (1.8%) | 7 (5.4%) | 1 (1.6%) | ||
| Lateral meniscus | ||||||
| Intact | 96 (71.1%) | 42 (75%) | 0.860 | 92 (70.8%) | 46 (75.4%) | 0.746 |
| Degenerative tear | 34 (25.2%) | 13 (23.2%) | 33 (25.4%) | 14 (23%) | ||
| Traumatic tear | 5 (3.7%) | 1 (1.8%) | 5 (3.8%) | 1 (1.6%) | ||
| Discoid meniscus | 11 (8.1%) | 4 (7.1%) | 0.814 | 10 (7.7%) | 5 (8.2%) | 0.904 |
Data are presented as median (25th–75th percentile) and frequencies (%)
Statistically significant associations are demonstrated in bold
pMFL posterior meniscofemoral ligament
Incidence of Meniscal Pathology According to pMFL Location
The number of meniscal tears and discoid menisci in each group is summarized in Table 2. Seventy-one patients had a medial meniscal tear and 53 patients had a lateral meniscal tear. The incidence of medial meniscal tears in the high and low type MFL group were 36.3% (49/135) and 39.3% (22/56) for reader 1, respectively, and 34.6% (45/130) and 42.6% (26/61) for reader 2, respectively. The incidence of lateral meniscal tears in the high and low type MFL group were 28.9% (39/135) and 25.0% (14/56) for reader 1, respectively, and 29.2% (38/130) and 24.6% (15/61) for reader 2, respectively. There was no significant difference in the incidence of meniscal tears according to pMFL type for either reader.
The prevalence of a discoid meniscus was 7.9% among all patients (15/191). There was no significant relationship between pMFL attachment site and the incidence of discoid menisci for either reader.
The Thickness of the pMFL and Its Relationship to Meniscal Pathology
There was no significant difference in pMFL thickness between patients with medial meniscal tears and those with intact medial meniscus. For reader 1, patients with lateral meniscal tear had a greater thickness of pMFL than patients with intact lateral meniscus, but the difference was at the margin of statistical significance (p = 0.064). For reader 2, there was a significant difference in pMFL thickness between patients with lateral meniscal tears and those without lateral meniscal tears (p = 0.047). In particular, patients with degenerative lateral meniscal tear had a significantly higher thickness of pMFL than patients with intact lateral meniscus for both readers (p = 0.039 and p = 0.035, respectively). However, there was no significant difference in pMFL thickness between patients with traumatic lateral meniscal tears and those with intact lateral meniscus for either reader.
In addition, there was no significant difference in pMFL thickness between patients with a discoid meniscus and those without a discoid meniscus (Table 3).
Table 3.
Median thickness of the pMFL according to the meniscal tear or discoid meniscus
| Reader 1 | Reader 2 | |||
|---|---|---|---|---|
| Thickness of the pMFL (mm) | p | Thickness of the pMFL (mm) | p | |
| Medial meniscus | ||||
| Intact MM (A) (n = 120) | 2.69 (2.12–3.56) | 2.7 (2.2–3.52) | ||
| MM Tear (B) (n = 71) | 2.8 (2.28–3.38) | (A,B) 0.854 | 2.75 (2.3–3.37) | (A,B) 0.985 |
| Degenerative MM Tear (C) (n = 63) | 2.69 (2.25–3.44) | (A,C) 0.946 | 2.75 (2.29–3.37) | (A,C) 0.818 |
| Traumatic MM Tear (D) (n = 8) | 2.96 (2.86–3.08) | (A,D) 0.370 | 2.88 (2.68–3.44) | (A,D) 0.400 |
| Lateral meniscus | ||||
| Intact LM (A) (n = 138) | 2.67 (2.11–3.46) | 2.57 (2.16–3.42) | ||
| LM Tear (B) (n = 53) | 2.95 (2.28–3.56) | (A,B) 0.064 | 2.99 (2.44–3.55) | (A,B) 0.047 |
| Degenerative LM Tear (C) (n = 47) | 2.95 (2.52–3.6) | (A,C) 0.039 | 2.99 (2.48–3.56) | (A,C) 0.035 |
| Traumatic LM Tear (D) (n = 6) | 2.6 (2.25–3.05) | (A,D) 0.841 | 2.75 (2.31–3.25) | (A,D) 0.920 |
| Discoid Meniscus | ||||
| Discoid Meniscus (−) | 2.69 (2.24–3.49) | 0.967 | 2.75 (2.22–3.45) | 0.998 |
| Discoid Meniscus (+) | 2.95 (2.04–3.67) | 3.00 (2.00–3.67) | ||
Data are presented as median (25th–75th percentile)
Statistically significant associations are demonstrated in bold
pMFL posterior meniscofemoral ligament, MM medial meniscus, LM lateral meniscus
The Volume of the pMFL and Its Relationship to Meniscal Pathology
There was no significant difference in pMFL volume between patients with medial meniscal tears and those with intact medial meniscus. There was a significant difference in the volume of the pMFL between the lateral meniscal tear and intact group and between the degenerative lateral meniscal tear and intact group (p = 0.034 and p = 0.024, respectively). However, there was no significant difference in pMFL volume between traumatic lateral meniscal tear and intact group (p = 0.912). In addition, there was no significant difference in pMFL volume between patients with a discoid meniscus and those without a discoid meniscus (Table 4).
Table 4.
Median volume of the pMFL according to the meniscal tear or discoid meniscus
| Volume of the pMFL (mm3) | p | |
|---|---|---|
| Medial meniscus | ||
| Intact MM (A) (n = 120) | 161.8 (125.9–206.7) | |
| MM Tear (B) (n = 71) | 171.8 (136.7–210.3) | (A,B) 0.193 |
| Degenerative MM Tear (C) (n = 63) | 171.1 (134.1–210.3) | (A,C) 0.338 |
| Traumatic MM Tear (D) (n = 8) | 187 (169.1–224.9) | (A,D) 0.128 |
| Lateral Meniscus | ||
| Intact LM (A) (n = 138) | 160.2 (126.8–201.3) | |
| LM Tear (B) (n = 53) | 178.9 (142.1–225.3) | (A,B) 0.034 |
| Degenerative LM Tear (C) (n = 47) | 180.1 (145.2 – 229.0) | (A,C) 0.024 |
| Traumatic LM Tear (D) (n = 6) | 171.5 (112.8 – 191.0) | (A,D) 0.912 |
| Discoid Meniscus | ||
| Discoid Meniscus (−) | 167.1 (130.5–209.8) | 0.627 |
| Discoid Meniscus (+) | 168.5 (92.8–208.7) | |
Data are presented as median (25th–75th percentile)
Statistically significant associations are demonstrated in bold
pMFL posterior meniscofemoral ligament, MM medial meniscus, LM lateral meniscus
Discussion
Anterior and posterior MFLs reinforce the anterior and posterior attachments of the lateral meniscus, and the pMFL reduces anteroposterior laxity of the knee as it pulls the posterior horn of the lateral meniscus anteromedially during flexion [4]. Loss of the MFL results in a 10% increase in the peak contact pressure on the articular cartilage when the knee is loaded in axial compression [6]. Based on his experience, our hospital’s orthopedic surgeon suggested that there is a positive relationship between the incidence of meniscal tears and the attachment type of the pMFL. He suggested that the high type pMFL was more often associated with meniscal tears than the low type pMFL. In our study, however, the type of pMFL was not associated with lateral meniscal tear as well as medial meniscal tear. In addition, the type of pMFL was not associated with the presence of discoid lateral meniscus. Kim et al. reported that patients with complete discoid lateral meniscus had thicker and high-riding pMFL than those with non-discoid lateral meniscus [10]. Furthermore, they showed that in patients with complete discoid lateral meniscus, the lateral meniscus tear was seen more often with high-riding pMFL than low-riding pMFL. They suggested that a thicker pMFL with a higher location could cause wider excursion of the lateral meniscus and that this repetitive motion could increase the burden on the lateral meniscus and result in meniscal injury. However, there are some differences in their research method compared to our study. First, Kim et al.’s study included 109 patients with complete discoid lateral meniscus and another 100 patients with non-discoid lateral meniscus as control group [10]. In contrast, our study included consecutive patients who had knee pain and underwent knee MRI, and only 7.9% (15/191) of patients had discoid meniscus. The reported incidence of discoid meniscus ranges from 0.4 to 17% for the lateral meniscus [11, 12]. The prevalence of discoid meniscus in our study is similar to the literature. Fewer patients with discoid meniscus in our study may explain that the attachment type and size of the pMFL were not related to discoid meniscus. Second, there is a difference between Kim et al.’s criteria and our criteria for the location of the pMFL [10]. Kim et al. divided the types based on the relative positions of the PCL and pMFL, but we divided the types based on attachment site of the medial femoral condyle. In the Kim et al.’s study, 23.8% of the non-discoid meniscus group had high-riding pMFL. However, our study shows that incidence of high type pMFL was nearly 70% for both readers, although most patients had non-discoid meniscus. However, the type of pMFL used in Kim et al.’s study may be somewhat controversial in that the positional relationship between the PCL and pMFL may change according to the knee posture. Despite these differences, there is a common point between our study and Kim et al.’s study. Kim et al. reported that the insertion of the pMFL was not related to the lateral meniscus tear in the non-discoid meniscus group (p = 1.000). In our study, most patients had non-discoid meniscus, and there was no association between type of the pMFL and meniscal tear, which was similar to Kim et al.’s study. These findings mean that the location of pMFL itself is not significantly related to the lateral meniscal tear.
In our study, the high type pMFL had significantly greater thickness and volume than the low type pMFL. In addition, patients with degenerative lateral meniscal tear had significantly higher thickness and volume of the pMFL than patients with intact lateral meniscus. Although some studies have reported that thick and high location of the pMFL were associated with lateral meniscus pathology, there have been no reports about the relationship between the location and thickness of the pMFL [7, 8, 10]. Moreover, the relationship between the thickness of pMFL and lateral meniscal tear has not been reported, including patients with non-discoid lateral meniscus. In our study, the association between the size of the pMFL and the degenerative lateral meniscus tear may be related to the biomechanics and functions of the pMFL. According to the cadaveric biomechanical study, the degree of lateral meniscal displacement was correlated with the magnitude of tension of the pMFL [13]. Thick pMFL increases the magnitude of tension of pMFL, which in turn could increase the burden to the lateral meniscus and eventually cause meniscal tears. However, on the other hand, considering the function of pMFL, thick pMFL may be the result of the degenerative lateral meniscal tear, not the cause. The MFL functions as a stabilizer of the posterior horn of the lateral meniscus, which reduces tibiofemoral contact pressure [6, 14]. Degenerative meniscal tears occur spontaneously or occur from minor, fairly innocuous trauma over a relatively long period of time. The meniscus has an important role in load bearing and load transmission [15]. With these in mind, it is possible that degenerative lateral meniscal tears resulted in compensatory hypertrophy of the pMFL. Further studies are needed to determine the causal relationships between the pathology of the lateral meniscus and the size of the pMFL.
The thickness of pMFL in the mid-sagittal plane is often and easily used to estimate the size of the pMFL [10]. However, the single-slice measurement of the thickness may have limitations in evaluating quantitative parameters of pMFL. We measured the ligament volume through manual image segmentation with ITK-SNAP software, and a relatively high correlation was found between the pMFL thickness and volume (R > 0.8). However, manual segmentation requires the operator to select voxels corresponding to the ligament while moving slice-by-slice through the MR image stack, which is time-consuming and prone to observer variability. There have been several reports of efforts to develop automated segmentation methods for cartilage and meniscus in knee MRI [16, 17]. The knee ligaments are difficult to segment automatically due to the unclear tissue boundary relative to background tissue [18]. However, it has recently been reported that automated segmentation of the anterior cruciate ligament using convolutional neural network (U-Net)-based deep learning model achieved high speed, accuracy, and reliability [18, 19]. To support the findings of our study, the use of automated image processing technology would be more helpful in future research.
Our study had several limitations. First, we could not completely exclude the possibility of selection bias and confounding factors because this was a retrospective study. Furthermore, we only included symptomatic patients who had knee MRI data, which could also contribute to selection bias. Third, we did not correlate pMFL results to clinical outcomes. Fourth, the number of surgical confirmations of meniscus tear was low at 25%, in other cases diagnosed by MR imaging. However, this study did not focus on the accuracy of radiographic techniques or imaging signs, but focused on the relationship between the meniscofemoral ligament and meniscus pathology.
In conclusion, high type pMFLs tended to be thicker and larger than low type pMFLs and higher thickness and volume of the pMFL was significantly related to the degenerative lateral meniscal tear. However, the attachment type of the pMFL itself was not significantly related to the lateral meniscal tear as well as the medial meniscal tear. Further studies are needed on the causal relationships between the lateral meniscal lesion and the quantitative parameters of the pMFL.
Supplementary Information
Below is the link to the electronic supplementary material.
Declarations
Competing Interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Deckey DG, Tummala S, Verhey JT, Hassebrock JD, Dulle D, Miller MD, Chhabra A. Prevalence, biomechanics, and pathologies of the meniscofemoral ligaments: a systematic review. Arthrosc Sports Med Rehabil. 2021;3(6):e2093–e2101. doi: 10.1016/j.asmr.2021.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Röhrich S, Kainberger F, Hirtler L. Evaluation of age-dependent morphometrics of the meniscofemoral ligaments in reference to the posterior cruciate ligament in routine MRI. Eur Radiol. 2018;28(6):2369–2379. doi: 10.1007/s00330-017-5128-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Gupte CM, Smith A, McDermott ID, Bull AMJ, Thomas RD: Meniscofemoral ligaments revisited. Anatomical study, age correlation and clinical implications. J Bone Joint Surg Br 84(6):846–851, 2002 [DOI] [PubMed]
- 4.Lee BY, Jee WH, Kim JM, Kim BS, Choi KH. Incidence and significance of demonstrating the meniscofemoral ligament on MRI. Br J Radiol. 2000;73(867):271–274. doi: 10.1259/bjr.73.867.10817042. [DOI] [PubMed] [Google Scholar]
- 5.Bintoudi A, Natsis K. Tsitouridis I: Anterior and posterior meniscofemoral ligaments: MRI evaluation. Anat Res Int. 2012;2012:839724. doi: 10.1155/2012/839724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Amadi HO, Gupte CM, Lie DTT, McDermott ID, Amis AA, Bull AMJ. A biomechanical study of the meniscofemoral ligaments and their contribution to contact pressure reduction in the knee. Knee Surg Sports Traumatol Arthrosc. 2008;16(11):1004–1008. doi: 10.1007/s00167-008-0592-0. [DOI] [PubMed] [Google Scholar]
- 7.Kim EY, Choi SH, Ahn JH. Atypically thick and high location of the Wrisberg ligament in patients with a complete lateral discoid meniscus. Skeletal Radiol. 2008;37(9):827–833. doi: 10.1007/s00256-008-0524-6. [DOI] [PubMed] [Google Scholar]
- 8.Ahn JH, Wang JH, Kim DU, Lee DK, Kim JH. Does high location and thickness of the Wrisberg ligament affect discoid lateral meniscus tear type based on peripheral detachment? Knee. 2017;24(6):1350–1358. doi: 10.1016/j.knee.2017.07.012. [DOI] [PubMed] [Google Scholar]
- 9.Araki Y, Yamamoto H, Nakamura H, Tsukaguchi I. MR diagnosis of discoid lateral meniscus of the knee. Eur J Radiol. 1994;18(2):92–95. doi: 10.1016/0720-048X(94)90271-2. [DOI] [PubMed] [Google Scholar]
- 10.Kim JE, Choi SH. Is the location of the Wrisberg ligament related to frequent complete discoid lateral meniscus tear? Acta Radiol. 2010;51(10):1120–1125. doi: 10.3109/02841851.2010.520026. [DOI] [PubMed] [Google Scholar]
- 11.Greis PE, Bardana DD, Holstrom MC, Burks RT. Meniscal injury: basic science and evaluation. J Am Acad Orthop Surg. 2002;10(3):168–176. doi: 10.5435/00124635-200205000-00003. [DOI] [PubMed] [Google Scholar]
- 12.Kim JH, Ahn JH, Kim JH, Wang JH. Discoid lateral meniscus: importance, diagnosis, and treatment. J Exp Orthop. 2020;7(1):81. doi: 10.1186/s40634-020-00294-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Poynton A, Moran CJ, Moran R, O'Brien M. The meniscofemoral ligaments influence lateral meniscal motion at the human knee joint. Arthroscopy. 2011;27(3):365–371. doi: 10.1016/j.arthro.2010.07.025. [DOI] [PubMed] [Google Scholar]
- 14.Forkel P, Herbort M, Sprenker F, Metzlaff S, Raschke M, Petersen W. The biomechanical effect of a lateral meniscus posterior root tear with and without damage to the meniscofemoral ligament: efficacy of different repair techniques. Arthroscopy. 2014;30(7):833–840. doi: 10.1016/j.arthro.2014.02.040. [DOI] [PubMed] [Google Scholar]
- 15.McDermott I. Meniscal tears, repairs and replacement: their relevance to osteoarthritis of the knee. Br J Sports Med. 2011;45(4):292–297. doi: 10.1136/bjsm.2010.081257. [DOI] [PubMed] [Google Scholar]
- 16.Gaj S, Yang M, Nakamura K, Li X. Automated cartilage and meniscus segmentation of knee MRI with conditional generative adversarial networks. Magn Reson Med. 2019;84(1):437–449. doi: 10.1002/mrm.28111. [DOI] [PubMed] [Google Scholar]
- 17.Norman B, Pedoia V, Majumdar S. Use of 2D U-Net convolutional neural networks for automated cartilage and meniscus segmentation of knee MR imaging data to determine relaxometry and morphometry. Radiology. 2018;288(1):177–185. doi: 10.1148/radiol.2018172322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Flannery SW, Kiapour AM, Edgar DJ, Murray MM, Fleming BC. Automated magnetic resonance image segmentation of the anterior cruciate ligament. J Orthop Res. 2021;39(4):831–840. doi: 10.1002/jor.24926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Awan MJ, Rahim MSM, Salim N, Rehman A, Garcia-Zapirain B. Automated knee MR images segmentation of anterior cruciate ligament tears. Sensors. 2022;22(4):1552. doi: 10.3390/s22041552. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.