Diagnostic Accuracy of Ultrasound in the Evaluation of Collateral Ligaments and Meniscal Injuries of Knee in Comparison with Magnetic Resonance Imaging

Introduction

The diagnosis of knee joint disorders was a great challenge for radiologists of yesteryears, as they had access only to limited diagnostic tools like conventional radiographs, positive or negative contrast arthrography, angiography and traditional 2D-scintigraphy. These revealed nothing about the internal architecture, internal derangement, or disease activity, except for some indirect evidence to suggest an ongoing disease. Despite these limitations, they had done commendable work with a systematic approach to the disease. The introduction of planar imaging modalities starting from ultrasonography to today's hybrid tools like Positron emission tomography-computed tomography (PET-CT), Positron emission tomography–magnetic resonance imaging (PET-MRI) with their multiplanar capabilities and superb resolution brought a new era in the imaging of bone and joint disorders. The knee is the largest hinge type synovial joint formed by femoral condyles, tibial plateau, and patella.¹ Since these bones do not interlock, the stability of the joint is provided by the ligaments, tendons, menisci, and capsule.² During the bipedal locomotion, the knee is subjected to high weight-bearing. However, the complex arrangement of the intra and extracapsular ligaments helps the joint to counter the biomechanical demands imposed on the joint.¹

The knee joint is subjected to a variety of trauma especially in young adults and sportsmen which can result in various types of bony, meniscal, and ligamentous injuries that may warrant surgical management and rest for a long period. Imaging plays a key role in diagnosis, prognosis and treatment planning. Since the injury to the joint is recognized as a high-risk factor for the development of osteoarthrosis, early detection of trauma is very important.³

Radiographs and computed tomography are primarily used for diagnosing bone injuries, whereas ultrasound and MRI are used for assessing soft tissues.⁴ The disadvantages of the former are the use of ionizing radiation and limited soft-tissue resolution. Though MRI is considered gold standard in the current imaging of knee, ultrasound has a very significant role in proper evaluation of knee.⁵ Musculoskeletal Ultrasound (MSUS) machines with high-frequency (12 - 15 MHz) linear array transducers are used to evaluate superficial soft tissue structures such as muscle, tendon, ligament, and bursa since they provide excellent image resolution.⁶ The ultrasound has several unique strengths over MRI. It has higher spatial resolution than MR imaging, and allows a better evaluation of the superficial structures. The capability of dynamic imaging is crucial in differentiating a partial tear from complete tears of tendons and real-time comparison with the contralateral knee. USG is of utmost help in evaluating claustrophobic and otherwise MR contraindicated patients. Easy accessibility, portability, and low cost makes it an indispensable modality in the current imaging practice.⁵ The major disadvantage of ultrasound even with high resolution, high frequency equipment is its operator dependence, long learning curve and absolute requirement of trained radiologists. This study assessed the competence of ultrasound to detect ligament and meniscal injuries of the knee joint in comparison with MRI.

Materials and Methods

This observational study approved by the institutional ethical committee was conducted in the Department of Radiodiagnosis for a period of six months. In this study, a total of 110 patients were selected for USG examination using the inclusion criteria. The age of the patients ranged between 18 to 80 years. After getting the consent, the problematic knee was evaluated by USG with a high-frequency probe. The findings were recorded in a data sheet. Then the patient was evaluated using 1.5 T MRI and findings were recorded. The mean age and sex ratio were also calculated.

Equipment

USG machine: GE LOGIQ F6 using 7-12 MHz transducer, USA. MRI machine: GE SIGNA HDXT 1.5 T, USA.

Inclusion criteria

1. All patients with a history of knee trauma aged between 18 to 80 years.

2. Both inpatients and outpatients.

3. Males and females.

4. All patients who provided consent for the study.

Exclusion criteria

1. Patients with a previous history of knee surgery.

2. Patients for whom MRI was contraindicated.

3. Patients with tumors in the joint.

4. Patients who did not give consent for study purposes.

Technique for Ultrasound Examination of Knee General technique

Ultrasound examination was done by using a high frequency probe of frequency 7-9 MHz. It was done in both supine and prone positions. The protocol followed is described below.

1) Medial

The structures evaluated in this aspect were, medial collateral ligament, body and anterior horn of medial meniscus. The position of the patient was supine with hip slightly externally rotated. The evaluation was started in the coronal plane and medial collateral ligaments (MCL) was identified by placing the probe along medial joint line, moving the probe anteriorly and posteriorly until the fibrillar appearance of the ligament was seen. It was evaluated in both short and long axis.^5,7 MCL appeared as broad and flat fibrillar structure having two layers (superficial and deep). A thin layer of fatty areolar tissue or a bursa was observed between the layers.⁸

2) Lateral knee

The structures evaluated were the lateral collateral ligaments (LCL), body and anterior horn of the lateral meniscus. The position of the patient was supine with hip internally rotated and knee slightly flexed. The proximal end of the transducer was stabilized in the groove for the popliteal tendon, and the distal aspect of the transducer was rotated posteriorly to identify the fibular head. The LCL was seen in this plane. The body of the lateral meniscus was identified along the joint line and transducer moved anteriorly to identify the anterior horn.^5,7

Diagnostic Criteria for Tears in USG and MRI Meniscal tears

In USG, normal meniscus appear as triangular hyperechoic structure (Figure 1a). Meniscal tears appear as focal hypoechoic or anechoic linear defects extending to the superior or inferior meniscus surfaces.^5,9-11 (Figure 1b) In case of a bucket-handle lesion, posterior horn appear short, and the front part is dislocated into the knee joint.¹²

In MRI, the menisci appear as low signal intensity structures in all MR sequences. On sagittal images, they appear as either a “bow- tie” structure peripherally or opposing triangles centrally and either triangular or wedge – shaped on coronal images (Figure 1c). MR imaging criteria for diagnosing a tear include meniscal distortion in the absence of prior surgery or increased intrasubstance signal intensity extending up to the articular surface (Figure 1d). The most commonly used indirect signs are a parameniscal cyst, meniscal extrusion, and subchondral marrow edema.¹³

Collateral ligament injury

In USG, torn collateral ligaments show thickening and heterogeneously hypo echogenicity in the ligament on USG.¹⁴ (Figure 1e) Medial collateral injuries are more common than lateral collateral injuries.¹⁵

MRI criteria

Grade 1: Lesions are defined as high signal intensity superficial to the MCL/LCL representing edema with intact MCL/LCL fibers.

Grade 2: Lesions in which fluid signal extend partially through the MCL/LCL, although some fibers remain intact.

Grade 3: Lesions with complete discontinuity of the MCL/LCL fibers seen along with surrounding edema, consistent with a complete rupture (Figure 1f).^16,17

Statistical analysis

Statistical analysis was used to calculate sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy, in order to assess the reliability of the USG and MRI results.

The composite data was tabulated and studied for correlation with MRI findings and grouped into four categories:

1. True-positive - If the USG diagnosis was confirmed by MRI.

2. True-negative - When USG negative for lesion and confirmed by MRI.

3. False-positive - When USG showed lesion but the MRI was negative.

4. False-negative- When MRI was positive but the USG showed negative finding.

To evaluate the sensitivity, specificity, PPV, NPV and accuracy of USG, the findings of MRI were taken to be the true diagnosis.

• Sensitivity was calculated from the number of true positive results divided by the sum of the true positive results and the false negative results.

• Specificity was calculated from the number of true negative results divided by the sum of the true negative results and the false positive results.

• Positive predictive value was calculated from the number of true positives divided by sum of true positives and false positives.

• Negative predictive value was calculated from the number of true negatives divided by sum of true negatives and false negatives.

• Accuracy was calculated by dividing the sum of true positives and true negatives by total number of examinations.

Results

Out of 110 patients, 74 (67%) were males and 36 were females (32.7%) The mean age of male patients was 34.83 years with a standard deviation of 12.65, while the mean age of female patients was 37.83 years with a standard deviation of 14.22.

The proportion of MCL injury in male patients was 1.4%, while it was 0% in females. In this study, no LCL injuries were found. Medial meniscal injury in males was 24% compared to 17% in females. Lateral meniscus (LM) injury in males was 15% and 5.5% in females (Graph 1). Among collateral ligaments injuries, MCL showed a higher proportion of 0.9% compared to LCL (0%). In case of meniscal injuries, medial meniscus (MM) had a higher proportion of 21% and it was only 11% in LM. Out of 110 patients, USG detected medial meniscal tear in 19 patients. USG detected normal medial meniscus in 84 patients which was confirmed by MRI. Though USG detected tear in three patients, MRI showed normal appearance of medial meniscus in these patients. USG missed medial meniscal tear in four patients which was clearly depicted in MRI. The sensitivity, specificity, PPV, NPV and accuracy were 83%, 96.5%, 86%, 95% and 93.6%, respectively.

Out of 110 patients, USG detected lateral meniscal tear in 10 patients and normal lateral meniscus in 94 patients which was confirmed with MRI. Though USG detected tear in three patients, following MRI showed normal appearance. USG did not detect lateral meniscal tear in three patients which was clearly depicted in MRI. The sensitivity, specificity, PPV, NPV and accuracy were 77%, 97%, 77%, 97% and 94.5%, respectively.

Out of 110 patients, USG detected medial collateral ligament tear in one patient and normal medial collateral ligament in 109 patients which was also confirmed by MRI. Though USG detected medial collateral ligament tear in one patient, following MRI evaluation showed normal medial collateral ligament. The sensitivity, specificity, PPV, NPV and accuracy were 100%, 99%, 50%, 100% and 99%, respectively.

Out of 110 patients, USG detected normal lateral collateral ligament in 109 patients which was also confirmed in MRI. Though USG detected tear in one patient, MRI showed normal appearance of lateral collateral ligament in that patient. The sensitivity, specificity, PPV, NPV and accuracy were 0%, 99%, 0%, 100% and 99%, respectively.

Discussion

Knee injuries are one of the common joint problems for sports persons and young adults. Ligamentous and meniscal injuries top the list and if not detected and treated properly at the earliest, may result in temporary or permanent disability, including early osteoarthrosis. Proper clinical evaluation with appropriate imaging plays a crucial role in treatment planning. Though MRI is the chosen modality in the evaluation of joints, Ultrasound is an alternate sensitive modality for the evaluation of joints in situations where MRI is either not available or contraindicated.

Analysis of data in this study showed that the knee injuries are more common in young adult males. This higher proportion is likely due to increased physical activity in these groups, an opinion shared by other authors like Unlu et al. Among the menisci, MM has a higher proportion of injuries of 21%.¹⁸ Medial collateral ligaments are injured more than (0.9%) than LCL. In the present study, the P value for LM and MM evaluation was <0.001 indicating that it was highly statistically significant, whereas in the case of MCL, P value was 0.01 which shows it was statistically significant.

The observations in this study were compared with a similar study conducted by Singh A, Mangat I et al. in the year 2018.¹⁷ This study concluded that USG showed promising results in the evaluation of ligaments and meniscal injuries with a sensitivity of 84.6% in both collateral ligament tear, specificity of 100% for MCL tear and 97.8% for LCL tear. Their results for MM are comparable with the present study but with reduced sensitivity for detecting LM. These variations can be explained by difference in sample size, inter-observer variation and the difference in the USG examination technique they have adopted.

In another study done by Singh B et al., in 2016, 50 patients in the age range of 11 to 65 years were assessed for the sensitivity, specificity, and accuracy of USG in detecting ligament and meniscal injuries.¹⁹ The sensitivity of USG in detecting MCL was 83% and 80% for LCL tears. Both MCL and LCL have similar specificity and accuracy of 97 and 96, respectively. In this study, the mean sensitivity, specificity and mean accuracy was calculated as 83%, 89% and 86%, respectively for both MM and LM. Compared to the above study, the sensitivity in detecting medial collateral ligament injuries was higher, but lower in the case of lateral collateral ligament injuries. However, specificity was almost similar. In the evaluation of menisci, sensitivity, specificity and accuracy were similar. As a conclusion to their study, they have suggested that the USG can be used as a screening tool in the imaging of knee injuries, and if positive, it must be confirmed with MRI. If negative, follow up with MRI, especially when there is no improvement in the symptoms.

A single group cohort study of posterolateral corner injury of the knee was done by Sekiya JK et al.²⁰ in 2010, including both static and dynamic US stress tests. Their static USG revealed a sensitivity of 92%, specificity of 75%, PPV of 92% and NPV of 75%. Though the specificity was comparable with the present study, they reported high sensitivity, probably due to the smaller sample size and relatively younger patients. Dynamic component of the study may be an added reason. Ravichandra G et al., in a double-blinded study of USG with arthroscopic correlation in 50 patients reported meniscal tears in 34 patients.²¹ The sensitivity, specificity, PPV, NPV and accuracies were 62.5%, 80.7%, 75%, 70% and 97% for MM and 23%, 89%, 42.8%, 77% and 95% for LM, respectively. Compared to this, the current study showed higher sensitivity of MM and specificity of MM and LM, but lower sensitivity of MM, PPV and NPV for MM and LM were higher in the present study. A very similar study by Cook et al., conclusively demonstrated that USG can detect meniscal tears with a high degree of specificity and sensitivity. ²²

A study by Hussein SMA et al., opined that the USG can be used as the first modality in evaluating trauma in situations where MRI is either unavailable or contraindicated.²³ In the evaluation of meniscal injuries, the study revealed almost 100% sensitivity and specificity except for reduced sensitivity of 83% for the posterior horn of MM. The current study demonstrated the specificity of more than 90% in detecting meniscal tears; however sensitivity for detecting meniscal injuries was significantly low. A double-blinded prospective study by Khan Z et al. assessed the diagnostic accuracy of USG in detecting knee injuries in comparison with MRI and arthroscopy evaluated both menisci and cruciate ligaments.²⁴ Compared to their study, the specificity in this study was similar, while sensitivity was reduced. Similarly, the NPV was almost similar to this study, but the PPV was reduced. This study emphatically confirms that the USG has a significant role in the evaluation of meniscal and ligament inquires of the knee, though MRI is the gold standard non-invasive imaging technique.

USG evaluation of meniscal injuries in 15 patients with a mean age 30 years by Mehdy NS et al., reported a sensitivity of 88%, slightly more than the result of 77- 83% obtained in this study.²⁵ Yaseen KM et al., in 2019 in their study involving 50 patients with suspected knee injury reported a sensitivity of 91%, specificity of 20% and accuracy of 70 % for ultrasound in detecting meniscal injuries.²⁶ These results were almost comparable with the current study except for a minimal drop of sensitivity.

A comparison of the present study with existing literature on the subject agrees that USG is a reliable modality for imaging injured knee with a reasonable degree of sensitivity, specificity, NPV, PPV and accuracy. The majority of the reported results validate that currently MRI is the gold standard of imaging for knee injuries. Incorporation of dynamic evaluation with static imaging as reported in some studies will improve the efficacy of ultrasonography of the knee drastically. The discordance in some of the parameters reported by different authors may be due to a host of factors like machine sensitivity, the competence of personnel, the difference in sample selection, and patient-related factors. The main hurdle before the ultrasound to occupy the prime position in the evaluation of knee and other joint disorders are its high dependency on equipment and operators. The only way to overcome these drawbacks is the constant upgradation of equipment and continuous acquisition of knowledge and skill by radiologists in musculoskeletal ultrasound which will help in a great way to achieve the goal of an ultrasound examination of the knee with equal specificity, sensitivity, NPV, PPV and accuracy as that of MRI.

Conclusion

After analyzing the results, the following observations were made. The knee ligaments and meniscal injuries are more common in males than females and in young adults than old people. Ultrasound shows reasonably good sensitivity, specificity, NPV, PPV and accuracy in detecting meniscal and medial collateral ligaments injury. Hence the USG can be done as a screening test if MRI is not available or contraindicated. In case of lateral collateral ligament, P value cannot be calculated as there were no cases with lateral collateral tear in MRI and USG.

Ethics approval and consent to participate

This study obtained an ethical clearance from the KMCT medical college institutional ethical committee with ref no: KMCT/RP 2020/IEC/05. Consent from all the patients were also obtained before the commencement of study.

Competing interests

Nil