Quantitative magnetic resonance imaging during knee hyperextension for the evaluation of screw-home mechanism abnormalities in patients with anterior cruciate ligament injury: a clinical study

Abstract

Background

Anterior cruciate ligament (ACL) injury, a common sports-related knee ligament injury, often leads to screw-home mechanism (SHM) dysfunction and joint instability. Clinical testing and conventional magnetic resonance imaging (MRI) do not provide satisfactory quantitative assessment of functional instability post-acute ACL injury. To improve ACL injury diagnosis and treatment, this study examined the ability of MRI during knee hyperextension to clinically evaluate postinjury SHM abnormalities.

Methods

This retrospective study enrolled 56 participants who underwent positional MRI of the knee in hyperextension from June 2020 to June 2025, including 21 controls, 15 patients with acute partial tear of the ACL, and 20 with acute complete rupture of the ACL (acute injury was defined as an injury-to-MRI interval ≤14 days). None of the enrolled patients had knee extension deficit, and standard hyperextension position was maintained during scanning to simulate terminal knee extension. The standardized tibiofemoral translation distances and tibiofemoral rotation angle were measured. One-way analysis of variance was used for intergroup comparison, and receiver operating characteristic (ROC) curve analysis was applied to determine diagnostic efficacy, with P<0.05 indicating statistical significance.

Results

In the control group, the tibiofemoral rotation angle and standardized lateral tibiofemoral translation during hyperextension were 6.27±2.35° and 0.016±0.035, respectively, which is consistent with the physiological function of a normal SHM; these two parameters were −4.51±5.88° and 0.073±0.059 in the partial tear group and −6.18±4.21° and 0.103±0.051 in the complete rupture group, respectively. Compared with controls, patients showed significantly elevated standardized lateral tibiofemoral translation and reduced tibiofemoral rotation angle (F=17.218 and F=52.013, respectively; P<0.001). No significant differences in these kinematic parameters were observed between the partial and complete tear groups or between patients with isolated ACL injury and those with concomitant injuries to other knee-joint-stabilizing structures (P>0.05). Standardized medial tibiofemoral translation showed no significant intergroup difference (P=0.269) and had limited diagnostic value. In the ROC curve analysis, the tibiofemoral rotation angle had the highest single-index efficacy in diagnosing ACL injury-related SHM abnormalities [area under the curve (AUC) =0.985; sensitivity =100%; specificity =97.1%; P<0.001]. When this rotation angle was combined with standardized lateral tibiofemoral translation, diagnostic performance further improved (AUC =0.996; P<0.001). The intraclass correlation coefficients for interobserver reliability of all measurements were >0.95 (P<0.001), indicating excellent measurement consistency.

Conclusions

By simulating the physiological characteristics of the knee joint in terminal extension, MRI in the hyperextended position can quantify the abnormal kinematic parameters of the tibiofemoral joint after acute ACL injury and accurately evaluate the rotational and translational stability of the joint, with an extremely high sensitivity (100%) and specificity (97.1%) demonstrated by the tibiofemoral rotation angle. This technique is simple, noninvasive, and equal in duration to conventional MRI and is thus a practical, reliable, and objective evaluation method for clinical practice.

Introduction

The anterior cruciate ligament (ACL) is one of the most important ligaments of the knee joint, providing rotational and translational stability to the joint. With the growth of competitive sports and the widespread implementation of national fitness campaigns, the incidence of sports-related knee joint injuries is increasing each year. Among these injuries, ACL injury is one of the most common ligamentous injuries of the knee joint, with a particularly high incidence in sports requiring sudden deceleration, direction changes, or jumping (such as basketball, football, badminton, and skiing) (1). ACL injury is often accompanied by injuries to other stabilizing structures of the knee joint, with those of the medial and lateral menisci and the medial collateral ligament being the most common. These combined injuries can further aggravate knee-joint instability and significantly increase the long-term risk of posttraumatic osteoarthritis in patients (2,3).

The degree of knee joint instability ranges from mild to severe and can be determined according to rotational or translational instability (4). The diagnosis of an ACL tear is achieved through a combination of clinical tests and magnetic resonance imaging (MRI). The main clinical tests include the lever test, Lachman test, anterior drawer test, and pivot shift test, but these have limited sensitivity and specificity for partial tears and acute injuries (5). However, patients with ACL injuries who exhibit similar MRI manifestations may have varying degrees of clinical instability (6,7). In a 2010 study, Ahn et al. found that some patients with complete tears on MRI appeared to have only mild instability on clinical tests (Lachman grade ≤1), and follow-up MRI after nonsurgical treatment indicated recovery of ligament continuity and signal (8). A growing body of scientific evidence suggests that for patients with ACL injury, knee joint stability can be restored through a properly constructed rehabilitation program (9).

Therefore, reliable and practical measurement methods are needed to improve the assessment of knee-joint stability (10). Evaluating knee-joint stability requires the detection of subtle changes in the alignment of the tibiofemoral joint. In contrast to nonimaging methods, MRI can measure the tibiofemoral rotation angle (TRA) with high reliability and separately determine the medial and lateral translation distances of the tibiofemoral joint. In 2023, Griffith et al. reported that positional MRI (including the hyperextension position), by quantifying tibiofemoral translation and rotation, can more accurately reflect the dynamic stability of the knee joint as compared to conventional MRI (11). Critically, in the last 20 degrees of knee extension from flexion to full extension, the tibia externally rotates 15 degrees relative to the femur to achieve full extension. This physiological movement is referred to as the screw-home mechanism (SHM). This mechanism is primarily based on the asymmetry of the medial and lateral femoral condyles and their articular surfaces, in addition to the passive tension of the ACL and the lateral pull of the quadriceps femoris (12). Tibial external rotation can tighten the cruciate, collateral, and other capsular ligaments, thereby allowing the knee joint to maintain stability in the fully extended position (13,14). However, the study by Griffith et al. involved several limitations, including the fact that the participants were all patients with chronic knee injuries.

Early identification of patients who can benefit from early ACL reconstruction or simple rehabilitation is crucial for reducing resource consumption and overtreatment. Consequently, in our study, we attempted to use MRI during knee hyperextension to simulate the terminal extension state of the knee joint, quantitatively analyze the tibiofemoral kinematic parameters of patients with acute ACL injury and healthy controls, evaluate the efficacy of these parameters in diagnosing SHM abnormalities in patients with ACL injury, and thus provide a noninvasive and repeatable evaluation tool for clinical practice. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2628/rc).

Methods

Study participants

This single-center retrospective case-control study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Medical Ethics Committee of Central People’s Hospital of Zhanjiang (No. ZX-2025051-01). All participants provided written informed consent. A total of 56 patients who underwent MRI during knee hyperextension at Central People’s Hospital of Zhanjiang from June 2020 to June 2025 (mean age 25.63±5.58 years) were enrolled in this study and placed into a control group or knee-injury group.

The inclusion criteria for the control group were an age of 18–60 years and no history of ACL injury or knee-joint surgery. Ultimately, 21 participants (12 females and 9 males) were enrolled in the control group. Meanwhile, the knee-injury group included patients who underwent MRI examination for acute knee injury (5 females and 30 males). The exclusion criteria for this group were a history of knee-joint surgery and age >60 years. Two radiologists with 11 years of experience in musculoskeletal imaging diagnosis determined the type and severity of the knee injury. According to uniform diagnostic criteria and the double-blind reading protocol detailed in the Image analysis section, the radiologists further divided the patients in the knee-injury group into a partial tear group (15 cases) and a complete rupture group (20 cases).

In this study, acute ACL injury was defined as an interval from injury onset to MRI examination of ≤14 days. The primary reasons for only enrolling patients with acute injury and excluding those with chronic injury are as follows: (I) chronic ACL injury is often accompanied by muscle atrophy around the knee joint, adaptive kinematic compensation formed by altered neuromuscular control, and secondary changes of intra-articular structures, including secondary cartilage injury and meniscal injury. These factors could directly affect the kinematic parameters of tibiofemoral translation and rotation while severely interfering with the accurate evaluation of ACL injury’s direct impact on the SHM in this study. (II) One of the key objectives of this study was to develop a method that can facilitate the early identification of patients who can benefit from early ACL reconstruction or standardized rehabilitation treatment. Acute injury, lacking secondary compensation and structural adaptation, allows for a more accurate observation of immediate joint stability abnormalities after ACL injury, aligning closely with the study’s research objectives.

In our study, an upper age limit of 60 years was set for enrollment, mainly due to the following considerations: (I) older adult patients aged over 60 years generally have varying degrees of degenerative changes of the knee joint, including osteoarthritis, degenerative meniscal injury, and cartilage degeneration. These can impair the normal function of the knee SHM and the baseline kinematic characteristics of the tibiofemoral joint, thereby introducing bias into the results. (II) The incidence of sports-related acute ACL injury among those aged over 60 years is extremely low. The primary injury mechanism, clinical treatment decision-making, and functional prognosis of this population are significantly different from those of the young and middle-aged patients.

Imaging acquisition

All MRI scans were conducted with a 3.0T Signa Pioneer MRI system (GE HealthCare, Chicago, IL, USA) equipped with a knee-specific coil. A soft cushion of appropriate height was placed under the patient’s ankle to elevate it, and the patient’s knee joint was maintained in hyperextension by gravity (Figures 1,2). The imaging protocol consisted of conventional knee sequences, including proton density-weighted fat-suppressed sagittal, coronal, and axial sequences, as well as a T1-weighted sagittal imaging sequence (the details of the parameters are provided in Table 1). The total duration of a single unilateral knee hyperextension position for MRI examination was approximately 7–8 minutes, which was consistent with that of a conventional MRI examination.

Figure 1.

Standardized positioning of the knee joint in hyperextension for MRI scanning. A soft cushion of appropriate height was placed under the patient’s ankle to elevate it, and the knee joint was maintained in hyperextension by gravity. MRI, magnetic resonance imaging.

Figure 2.

Definition of the physiological range of knee hyperextension. The standard anatomical position of full knee extension is 0° in most individuals; a mild hyperextension (approximately 5–10° beyond 0°) is considered a normal variant.

Table 1. Scanning parameters.

Parameter	Sagittal fat suppression PD	Sagittal T1FSE	Coronal fat suppression PD	Axial fat suppression PD
TE (ms)	42	20	42	42
TR (ms)	2,209	713	2,208	2,648
Flip angle (°)	111	90	111	111
Suppression method	FS	–	FS	FS
Resolution (mm)	0.6×0.8	0.6×0.7	0.5×0.6	0.6×0.8
Slice thickness (mm)/gap (mm)	4/0.4	4/0.4	4/0.4	4/0.4
2D or 3D acquisition	2D	2D	2D	2D
Acquisition matrix	320×256	320×256	288×256	320×224
Number of averages	2	1	2	2
Time	1 min and 42 s	36 s	1 min and 51 s	1 min and 30 s

FS, fat suppression; PD, proton density; T1FSE, T1-weighted fast spin echo; TE, echo time; TR, repetition time.

Image analysis

Data measurement was independently performed by two attending physicians with more than 11 years of experience in musculoskeletal imaging diagnosis under a blinded method. One of the physicians repeated the measurements 2 weeks later for evaluation of interobserver and intraobserver reliability. All measurements were performed on the same workstation of the picture archiving and communication system with standardized measurement tools and parameter settings. The measured indicators were (I) the tibiofemoral translation distance and anteroposterior diameter of the medial and lateral femoral condyles and (II) the TRA.

The measurement of the tibiofemoral translation distance and anteroposterior diameter of the medial and lateral femoral condyles was as follows: on sagittal proton density-weighted fat-suppressed images, the midcoronal level of the medial and lateral femoral condyles was localized with reference to axial cross-sectional images. Vertical tangents perpendicular to the cortical bone were drawn along the posterior margin of the medial/lateral femoral condyle and the posterior cortical margin of the corresponding medial/lateral condyle of the proximal tibia. The two tangents were parallel to one another, and the distance between them was defined as the tibial translation distance, with the anterior translation of the tibia relative to the femur being set as positive and posterior translation as negative. On the same sagittal slice, the distance between the anterior and posterior cortical margins of the medial and lateral femoral condyles was respectively measured (Figure 3).

Figure 3.

Standardized measurement method for tibiofemoral translation distance. On sagittal proton density-weighted fat-suppressed images, the midcoronal level of the medial and lateral femoral condyles was localized with reference to axial cross-sectional images. Vertical tangents perpendicular to the cortical bone were drawn along the posterior margin of the medial/lateral femoral condyle and the posterior cortical margin of the corresponding medial/lateral condyle of the proximal tibia, respectively. The two tangents were parallel to each other, and the distance between them was defined as the tibial translation distance. (A,B) A 10.3-mm anterior medial tibial translation, with an anteroposterior diameter of 57.4 mm for the medial femoral condyle. (C,D) A 3.8-mm anterior lateral tibial translation, with an anteroposterior diameter of 64.6 mm for the lateral femoral condyle.

The measurement of the TRA was as follows: for the distal femur, the axial slice passing through the posterior-most margins of the medial and lateral femoral condyles was selected; for the proximal tibia, the axial slice immediately superior to the fibular head passing through the posterior-most margins of the medial and lateral tibial condyles was selected. The intercondylar line of the distal femur (connecting the posterior-most margins of the medial and lateral femoral condyles) and the intercondylar line of the proximal tibia (connecting the posterior-most margins of the medial and lateral tibial condyles) were each drawn. The included angle between the two lines was defined as the TRA, with the external rotation of the tibia relative to the femur being set as positive and the internal rotation as negative (Figure 4).

Figure 4.

Standardized measurement method for tibiofemoral rotation angle. Axial proton-weighted fat-suppressed images showing (A) the distal right femur and (B) proximal tibia. A line was drawn through the posterior cortex of the medial and lateral femoral condyles, and another line was drawn through the medial and lateral tibial condyles at the level immediately superior to the proximal fibula. The included angle between these two lines is the tibiofemoral rotation angle, with positive values indicating tibial external rotation relative to the femur. In this case, the tibia externally rotated 14.8° relative to the femur.

Considering that interindividual differences in physique and osseous structure dimensions may affect the absolute measurements of tibiofemoral translation distance and thus introduce potential bias in intergroup comparisons, we selected the local osseous anatomical diameter of the knee joint that directly matched the measurement plane as the standardized reference, specifically, the anteroposterior diameter of the ipsilateral medial and lateral femoral condyles. Standardized correction of the tibiofemoral translation distance was performed according to the ratio of the ipsilateral tibiofemoral translation distance to the anteroposterior diameter of the ipsilateral lateral femoral condyle.

Based on the above-described MRI measurement protocols, we performed standardized diagnostic classification of ACL injury for all participants using the same set of hyperextension-position MRI scans. The quality control protocol and diagnostic criteria were as follows: two radiologists with more than 11 years of experience in musculoskeletal imaging diagnosis independently completed the diagnostic classification of ACL injury for all participants in a double-blinded manner. When there were discrepancies in the diagnostic results between the two radiologists, a chief physician with more than 20 years of experience in musculoskeletal imaging diagnosis was invited to perform the final image reading and judgment, and cases were assigned to the corresponding group only after a consistent diagnosis was reached. In the diagnosis of acute ACL tear, partial tear was defined as an irregular or wavy ligament contour accompanied by altered signal intensity within the ligament parenchyma and a preserved continuity of ACL fibers in at least one T2-weighted image section (Figure 5); meanwhile, complete rupture was defined as discontinuity of the ligament parenchyma in all MRI sections (Figure 6).

Figure 5.

MRI manifestations and kinematic parameter measurements in a patient with acute ACL partial tear. (A-C) PDWI-FS multiplanar images showing an irregular contour and signal intensity changes in the ACL (arrows), consistent with a partial tear. (D) A 6.3-mm anterior lateral tibial translation. (E,F) An internal tibial rotation of 4.1° relative to the femur. ACL, anterior cruciate ligament; MRI, magnetic resonance imaging; PDWI-FS, proton density-weighted imaging with fat suppression.

Figure 6.

MRI manifestations and kinematic parameter measurements in a patient with acute ACL complete rupture. (A-C) PDWI-FS multiplanar images showing an irregular morphology and discontinuity of the ACL with signal intensity changes (arrows), confirming a complete tear. (D) A 5.3-mm anterior lateral tibial translation. (E,F) An internal tibial rotation of 7.3° relative to the femur. ACL, anterior cruciate ligament; MRI, magnetic resonance imaging; PDWI-FS, proton density-weighted imaging with fat suppression.

Statistical analysis

SPSS 26.0 statistical software (IBM Corp., Armonk, NY, USA) was used for data analysis. The Shapiro-Wilk test was used to verify the normality of the measurement data, and all data were confirmed to conform to the normal distribution without data transformation. Measurement data are expressed as the mean ± standard deviation. One-way analysis of variance was used to compare multiple groups (control, partial tear, and complete rupture), and the least significant difference t-test was used for pairwise comparisons; an independent-samples t-test was used to compare the partial tear and complete rupture groups. The intraclass correlation coefficient (ICC) was used to evaluate the consistency of the two radiologists’ measurement results, with an ICC >0.90 indicating extremely high reliability. The partial tear group and the complete rupture group were combined into the injury group. Receiver operating characteristic (ROC) curve analysis was performed with reference to the control group to calculate the area under the curve (AUC), sensitivity, specificity, and Youden index and to determine the optimal cutoff value for each indicator for identifying injuries. A P value <0.05 was considered statistically significant.

Results

General characteristics of the participants

The general characteristics of the participants are provided in Table 2.

Table 2. General characteristics of the participants.

Characteristics	Value
Age (years)	25.63±5.58
Gender
Male	39 (69.64)
Female	17 (30.36)
Examination site
Left knee	28 (50.00)
Right knee	28 (50.00)
Injury to other stabilizing structures
No	34 (60.71)
Yes	22 (39.29)

Data are presented as n (%) for categorical variables and mean ± standard deviation for continuous variables.

Descriptive statistics of core indicators

Standardized lateral tibiofemoral translation (SLTT)

SLTT, defined as the ratio of lateral tibiofemoral translation distance to the anteroposterior diameter of the ipsilateral lateral femoral condyle, showed a distinct gradient pattern across the three groups (Table 3). In the control group, this indicator had the narrowest range of variation (−0.051 to 0.073), with a median of 0 and a mean value of 0.016±0.035, indicating that lateral tibiofemoral translation was maintained at an extremely low level with excellent joint stability under physiological conditions. In the partial tear group, the range of variation was larger (−0.052 to 0.17), which was significantly higher than that in the control group, with a median of 0.071 and a mean value of 0.073±0.059. In the complete rupture group, the range of variation was even larger (0.029 to 0.194), with a median of 0.098 and a mean value of 0.103±0.051, the highest among the three groups. The degree of lateral tibiofemoral translation increased progressively with the severity of ACL injury, suggesting that this indicator is closely correlated with the severity of ligament injury.

Table 3. Descriptive statistics of standardized lateral tibiofemoral translation (positive value indicates anterior tibial translation).

Group	Minimum	Maximum	Median	Mean ± SD
Control	−0.051	0.073	0	0.016±0.035
Partial tear	−0.052	0.17	0.071	0.073±0.059
Complete rupture	0.029	0.194	0.098	0.103±0.051

SD, standard deviation.

Standardized medial tibiofemoral translation (SMTT)

Across the three groups, no distinct gradient pattern was observed in the SMTT, defined as the ratio of medial tibiofemoral translation distance to the anteroposterior diameter of the ipsilateral medial femoral condyle (Table 4). In the control group, this indicator had the greatest range (0.061 to 0.201) among the three groups, with a median of 0.125 and a mean value of 0.126±0.047. The partial tear group had the narrowest range (0 to 0.196), with a median of 0.083 and a mean value of 0.093±0.063. The complete rupture group had a range between those of the control group and partial tear group (0 to 0.182), with a median of 0.115 and a mean value of 0.105±0.052. Overall, SMTT was not related to injury severity, suggesting that medial tibiofemoral translation is only minimally affected by ACL injury and thus should be examined in combination with other indicators in the clinical evaluation of injury.

Table 4. Descriptive statistics of standardized medial tibiofemoral translation (positive value indicates anterior tibial translation).

Group	Minimum	Maximum	Median	Mean ± SD
Control	0.061	0.201	0.125	0.126±0.047
Partial tear	0	0.196	0.083	0.093±0.063
Complete rupture	0	0.182	0.115	0.105±0.052

SD, standard deviation.

TRA

There were significant differences in the TRA between the control group, partial tear group, and complete rupture group. The angles in the control group were all positive (6.27±2.35°), indicating external rotation; those of the partial tear group were mainly negative (−4.51±5.88°), indicating limited external rotation; and those of the rupture group were all negative, with a larger absolute value (−6.18±4.21°) and a greater degree of limitation in external rotation (Table 5).

Table 5. Descriptive statistics of tibiofemoral rotation angle (positive value indicates tibial external rotation relative to the femur).

Group	Minimum	Maximum	Median	Mean ± SD
Control	4	14.2	5.7	6.27±2.35
Partial tear	−15	5.9	−4	−4.51±5.88
Complete rupture	−15.2	0.5	−6.8	−6.18±4.21

SD, standard deviation.

Intergroup comparisons

There were significant differences in the SLTT distance and TRA between the three groups (F=17.218 and F=52.013, respectively; P<0.001); in contrast, there was no statistically significant difference in the SMTT distance between the groups (F=1.344; P=0.269). Specifically, the lateral tibiofemoral translation in the partial tear group and rupture group was significantly higher than that in the control group (P<0.05) and increased with the aggravation of injury; meanwhile, the TRA in the partial tear group and rupture group was significantly lower than that in the control group (P<0.05) (Table 6).

Table 6. Comparison of tibiofemoral translation and rotation angle of the normal group, partial tear group, and complete rupture group.

Parameter	Control (n=21)	Partial tear (n=15)	Complete rupture (n=20)	F	P
SLTT	0.016±0.035	0.073±0.059*	0.103±0.051*	17.218	<0.001
SMTT	0.126±0.047	0.093±0.063	0.105±0.052	1.344	0.269
TRA (°)	6.27±2.35	−4.51±5.88*	−6.18±4.21*	52.013	<0.001

Data are presented as mean ± standard deviation. *, P<0.05 vs. control group. SLTT, standardized lateral tibiofemoral translation; SMTT, standardized medial tibiofemoral translation; TRA, tibiofemoral rotation angle.

There were no statistically significant differences in any of the three indicators between the partial tear group and the complete rupture group (Table 7).

Table 7. Comparison of standardized tibiofemoral translation and tibiofemoral rotation angle between the partial tear group and complete rupture group.

Parameter	Partial tear (n=15)	Complete rupture (n=20)	t	P
SLTT	0.073±0.059	0.103±0.051	−1.618	0.115
SMTT	0.093±0.063	0.105±0.052	−0.618	0.541
TRA (°)	−4.51±5.88	−6.18±4.21	0.98	0.335

Data are presented as mean ± standard deviation. SLTT, standardized lateral tibiofemoral translation; SMTT, standardized medial tibiofemoral translation; TRA, tibiofemoral rotation angle.

Cases with accompanying injuries to other stabilizing structures in the partial tear group and rupture group were examined and compared between the groups, but no statistically significant differences in SLTT, SMTT, or TRA were found (Table 8).

Table 8. Subgroup comparison of tibiofemoral kinematic parameters between ACL partial tear and complete rupture in patients with concomitant knee stabilizing structure injuries.

Parameter	Partial tear (n=7)	Complete rupture (n=15)	t	P
SLTT	0.078±0.042	0.104±0.049	−1.2	0.244
SMTT	0.103±0.057	0.111±0.047	−0.366	0.718
TRA (°)	−5.19±4.86	−5.73±4.27	0.265	0.794

Data are presented as mean ± standard deviation. ACL, anterior cruciate ligament; SLTT, standardized lateral tibiofemoral translation; SMTT, standardized medial tibiofemoral translation; TRA, tibiofemoral rotation angle.

Within the injury group, intergroup comparison was performed between patients with isolated ACL injury and those with ACL injury combined with injury to other stabilizing structures, but no statistically significant differences were found in SLTT, SMTT, or TRA between these groups (Table 9).

Table 9. Intergroup comparison of tibiofemoral kinematic parameters between isolated and combined ACL injury.

Parameter	Isolated ACL injury (n=13)	ACL + concomitant injury (n=22)	t	P
SLTT	0.080±0.068	0.096±0.048	−0.815	0.421
SMTT	0.085±0.067	0.109±0.049	−1.22	0.231
TRA (°)	−5.31±6.10	−5.55±4.36	0.136	0.893

Data are presented as mean ± standard deviation. ACL, anterior cruciate ligament; SLTT, standardized lateral tibiofemoral translation; SMTT, standardized medial tibiofemoral translation; TRA, tibiofemoral rotation angle.

Reliability analysis of measurements

Extremely high reliability for measurements of SLTT, SMTT, and TRA was observed for both interobserver agreement between the two radiologists and intraobserver agreement for radiologist A at two separate time points. The ICCs for all three indicators were >0.95, all 95% CIs were close to 1, and all P values were <0.001. This indicates that the measurement method in this study has strong consistency and limited error and that it can provide stable, reliable measurement results for data support in statistical analysis and clinical evaluation (Tables 10,11).

Table 10. Reliability analysis from measurements by two radiologists.

Parameter	Radiologist A	Radiologist B	ICC	95% CI	P
SLTT	0.062±0.061	0.060±0.060	0.964	0.939–0.979	<0.001
SMTT	0.110±0.054	0.110±0.050	0.995	0.992–0.997	<0.001
TRA	−1.06±7.09	−1.19±6.82	0.991	0.985–0.995	<0.001

Data are presented as mean ± standard deviation, unless otherwise indicated. CI, confidence interval; ICC, intraclass correlation coefficient; SLTT, standardized lateral tibiofemoral translation; SMTT, standardized medial tibiofemoral translation; TRA, tibiofemoral rotation angle.

Table 11. Intraobserver agreement test from radiologist A’s measurements.

Parameter	Radiologist A: first measurement	Radiologist A: second measurement	ICC	95% CI	P
SLTT	0.062±0.061	0.063±0.060	0.996	0.993–0.998	<0.001
SMTT	0.110±0.054	0.110±0.050	0.997	0.995–0.998	<0.001
TRA	−1.06±7.09	−1.09±6.92	0.992	0.987–0.995	<0.001

Data are presented as mean ± standard deviation, unless otherwise indicated. CI, confidence interval; ICC, intraclass correlation coefficient; SLTT, standardized lateral tibiofemoral translation; SMTT, standardized medial tibiofemoral translation; TRA, tibiofemoral rotation angle.

ROC curve analysis of diagnostic efficacy

ROC curve analysis of the injury group (combination of partial tear group and complete rupture group) and the control group indicated that the TRA had the highest diagnostic efficacy among the single indicators. It had an AUC as high as 0.985 (95% CI: 0.955–1.000; P<0.001), a sensitivity of 100.0%, a specificity of 97.1%, a Youden index of 0.971, and an optimal cutoff value of 2.65°, indicating it could efficiently distinguish ACL injury from a healthy knee. The SLTT had good diagnostic efficacy, with an AUC of 0.874 (95% CI: 0.783–0.965; P<0.001), a sensitivity of 85.7%, a specificity of 74.3%, a Youden index of 0.600, and an optimal cutoff value of 0.055. The SMTT had an AUC of 0.609 (95% CI: 0.458–0.760; P=0.182), which was not statistically significant, with a Youden index of only 0.285, indicating limited diagnostic efficacy and unsuitability to diagnosing ACL injury. Finally, the combined diagnosis of SLTT and TRA achieved the highest overall diagnostic efficacy, with an AUC of 0.996 (95% CI: 0.986–1.000; P<0.001), a sensitivity of 100.0%, a specificity of 97.1%, a Youden index of 0.971. Its diagnostic efficacy was slightly higher than that of the single TRA, indicating that combined detection can further improve the accuracy in the diagnosis of ACL injury (Table 12).

Table 12. ROC curve analysis of tibiofemoral translation and rotation angle in differentiating injured and healthy knees.

Parameter	AUC (95% CI)	P	Sensitivity (%)	Specificity (%)	Youden index	Cutoff value
SLTT	0.874 (0.783–0.965)	<0.001	85.7	74.3	0.600	0.055
SMTT	0.609 (0.458–0.760)	0.182	71.4	57.1	0.285	0.102
TRA	0.985 (0.955–1.000)	<0.001	100.0	97.1	0.971	2.65
SLTT + TRA	0.996 (0.986–1.000)	<0.001	100.0	97.1	0.971	0.772

The optimal cut-off value was determined by the maximum Youden index. AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic; SLTT, standardized lateral tibiofemoral translation; SMTT, standardized medial tibiofemoral translation; TRA, tibiofemoral rotation angle.

Discussion

The incidence of sports injuries is increasing year on year, with ACL injury being common among them. When injured, the patient’s athletic ability and health will decline, and further complications may occur. Studies have shown that the majority of patients with ACL injury, especially young and active patients, opt for surgical treatment (1). However, the optimal treatment for an ACL tear remains controversial, and surgery is not suitable for all patients (15). Depending on age and activity level, a variety of treatment methods (including nonsurgical and surgical techniques) can be used to help athletes resume activity in their sport (16).

The assessment of knee-joint stability after ACL injury is clinically challenging. Moreover, the association between MRI-detected structural injury and actual instability remains unclear, and some patients with ACL tears do not exhibit obvious symptoms of knee-joint instability (6,8). A systematic review and meta-analysis published in 2022 focused specifically on the performance of various clinical tests in diagnosing acute ACL injury. The pooled sensitivities of the Lachman test, anterior drawer test, pivot shift test, and lever test were 79% (95% CI: 74–83%), 78% (95% CI: 73–83%), 55% (95% CI: 49–60%), and 82% (95% CI: 76–87%), respectively; the pooled specificities were 91% (95% CI: 87–94%), 91% (95% CI: 86–95%), 96% (95% CI: 91–98%), and 88% (95% CI: 82–93%), respectively, while the corresponding AUC values were 0.927, 0.919, 0.891, and 0.904, respectively. These findings support the effectiveness of physical examination tests in diagnosing ACL injury, among which the pivot shift test is valuable for confirming acute ACL injury (17). However, clinical tests have inherent limitations, such as high subjectivity and limited interobserver reproducibility (18). Meanwhile, although conventional MRI has high diagnostic accuracy, it is highly dependent on scanning quality and radiologists’ experience and cannot provide a quantitative assessment of the joint’s functional stability.

Based on the physiological characteristics of the terminal extension of the knee joint, this study leveraged the advantages of MRI—such as being unaffected by surrounding muscles, ligaments, or soft tissues and being highly accurate and noninvasive—to quantitatively evaluate the joint stability of patients with ACL injury (14). The results showed that the tibial external rotation angle during hyperextension in the control group was 6.27±2.35°, which was similar to that of the normal knee joint (8.1±3.3°) reported by Griffith et al. (11), confirming that MRI during knee hyperextension can accurately capture the SHM. This result is significant because this measurement can serve as a reliable physiological reference value for the clinical evaluation of SHM function. In a 2013 study, Ng et al. (19) inferred the TRA by measuring the difference in the degree of anterior translation of the medial and lateral tibiofemoral joints between the extended and flexed positions. They found that normal SHM was observed in 80% of participants with intact ACLs, whereas loss of the SHM was present in 51% of patients with ACL tears, and this loss was significantly correlated with a positive pivot shift test. These findings confirmed that ACL injury is closely associated with abnormal tibiofemoral kinematics and clinical knee instability, which was consistent with the conclusions of our study. However, this approach only enables indirect inference of tibiofemoral joint rotation and not direct measurement (19).

In our study, there was greater anterior tibial translation and impairment of external rotation in knees with partial ACL tears. At the optimal cutoff value, 9 of the 15 patients with partial ACL tears had excessive anterior tibial translation, and 14 had limited external rotation. In the 20 patients with complete ACL rupture, 17 had excessive anterior translation of the lateral tibia, and 20 had limited external rotation, which was more severe than that in the partial tear group. In addition, intergroup comparison showed that the TRA had the highest diagnostic efficacy among the single indicators, while the lateral tibiofemoral translation had good diagnostic efficacy. In contrast, medial tibiofemoral translation, despite demonstrating a degree of diagnostic efficacy, was not found to be a suitable indicator for injury. Functional abnormalities caused by ACL tears are often multidirectional. Previous MRI studies have typically focused on morphological changes of the ligament, with tibial anterior translation being used as an indirect sign for reference, while the TRA has largely been ignored (20). According to the optimal cutoff value derived from our study, a tibial external rotation angle relative to the femur of <2.65° in the hyperextension position indicates SHM abnormalities and knee rotational instability. This cutoff value is close to the physiological range of tibial external rotation during terminal extension of the normal knee joint reported in previous studies (11,12). When this mechanism is impaired, patients cannot acquire passive stability during terminal stance, resulting in compensatory gait. This not only compromises gait efficiency but also disrupts the kinetic chain of the entire lower extremity. Meanwhile, to maintain balance, the load on the hip and ankle joints increases accordingly, which may trigger secondary pain in the hip, foot, or ankle. In addition, abnormal knee kinematics can alter the transmission path of ground reaction forces, further exacerbating nonphysiological wear of the joint. A further critical effect is the disruption of patellofemoral joint biomechanics. Studies have shown that abnormal tibial rotation patterns can indirectly lead to deviation of the patellar alignment within the femoral trochlear groove by altering the force line of the quadriceps femoris, which commonly manifests as a lateral tracking abnormality. This stress concentration can induce patellofemoral pain and may accelerate cartilage degeneration in this region (21,22).

In the partial tear and complete rupture groups, 7 and 15 cases, respectively, had accompanying injuries to other stabilizing structures. Comparison between the partial tear and complete rupture subgroups with concomitant injuries indicated no statistically significant differences in lateral tibiofemoral translation, medial tibiofemoral translation, or TRA. Similarly, comparison between patients with isolated ACL injury and those with ACL injury combined with injuries to other stabilizing structures also revealed no statistically significant differences in these indicators. These two intergroup comparisons highlight the core role of the passive tension of the ACL in maintaining knee joint stability during the terminal extension state and suggest that combined injury to other stabilizing structures has no significant superimposed effect on the kinematic parameters in the hyperextension position. Despite this, it should be noted that these injuries can adversely affect the long-term prognosis of patients. It has been reported that the incidence of knee osteoarthritis after isolated ACL injury ranges from 0% to 13%, whereas it ranges from 21% to 48% after ACL injury combined with meniscus injury (23). In most cases, when concomitant injury to the meniscus or collateral ligaments is present, surgical reconstruction is routinely recommended to prevent further damage to these structures. Studies have also indicated that nonsurgical treatment may be a viable option for patients with isolated ACL injury (24). Nevertheless, multiple factors should be taken into account in the determination of treatment strategy, including the patient’s age, activity level, and joint stability. Comparison between the partial tear group and complete rupture group revealed no statistically significant differences in lateral tibiofemoral translation, medial tibiofemoral translation, and TRA regardless of the presence of combined injuries to other stabilizing structures. This finding indicates that an anatomically partial tear does not equate to preserved functional joint stability, and a considerable proportion of patients with partial ACL tears develop knee instability in the terminal extension position comparable to that observed in patients with complete ruptures. In this context, the optimal cutoff values of lateral tibiofemoral translation and TRA on MRI during hyperextension provide an objective basis for the assessment of joint stability, which is an advantage not available with conventional MRI. There is an abundance of data confirming that knee-joint stability can be restored in patients with ACL injury through a properly constructed rehabilitation program (9). Additionally, MRI of the hyperextended knee can provide objective kinematic parameters during the follow-up of these patients, assisting clinicians in adjusting the rehabilitation regimen in a timely manner.

There were several limitations in this study that should be acknowledged. The most notable was that some patients with acute ACL injury had knee extension dysfunction caused by pain, joint effusion, quadriceps inhibition, or mechanical block. Such patients could not undergo standardized hyperextension position MRI examination, which means this quantitative evaluation technique is not applicable to all patients with acute ACL injury, thus limiting the external generalizability of our study conclusions and the widespread clinical application of this technique. A large-sample clinical study conducted in 2025 showed that among 10,692 patients scheduled for ACL reconstruction enrolled between 2002 and 2023, 6.3% (677/10,692) had varying degrees of preoperative knee extension dysfunction (25). In our study, patients who failed to achieve full knee extension were excluded from MRI scanning during hyperextension of the knee. These patients only underwent MRI in the conventional flexion position. We sought to clarify the etiology of extension dysfunction through imaging and clinical physical examinations to determine whether it was attributable to mechanical blockage or functional/reactive restriction. The former is the primary indication for surgery, whereas for the latter, rehabilitation should be performed to restore full knee extension as much as possible, regardless of whether the patient receives surgical or nonsurgical treatment. Studies have indicated that preoperative extension deficit is associated with an increased risk of extension deficit after ACL reconstruction (26).

Another limitation is that the optimal cutoff value was derived from a single center with a small sample size and was not cross-validated with the results of clinical tests; for instance, the interval of the values measured in this study corresponding to mild instability on clinical tests was not defined. Moreover, due to the lack of external validation in multicenter settings or across a variety of populations, these values may not be applicable to populations with different ages, ethnicities, or baseline knee laxities. In the future, we will conduct a large-sample, multicenter, prospective study to complete external validation of the diagnostic model and cutoff value, further optimize the diagnostic threshold based on the characteristics of different populations, and improve the generalizability of this method in clinical practice. Meanwhile, this quantitative tool can be used to evaluate the efficacy of certain types of rehabilitation. In future work, we can collaborate with the rehabilitation department to conduct a randomized controlled trial comparing the efficacy of traditional quadriceps-strengthening training and neuromuscular re-education techniques targeting rotational control, with restoration of the SHM angle as a primary outcome measure. This would help clarify whether using MRI of the knee in hyperextension for serial quantitative assessment can optimize rehabilitation regimens aimed at restoring the joint’s native biomechanical properties and provide personalized rehabilitation strategies for patients.

In addition, there remain potential challenges in the popularization and application of this technique in routine clinical practice. Conventional knee MRI examinations are mostly performed in the flexion position, while hyperextension position examination requires additional standardized patient positioning and verification by technologists. Furthermore, differences in positional cooperation among patients with acute injury may affect examination efficiency; however, the scanning sequences used in this technique are completely consistent with those of conventional knee MRI, and the total examination time is no different from that of conventional MRI; thus, the impact on the routine clinical workflow would be minimal. In terms of cost, this technique does not require additional investment in equipment, coils, or consumables. It can be performed with a conventional clinical 3.0T MRI system and a knee-specific coil, without imposing an additional economic burden on hospitals and patients. Finally, regarding personnel training requirements, scanning technologists need to master the standardized parameters for the hyperextension position, and radiologists and clinicians need to become proficient in the measurement of the quantitative indicators proposed in this study. To this end, we have provided detailed measurement criteria and descriptions, and the measurement method demonstrated excellent interobserver consistency (ICC >0.99). Overall, we believe this technique can be effectively learned through brief standardized training and requires a low technical threshold.

Conclusions

By simulating the physiological characteristics of the knee joint in terminal extension, MRI in the hyperextension position can quantify the abnormal kinematic parameters of the tibiofemoral joint after acute ACL injury and accurately evaluate the rotational and translational stability of the joint. Notably, the TRA has extremely high sensitivity (100%) and specificity (97.1%) in diagnosing injury. While being straightforward and noninvasive, this technique does not require more time than does conventional MRI examination, and thus it may serve as a practical, reliable, and objective evaluation method for clinical practice.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This single-center retrospective case-control study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Medical Ethics Committee of Central People’s Hospital of Zhanjiang (No. ZX-2025051-01), and all participants provided written informed consent.

Data Sharing Statement

Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2628/dss

qims-16-06-498-dss.pdf ^{(88.7KB, pdf)}

DOI: 10.21037/qims-2025-1-2628