Aiming at the problems of low accuracy and large difference of segmentation boundary distance in anterior cruciate ligament (ACL) image segmentation of knee joint, this paper proposes an ACL image segmentation model by fusing dilated convolution and residual hybrid attention U-shaped network (DRH-UNet). The proposed model builds upon the U-shaped network (U-Net) by incorporating dilated convolutions to expand the receptive field, enabling a better understanding of the contextual relationships within the image. Additionally, a residual hybrid attention block is designed in the skip connections to enhance the expression of critical features in key regions and reduce the semantic gap, thereby improving the representation capability for the ACL area. This study constructs an enhanced annotated ACL dataset based on the publicly available Magnetic Resonance Imaging Network (MRNet) dataset. The proposed method is validated on this dataset, and the experimental results demonstrate that the DRH-UNet model achieves a Dice similarity coefficient (DSC) of (88.01±1.57)% and a Hausdorff distance (HD) of 5.16±0.85, outperforming other ACL segmentation methods. The proposed approach further enhances the segmentation accuracy of ACL, providing valuable assistance for subsequent clinical diagnosis by physicians.
Humans ; Magnetic Resonance Imaging/methods* ; Anterior Cruciate Ligament/diagnostic imaging* ; Image Processing, Computer-Assisted/methods* ; Knee Joint/diagnostic imaging* ; Neural Networks, Computer ; Algorithms ; Deep Learning

OBJECTIVE: To retrospectively analyze the healing degree of graft in patients who underwent over-the-top reconstruction of the anterior cruciate ligament(ACL) using autologous tendon 1 year postoperatively to find out possible risk factors. METHODS: From January 2017 to March 2023, 207 patients with more than 1 year of follow-up and 1-year postoperative MRI were selected from that receiving ACL autograft tendon over-the-top reconstruction surgery including 137 males and 70 females, with a mean age of (28.01±9.71) years rangeding 12 to 56 years old;time from trauma to surgery was (5.80±3.78) months. All patients had intra-operative pivoshift tested asⅠtoⅡdegree. The 1-year postoperative magnetic resonance imaging (MRI) results were categorized by Howell's method, classifying degreesⅠandⅡas group BHD (better healing degree), while degrees Ⅲ and Ⅳ as group PHD (poorer healing degree). Potential influencing factors such as age, gender, whether it was a sports injury, time from injury to surgery, time to return to sports after surgery, body mass index (BMI), Beighton score, knee hyper-extension, diameter of the graft, whether it was remnant-preserving or not, whether the medial and lateral meniscus were resected, sutured or not, posterior lateral tibial slope, intercondylar notch width ratio (ICNWR), and lateral femoral condyle depth ratio (LFCDR) were analyzed by univariate analysis and Logistic regression to find the influencing factors. RESULTS: All patients were follow-up time for (18.59±4.63) months, and all the patients had good wound healing without any infection after surgery. There were 3 cases of re-ruptrure and 2 cases with secondary medial meniscus barrel handle-like tear 1 year after surgery, which received arthroscopic surgical treatment. Univariate analysis showed statistically significant difference in time from injury to surgery, time from surgery to return to sports, ICNWR, posterior lateral tibial slope, and BMI between the two groups(P<0.05). Multi-factorial regression analysis showed that smaller ICNWR, OR=0.684, 95%CI(0.528, 0.886), P=0.004; larger posterior lateral tibial slope, OR=1.557, 95%CI(1.222, 1.984), P<0.001;larger BMI, OR=1.724, 95%CI(1.369, 2.172), P<0.001;and shorter return-to-sport time<6 months vs. 6 to 12 months, OR=0.167, 95%CI(0.041, 0.672), P=0.012;<6 months vs.>12 months, OR=0.022, 95%CI(0.004, 0.108), P<0.001 might be risk factors for poor MRI healing degree 1 year postoperatively. CONCLUSION In patients received ACL autograft tendon over-the-top reconstruction, smaller ICNWR, larger posterior lateral tibial slope, larger BMI, and shorter time to return to sports may all be high-risk factors for poor graft healing degree on MRI 1 year postoperatively, and the surgical technique and the postoperative rehabilitation are supposed to be individualized for such patients.
Humans ; Male ; Female ; Adult ; Magnetic Resonance Imaging ; Anterior Cruciate Ligament Reconstruction/methods* ; Middle Aged ; Adolescent ; Young Adult ; Tendons/transplantation* ; Child ; Risk Factors ; Retrospective Studies ; Transplantation, Autologous ; Anterior Cruciate Ligament Injuries/diagnostic imaging* ; Anterior Cruciate Ligament/diagnostic imaging* ; Autografts ; Wound Healing

OBJECTIVE: To explore correlation between femoral mechanical axis and Blumensaat line (FMBL) angle of knee joint (angle between Blumensaat line and femoral mechanical axis), α angle (angle between Blumensaat line and axis of distal femur in sagittal plane) on EOS biplane imaging and non-contact anterior cruciate ligament(ACL) injury, and evaluate angle for its accuracy in predicting the populations prone to non-contact ACL injury. METHODS: From February 2018 to October 2020, EOS imaging and clinical data from 88 patients (176 knees) with unilateral non-contact ACL injury were retrospectively analyzed, including 53 males and 35 females, aged from 18 to 45 years old with an average of (30.3±6.2) years old, 48 patients on the left side and 40 patients on the right side. The patients were divided into ACL-affected group and ACL-health group according to side of ACL injuries, and 51 patients (51 knees) with non-ACL identified from EOS database were included in normal control group, including 28 males and 23 females, aged from 20 to 44 years old with an average of (31.6±5.5) years old, 26 patients on the left side and 25 patients on the right side. Full-length EOS imaging of skeleton extremitatis inferioris among three groups were reconstructed to 3D images of skeletal system with EOS software, and then FMBL angle and α angle were measured on the images. Univariate binary Logistic regression analysis was performed to determine the influence of the univariate(FMBL angle or α angle) on ACL status(normal or torn). And the angle cutoff value for univariate was selected based on receiver operating characteristics curve (ROC) to got the best accuracy. RESULTS: There was no statistically significant difference in age, gender and side distribution between ACL-injured group and normal control group(P>0.05). Statistical analyses (one-way ANOVA) indicated no significant difference in FMBL angle between ACL-injured knee group (32.8±2.3)° and ACL-injured contralateral knee group(32.5±2.3)°(P>0.05), but the values between two groups were significantly lower than that in normal control group (37.0±2.0)°(P<0.001). There was no statistically significant difference in α angle among three groups (P>0.05). Univariate binary Logistic regression analysis demonstrated that FMBL angle was risk factor for non-contact ACL injury[OR=0.433, 95%CI(0.330, 0.569), P<0.001]. The area under ROC curve for FMBL angle was 0.909[95%CI(0.861, 0.958), P<0.001], and the sensitivity and specificity were 70.5% and 98.0% respectively, cut-off value was 33.7°. CONCLUSION FMBL angle formed by Blumensaat line and femoral mechanical axis is one of the risk factors for non-contact ACL injury and has good predictive accuracy. The general population with FMBL angle below 33.7° may be increased risk for ACL injury.
Male ; Female ; Humans ; Adolescent ; Young Adult ; Adult ; Middle Aged ; Anterior Cruciate Ligament Injuries/diagnostic imaging* ; Retrospective Studies ; Imaging, Three-Dimensional ; Magnetic Resonance Imaging/methods* ; Anterior Cruciate Ligament/diagnostic imaging* ; Knee Joint/diagnostic imaging*

OBJECTIVE: To compare the posterior cruciate ligament(PCL) index with six different measurement methods, and analyze and verify its clinical diagnostic value in anterior cruciate ligament (ACL) injury. METHODS: The Magnetic resonance imaging (MRI) data of 225 knee joints in our hospital from May 2018 to March 2022 were retrospectively analyzed, aged from 18 to 60 years old, with a median of 32 years old. On the sagittal MRI images of 114 patients with ACL injury and 111 patients with intact ACL, Measure the straight-line distance (A) between the femoral attachment point and the tibial attachment point of the PCL on the MRI sagittal image and the maximum vertical distance (B) between the straight line and the arcuate mark point of the PCL on the sagittal image, calculate the PCL index and evaluate the diagnostic value of the PCL index for ACL injury. RESULTS: The PCL index of the ACL normal group and the ACL injury group were statistically described. There was no significant difference in PCL index 1, 2, 3 and 6 between the two groups(P>0.05). The difference of PCL index 4 and 5 between the two groups was statistically significant (P<0.001). This study only found that the PCL index 2, 6 in the ACL normal group had a negative correlation with the patient's age (correlation coefficient=-0.213, -0.819;P<0.05), and the PCL index 5 in the ACL injury group was significantly correlated with the patient's body mass index(BMI)had a negative correlation (correlation coefficient=-0.277, P<0.05). CONCLUSION The change of PCL index is helpful for the diagnosis of ACL injury, PCL index 4 and 5 can be used as effective reference indexes for diagnosing ACL injury in clinic.
Humans ; Adolescent ; Young Adult ; Adult ; Middle Aged ; Posterior Cruciate Ligament/diagnostic imaging* ; Anterior Cruciate Ligament Injuries/diagnostic imaging* ; Anterior Cruciate Ligament ; Retrospective Studies ; Knee Joint ; Magnetic Resonance Imaging/methods*

OBJECTIVE: To investigate the diagnostic value of multislice spiral CT (MSCT) and MRI in occult fracture of knee joint with meniscus and ligament injury. METHODS: From January 2020 to March 2021, 63 patients with knee occult fracture with meniscus and ligament injury, including 41 males and 22 females, aged from 21 to 67 years old, with an average of (44.35±8.77) years old, the course of disease ranged from 1 to 6 days, with an average of (4.64±1.75) days, the body mass index (BMI) was (19.85±2.78) kg/m². MSCT and MRI data of 63 patients were collected and statistically analyzed to evaluage their diagnist value. RESULTS: The detection of MRI for occult knee fravtures with meniscus and ligament injury, joint cavity effusion, bone marrow edema, and articular surface injury were 100.00% (63/63), 95.24% (60/63), 42.86% (27/63) and 36.51% (23/63), respectively. The detection rates of MSCT were 49.21% (31/63), 41.27% (26/63), 0.00% (0/63) and 1.59% (1/63), respectively, significantly lwver than that of MRI (P<0.05). The diagnostic sensitivity, specificity and accuracy of MRI were significantly higher than those of MSCT(P<0.05). CONCLUSION The sensitivity, specificity and accuracy of magnetic resonance imaging in the diagnosis of occult fracture of knee joint with meniscus and ligament injury are significantly better than that of MSCT. MRI has higher accuracy in the diagnosis of peripheral tissue diseases such as joint cavity, articular surface and bone marrow, and can reduce the risk of clinical misdiagnosis.
Male ; Female ; Humans ; Young Adult ; Adult ; Middle Aged ; Aged ; Tibial Meniscus Injuries/diagnostic imaging* ; Fractures, Closed/diagnostic imaging* ; Arthroscopy/methods* ; Knee Injuries/diagnostic imaging* ; Knee Joint/diagnostic imaging* ; Magnetic Resonance Imaging/methods* ; Ligaments ; Meniscus ; Tomography, Spiral Computed ; Fractures, Bone ; Anterior Cruciate Ligament Injuries

OBJECTIVE: To investigate the correlation between the quadriceps cross-sectional area (CSA) and quadriceps muscle volume (QMV) at different horizontal levels from the upper edge of the patella, and to determine the best observation position. METHODS: Thigh magnetic resonance imaging (MRI) images of 22 Chinese young men [age: (29±6) years] with anterior cruciate ligament (ACL) rupture were examined. The CSA was measured at 18, 15, and 12 cm above the upper edge of the pate-lla (denoted by CSA-18, CSA-15 and CSA-12 respectively), and the QMV and CSA were determined by semiautomatic segmentation. A curve model was established to estimate QMV. Bland-Altman analysis was performed to determine the confidence limits of the volumes. RESULTS: On the unaffected side, the mean QMV was (1 944.45±323.77) cm³. The quadriceps CSA at the upper edge of the patella at 18, 15, and 12 cm was (80.80±12.16) cm², (77.53±12.03) cm², and (72.68±10.51) cm², respectively. The coefficients of determination (R²), ascertained using curve estimation models, for the 3 positions were 0.819, 0.755, and 0.684 (P < 0.001), and the standard deviations of the volume estimated value (SEE) were 7.4%, 8.7%, and 9.8%. The fitting equations of the three horizontal positions were all good, but the fitting degree of CSA-18 was the highest. The Bland-Altman scatter plot showed that the arithmetic means of the QMV at 18, 15 and 12 cm from the upper edge of the patella 0.8 cm³, -1.1 cm³, and 0.9 cm³ and 95% limits of agreement (LoA) were (-268.8, 270.5), (-315.2, 313.1), and (-355.7, 357.5), respectively. The estimated QMV was in good agreement with the measured value. The difference between the estimated CSA-18 and measured values was the smallest. The results on the affected side were consistent. CONCLUSION The correlation between QMV and CSA in the young men with the upper edge of patella as baseline was reliable and consistent. Among them, CSA-18 had the highest correlation with the QMV. However, different observation sites could be selected for different injuries of the quadriceps.
Adult ; Anterior Cruciate Ligament Injuries/surgery* ; Anterior Cruciate Ligament Reconstruction ; Humans ; Magnetic Resonance Imaging ; Male ; Patella ; Quadriceps Muscle/diagnostic imaging* ; Young Adult

The anterior cruciate ligament (ACL) is an important stabiliser of the knee and is commonly torn in sports injuries. Common indications for imaging after ACL reconstruction include persistent symptoms, limitation of motion and re-injury. Important postoperative complications include graft failure, impingement, arthrofibrosis and graft degeneration. This article aimed to familiarise the radiologist with magnetic resonance (MR) imaging appearances of properly positioned intact ACL grafts and to provide a comprehensive review of MR imaging features of complications following ACL reconstruction.
Anterior Cruciate Ligament ; diagnostic imaging ; surgery ; Anterior Cruciate Ligament Reconstruction ; adverse effects ; methods ; Humans ; Knee Joint ; diagnostic imaging ; surgery ; Magnetic Resonance Imaging ; Postoperative Complications ; diagnostic imaging ; Transplants ; diagnostic imaging ; Wound Healing

OBJECTIVETo explore diagnostic value of MRI on posterior root tear of medial and lateral meniscus.

METHODSFrom January 2012 to January 2016, clinical data of 43 patients with meniscal posterior root tear confirmed by arthroscopy were retrospective analyzed, including 25 males and 18 females, aged from 27 to 69 years old with an average age of(42.5±8.3)years old;27 cases on the right side and 16 cases on the left side. MRI examinations of 43 patients with tear of posterior meniscus root confirmed by knee arthroscopies were retrospectively reviewed. MRI images were double-blinded, independently, retrospectively scored by two imaging physicians. Sensitivity, specificity and accuracy of MRI diagnosis of lateral and medial meniscus posterior root tear were calculated, and knee ligament injury and meniscal dislocation were calculated.

RESULTSForty-three of 143 patients were diagnosed with meniscus posterior root tears by arthroscopy, including 19 patients with lateral tears and 24 patients with medial tears. The sensitivity, specificity and accuracy in diagnosis of posterior medial meniscus root tears for doctor A were 91.67%, 86.6% and 83.9% respectively, and for doctor B were 87.5%, 87.4% and 87.4%, 19 patients with medial meniscal protrusion and 2 patients with anterior cruciate ligament tear. The sensitivity, specificity and accuracy in diagnosis of posterior lateral meniscus root tears for doctor A were 73.7%, 79.9% and 79% respectively, and for doctor B were 78.9%, 82.3% and 82.5%, 4 patients with lateral meniscus herniation and 16 patients with cruciate ligament tear. Kappa statistics for posterior medial meniscus root tears and posterior lateral meniscus root tears were 0.84 and 0.72.

CONCLUSIONSMRI could effectively demonstrate imaging features of medial and lateral meniscal root tear and its accompanying signs. It could provide the basis for preoperative diagnosis of clinicians, and be worthy to be popularized.

Adult ; Aged ; Anterior Cruciate Ligament Injuries ; diagnostic imaging ; Arthroscopy ; Female ; Humans ; Knee Injuries ; diagnostic imaging ; Magnetic Resonance Imaging ; Male ; Menisci, Tibial ; diagnostic imaging ; Middle Aged ; Retrospective Studies ; Sensitivity and Specificity ; Tibial Meniscus Injuries ; diagnostic imaging

Background: Nowadays, the anterior cruciate ligament (ACL) injury has been one of the most common diseases of the knee joint. The relationships between the ACL injury and the anatomical structures are still controversial. This study aimed to identify the anatomical risk factors of ACL injury by magnetic resonance imaging (MRI) of the knee. Methods: This was a retrospective study of 125 patients undergoing primary ACL reconstruction between July 2013 and May 2017. Another 125 patients without any organic knee joint injury were served as controls. The shape of intercondylar notch, the intercondylar notch width index, the intercondylar notch height index, the α angle, the β angle, and the medial and lateral tibial plateau slope were measured with MRI and compared. The data were compared by binary logistic regression to find the risk factors. Results: The two groups differed in the proportion of male patients (70.4% vs. 52.0%, χ = 8.911, P = 0.003), but gender was excluded as a risk factor for ACL injury with regression analysis (odds ratio = 1.476, 95% confidence interval [CI]: 0.689-3.160, P = 0.317). The injured group was found to have a smaller notch width index (95% CI = 7.960E-23-2.154E-9, P < 0.001), a larger β angle (95% CI = 1.311-1.785, P < 0.001), and a larger lateral tibial plateau slope (95% CI = 1.201-1.683, P < 0.001). The cutoff values of notch width index, β angle, and the lateral tibial plateau slope were 0.252, 38.5°, and 7.5°, respectively. Conclusions: In this study, a narrow intercondylar notch (intercondylar notch width index <0.252), a larger lateral tibial slope (>7.5°), and larger β angle (>38.5°) might be the factors associated with ACL injury. Trial Registration ChiCTR-RRC-17014116; http://www.chictr.org.cn/showproj.aspx?proj=24119.
Adult ; Anterior Cruciate Ligament Injuries ; diagnostic imaging ; etiology ; Anterior Cruciate Ligament Reconstruction ; Female ; Humans ; Knee Joint ; diagnostic imaging ; Logistic Models ; Magnetic Resonance Imaging ; Male ; Retrospective Studies ; Risk Factors

OBJECTIVESTo investigate whether tibial intercondylar hyperplasia can cause anterior cruciate ligament (ACL) injuries.

METHODSFrom February 2009 to October 2012, 58 cases (including 70 knees) with knee osteoarthritis (KOA) without trauma history were selected from the out-patient department of orthopaedics in Guang'anmen Hospital attached to China Academy of Chinese Medical Science. All cases were tested by MRI scanning, then vertical height of the medial and lateral intercondylar spine in coronal position and the area of abnormal signal of ACL in sagittal position were tested from patients. Correlation between the height of intercondylar spine and the area of abnormal signal of ACL were analyzed.

RESULTSFives cases of 58 cases( 70 knees) were excluded from episome in knee joint, and other 65 knees were in accordance with the inclusion criteria. Among them, there were 10 males and 48 females aged from 39 to 79 years old with an average of 61 years old. The vertical height of the medial intercondylar spine was (10.02 +/- 1.46) mm and lateral was (8.92 +/- 1.69) mm. The area of abnormal signal ACL was (318.42 +/- 130.10) mm2. There was positive correlation between the height of medial intercondylar spine and the area of abnormal signal of ACL (r=0.06, P=0.00). But there was no correlation between lateral intercondylar spine and the area of abnormal signal of ACL (P=0.10).

CONCLUSIONThe height of medial intercondylar spine and the area of abnormal signal of ACL is positively correlated. The clinical significance is in study, while the clinical meaning still need further study.

Adult ; Aged ; Anterior Cruciate Ligament ; abnormalities ; diagnostic imaging ; surgery ; Female ; Humans ; Hyperplasia ; diagnostic imaging ; surgery ; Magnetic Resonance Imaging ; Male ; Middle Aged ; Osteoarthritis, Knee ; diagnostic imaging ; surgery ; Radiography ; Tibia ; diagnostic imaging ; pathology ; surgery