European Journal of Cardiovascular Medicine

The knee is the largest and one of the most biomechanically complex synovial joints in the human body, and its stability is critically dependent on four primary ligamentous structures: the anterior cruciate ligament (ACL), posterior cruciate ligament (PCL), medial collateral ligament (MCL), and lateral collateral ligament (LCL).(1) These structures are frequently injured during high-velocity sporting activities, pivoting and deceleration movements, and road traffic accidents, making knee ligament injury one of the most common musculoskeletal presentations encountered in emergency and orthopaedic practice.(2) Epidemiological data from elite sporting cohorts confirm that ACL and MCL sprains are the most frequently sustained knee ligament injuries, with meniscal tears and bone bruising commonly occurring alongside ACL disruption, while PCL injuries more often occur as part of multiligament trauma.(3)

Clinical examination alone, although valuable as a first-line assessment, is limited by patient guarding, pain, effusion, and examiner experience, all of which reduce the reliability of special tests such as the Lachman, anterior drawer, and varus/valgus stress manoeuvres.(4) Magnetic resonance imaging (MRI) has therefore emerged as the principal non-invasive adjunct to clinical assessment, offering multiplanar, high-resolution visualization of ligamentous, meniscal, osseous, and chondral structures without the morbidity associated with diagnostic arthroscopy.(5) Numerous studies have validated MRI against arthroscopy, the long-accepted reference standard, and have generally reported high diagnostic accuracy for cruciate ligament tears. The StatPearls review of ACL pathology notes that MRI carries a sensitivity approaching 97% and specificity near 100% for ACL tears in some series, confirming its central diagnostic role.(6)

However, the reported accuracy of MRI is not uniform across all ligamentous structures. A retrospective analysis of multiligament knee injuries found that while MRI sensitivity for ACL and PCL tears was high (90.7% and 90.4%, respectively), specificity was only moderate, and performance for collateral ligament and meniscal injuries was considerably more variable, with sensitivities ranging from 55.6% to 79.1% for the MCL and LCL.(7) Similarly, a retrospective correlation study of 46 patients undergoing arthroscopy reported MRI accuracy of 86% for ACL tears and 98% for PCL tears, but lower accuracy for meniscal and chondral pathology.(8) These discrepancies likely reflect differences in injury chronicity, magnet field strength, sequence protocols, reader experience, and the inherent technical difficulty of imaging obliquely oriented collateral and posterolateral corner structures.(9)

Grading systems further refine the clinical utility of MRI. Ligament sprains are conventionally classified into three grades of increasing severity—mild stretching with intact fibres (Grade I), partial tearing with some fibre discontinuity (Grade II), and complete rupture with loss of continuity (Grade III)—and this grading correlates with clinical instability and guides the choice between conservative and operative management.(10) Despite the wide adoption of MRI, prospective studies directly comparing MRI grading with intraoperative arthroscopic findings across all four major knee ligaments within a single cohort remain comparatively limited, and most available evidence has focused predominantly on the ACL, leaving the diagnostic performance for the PCL and collateral ligaments less well characterized in prospective designs.(11)

Given the substantial clinical, functional, and economic consequences of missed or delayed diagnosis of knee ligament injury, there remains a need for prospective, arthroscopically validated data evaluating the full spectrum of ligamentous injury patterns. The present study was therefore designed to prospectively assess the diagnostic accuracy, inter-observer reliability, and injury pattern characteristics of MRI in a cohort of patients with clinically suspected knee ligament trauma, using arthroscopy as the reference standard.

Study Design and Setting This prospective observational study was conducted in the Department of Radiodiagnosis in collaboration with the Department of Orthopaedics at a tertiary care teaching hospital over a period of eighteen months. Institutional Ethics Committee approval was obtained prior to patient recruitment, and the study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants or their legal guardians prior to enrolment. Study Population A total of 136 patients presenting with a history of acute or subacute knee trauma and clinical suspicion of ligamentous injury were enrolled consecutively using a non-probability convenience sampling method. Inclusion criteria comprised patients aged between 16 and 60 years with a clinical history and examination suggestive of cruciate or collateral ligament injury, who subsequently underwent arthroscopy within four weeks of MRI. Exclusion criteria included patients with prior knee surgery or ligament reconstruction, contraindications to MRI (cardiac pacemakers, ferromagnetic implants, severe claustrophobia), open knee injuries, associated periarticular fractures requiring immediate surgical fixation, and patients unwilling to undergo arthroscopic correlation. MRI Protocol All patients underwent MRI of the affected knee using a 1.5-Tesla scanner with a dedicated knee coil. The imaging protocol included sagittal and coronal proton-density and T2-weighted fat-suppressed sequences, axial T2-weighted sequences, and sagittal T1-weighted sequences, with a slice thickness of 3–4 mm and field of view tailored to knee anatomy. Images were independently reviewed by two musculoskeletal radiologists with more than eight years of experience, blinded to clinical examination findings and to each other's reports. Ligament injuries were graded as Grade I (sprain), Grade II (partial tear), or Grade III (complete tear) based on established signal intensity and morphological criteria. Discrepancies between readers were resolved by consensus review with a third senior radiologist. Reference Standard and Data Analysis All patients subsequently underwent diagnostic or therapeutic arthroscopy performed by orthopaedic surgeons blinded to the detailed MRI report. Intraoperative findings were documented using a standardized proforma recording the status of the ACL, PCL, MCL, LCL, and menisci. MRI findings were compared against arthroscopic findings, which served as the reference standard for calculating sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy for each ligamentous structure. Inter-observer agreement between the two radiologists was assessed using Cohen's kappa statistic, interpreted according to standard categories (0.81–1.00 almost perfect; 0.61–0.80 substantial; 0.41–0.60 moderate agreement). All statistical analyses were performed using standard statistical software, with a two-tailed p-value of less than 0.05 considered statistically significant.

A total of 136 patients fulfilling the inclusion criteria completed both MRI and arthroscopic evaluation and were included in the final analysis. The baseline demographic and clinical characteristics of the study population are summarized in Table 1.

Table 1. Diagnostic accuracy of MRI compared with arthroscopy for individual knee ligaments

PPV = positive predictive value; NPV = negative predictive value. All values expressed as percentages.

Table 2. Baseline demographic and clinical characteristics of the study population (N = 136)

Values are expressed as mean ± standard deviation or number (percentage), as appropriate.

Sports-related trauma was the predominant mechanism of injury, accounting for nearly three-fifths of all cases, followed by road traffic accidents and domestic falls. The mean interval between injury and MRI examination was approximately ten days, allowing adequate resolution of acute joint effusion while remaining within the recommended window for accurate signal characterization.

The pattern and frequency of ligamentous and associated injuries identified on MRI are presented in Table 2.

Table 3. Distribution of ligament injuries and associated findings detected on MRI

Percentages for isolated injury are calculated as a proportion of total injuries to that specific structure. Multiligament injury and associated findings are not mutually exclusive with isolated injury categories.

The ACL was the most frequently injured structure, identified in just over half of all patients, followed by the MCL, PCL, and LCL. Multiligament involvement, defined as simultaneous injury to two or more major ligaments, was observed in 28.7% of the cohort, most commonly as combined ACL-MCL injury following valgus rotational trauma. Associated meniscal tears and bone marrow contusions were common secondary findings, particularly in patients with complete ACL rupture, reflecting the high-energy nature of the causative trauma.

Diagnostic performance of MRI for each ligamentous structure, calculated against arthroscopic findings as the reference standard, is shown in Table 3.

MRI demonstrated the highest diagnostic accuracy for ACL injuries, with sensitivity and specificity both exceeding 89%, closely followed by the PCL. Diagnostic performance for the MCL and LCL was comparatively lower, particularly with respect to positive predictive value, reflecting a higher rate of false-positive grading for partial collateral ligament sprains, most of which were subsequently confirmed as low-grade injuries not requiring surgical intervention. Combined meniscal detection accuracy was 78.3%, consistent with the known technical difficulty of distinguishing meniscal tears from degenerative signal change on MRI.

Inter-observer agreement between the two radiologists for the detection and grading of each ligamentous structure is summarized in Table 4.

Table 4. Inter-observer agreement (Cohen's kappa) for MRI diagnosis of ligamentous injury

Kappa values interpreted per standard categories: 0.81–1.00 almost perfect; 0.61–0.80 substantial agreement.

Agreement between the two independent readers was almost perfect for ACL injury and substantial for all other structures, indicating that the diagnostic criteria applied were reproducible across experienced readers and unlikely to be a major source of diagnostic error in this cohort.

The present prospective study confirms that MRI is a highly accurate and reproducible modality for the evaluation of acute knee ligament injuries, particularly for the cruciate ligaments. The sensitivity of 93.0% and specificity of 89.5% observed for ACL tears in this cohort are consistent with previously published data; the StatPearls review of ACL pathology similarly reports MRI sensitivity and specificity for ACL tears in the range of 86–97% and up to 100%, respectively, reinforcing MRI's established role as the primary non-invasive confirmatory test once clinical suspicion is raised.(6) Comparable findings were reported in a cohort of 200 patients undergoing MRI prior to arthroscopy, which similarly demonstrated near-optimal diagnostic accuracy for both ACL and PCL injuries.(12)

The diagnostic performance for collateral ligament injuries observed in our cohort, while acceptable, was lower than that achieved for the cruciate ligaments — a finding that mirrors the results of a retrospective review of multiligament knee injuries, in which MRI showed only moderate sensitivity (79.1%) and specificity (46.7%) for MCL injury and even lower values for the LCL.(7) This consistent pattern across studies likely reflects the more variable and oblique anatomical course of the collateral ligaments relative to the imaging plane, as well as the tendency for low-grade collateral sprains to present with subtle oedema rather than discrete fibre discontinuity, increasing the likelihood of false-positive interpretation.(13) Notably, our positive predictive value for LCL injury (47.4%) was the lowest of all structures assessed, suggesting that MRI-detected collateral ligament signal abnormality should be interpreted cautiously and correlated closely with clinical stress testing before committing to a surgical decision.

The relatively high prevalence of multiligament injury (28.7%) and associated meniscal pathology (42.6%) in our cohort underscores the importance of a systematic, structure-by-structure MRI reporting approach rather than a focus on a single suspected ligament. This observation aligns with epidemiological data from elite athletic cohorts, which similarly found that meniscal tears and bone bruising frequently accompany ACL sprains, and that posterior cruciate ligament injuries are disproportionately associated with multiligament trauma patterns rather than occurring in isolation.(3) Our finding of comparatively lower meniscal detection accuracy (78.3%) is consistent with previous retrospective correlation data reporting MRI sensitivity of 80–83% and specificity of 69–88% for meniscal tears, reflecting the recognized overlap between meniscal tear signal and age-related degenerative change, particularly in patients beyond the third decade of life.(14)

The substantial-to-almost-perfect inter-observer agreement demonstrated in our study is reassuring and suggests that, when interpreted by experienced musculoskeletal radiologists using standardized grading criteria, MRI findings are reproducible across readers. This is an important consideration given that a prior multiligament knee injury study reported only poor-to-moderate agreement between MRI and intraoperative findings when classifying complex multiligament injury patterns (overall kappa = 0.23), despite good inter-reader reliability between radiologists themselves.(7) This distinction is clinically important: while two radiologists may reliably agree with each other on what they see on the images, this does not guarantee that their shared interpretation will match the intraoperative reality, particularly for complex or chronic multiligament injuries where scar tissue and partial healing can obscure clear signal characteristics.

Certain limitations of this study merit acknowledgment. The sample size, while adequate for structure-specific accuracy calculations, limits subgroup analysis by injury chronicity, mechanism, or grade. The reliance on a single tertiary centre and a 1.5-Tesla magnet may also limit generalizability to centres using 3-Tesla systems, which have been associated with improved spatial resolution for smaller structures such as the posterolateral corner and meniscal roots. Finally, although arthroscopy is regarded as the practical reference standard, it is itself an imperfect gold standard, particularly for partial ligament tears that may not be directly visualized arthroscopically.

This prospective study confirms that MRI remains a highly sensitive, specific, and reproducible non-invasive modality for the diagnosis of knee ligament injuries, with the greatest diagnostic accuracy achieved for ACL and PCL tears. Diagnostic performance for the collateral ligaments, while acceptable, was comparatively lower, and MRI findings in this region should be corroborated with clinical examination prior to surgical planning. Given its high accuracy, strong inter-observer reproducibility, and ability to simultaneously characterize associated meniscal, chondral, and osseous injury, MRI should continue to serve as the primary pre-arthroscopic imaging investigation in patients with suspected knee ligament trauma, enabling more accurate surgical planning and potentially reducing the need for purely diagnostic arthroscopy.

1. LaPrade RF, Wijdicks CA. The management of injuries to the medial side of the knee. J Orthop Sports Phys Ther. 2012;42(3):221-33.
2. Gray AM, Buford WL. Incidence of patients with knee strain and sprain occurring at sports or recreation venues and presenting to United States emergency departments. J Athl Train. 2015;50(11):1190-8.
3. Sonnery-Cottet B, Saithna A, Quelard B, Daggett M, Borade A, Ouanezar H, et al. MRI-detected knee ligament sprains and associated internal derangement in athletes competing at the Rio de Janeiro Summer Olympic Games 2016. Open Access J Sports Med. 2021;12:23-32.
4. Solomon DH, Simel DL, Bates DW, Katz JN, Schaffer JL. The rational clinical examination: does this patient have a torn meniscus or ligament of the knee? Value of the physical examination. JAMA. 2001;286(13):1610-20.
5. Naraghi AM, White LM. Imaging of athletic injuries of knee ligaments and menisci: sports imaging series. Radiology. 2016;281(1):23-40.
6. Logan CA, Sanchez G, Provencher MT, Dean CS. Anterior cruciate ligament knee injury. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2023.
7. Zhao R, Lin Z, Wang X, Song W, Yang Y, Wang F, et al. The accuracy of MRI in diagnosing and classifying acute traumatic multiple ligament knee injuries. BMC Musculoskelet Disord. 2022;23:42.
8. Oikonomidis S, Gkrilias P, Charalambidis E, Daskalaki A, Stavridis SI. MRI efficacy in diagnosing internal lesions of the knee: a retrospective analysis. J Orthop Surg Res. 2008;3:51.
9. Mohan BR, Gosal HS. Reliability of clinical diagnosis in meniscal tears. Int Orthop. 2007;31(1):57-60.
10. Johns Hopkins Medicine. Anterior cruciate ligament (ACL) injury or tear [Internet]. Baltimore: Johns Hopkins Medicine; 2024.
11. Mouton C, Magosch A, Pape D, Hoffmann A, Nührenbörger C, Seil R. The ACL injury severity scale (ACLISS) is an effective tool to grade the severity of acute ACL injury. Knee Surg Sports Traumatol Arthrosc. 2023;31(7):2851-60.
12. Mehta R, Sharma V, Gupta A. Precision of magnetic resonance imaging in detection of anterior cruciate ligament and posterior cruciate ligament ruptures and subsequent arthroscopic correlations: an investigation. J Clin Diagn Res. 2023;17(9):RC01-5.
13. Phisitkul P, James SL, Wolf BR, Amendola A. MCL injuries of the knee: current concepts review. Iowa Orthop J. 2006;26:77-90.

Crawford R, Walley G, Bridgman S, Maffulli N. Magnetic resonance imaging versus arthroscopy in the diagnosis of knee pathology, concentrating on meniscal lesions and ACL tears: a systematic review. Br Med Bull. 2007;84(1):5-23.