This study assessed the accuracy of surface-measured OpenSim-derived tibiofemoral kinematics in functional activities. Ten subjects with unilateral, isolated grade II PCL deficiency performed level running and stair ascent. A dynamic stereo radiography (DSX) system and a Vicon motion capture system simultaneously measured their knee or lower extremity movement. Surface marker motion data from the Vicon system were used to create subject-specific models in OpenSim and derive the tibiofemoral kinematics. The surface-measured model-derived tibiofemoral kinematics in all 6 degrees of freedom (DOFs) were then compared with those measured by the DSX as the benchmarks. The differences between surface- and DSX-measured tibiofemoral kinematics were found to be substantial: the overall mean (±SD) RMS differences during running were 9.1±3.2°, 2.0 ± 1.2°, 6.4 ± 4.5° for the flexion-extension, abduction-adduction, and internal-external rotations, and 7.1± 3.2mm, 8.8± 3.7mm, and 1.9± 1.2mm for anterior-posterior, proximal-distal, and medial-lateral translations. The differences were more pronounced in the relatively higher speed running than in stair ascent. It was also found that surface-based measures significantly underestimated the mean as well as inter-subject variability of the differences between PCL-injured and intact knees in abduction-adduction, internal-external rotation, and anterior-posterior translation.

Availability of accurate three-dimensional (3D) kinematics of lumbar vertebrae is necessary to understand normal and pathological biomechanics of the lumbar spine. Due to the technical challenges of imaging the lumbar spine motion in vivo, it has been difficult to obtain comprehensive, 3D lumbar kinematics during dynamic functional tasks. The present study demonstrates a recently developed technique to acquire true 3D lumbar vertebral kinematics, in vivo, during a functional load-lifting task. The technique uses a high-speed dynamic stereo-radiography (DSX) system coupled with a volumetric model-based bone tracking procedure. Eight asymptomatic male participants performed weight-lifting tasks, while dynamic X-ray images of their lumbar spines were acquired at 30 fps. A custom-designed radiation attenuator reduced the radiation white-out effect and enhanced the image quality. High resolution CT scans of participants' lumbar spines were obtained to create 3D bone models, which were used to track the X-ray images via a volumetric bone tracking procedure. Continuous 3D intervertebral kinematics from the second lumbar vertebra (L2) to the sacrum (S1) were derived. Results revealed motions occurring simultaneously in all the segments. Differences in contributions to overall lumbar motion from individual segments, particularly L2-L3, L3-L4, and L4-L5, were not statistically significant. However, a reduced contribution from the L5-S1 segment was observed. Segmental extension was nominally linear in the middle range (20%-80%) of motion during the lifting task, but exhibited nonlinear behavior at the beginning and end of the motion. L5-S1 extension exhibited the greatest nonlinearity and variability across participants. Substantial AP translations occurred in all segments (5.0 ± 0.3 mm) and exhibited more scatter and deviation from a nominally linear path compared to segmental extension. Maximum out-of-plane rotations (<1.91 deg) and translations (<0.94 mm) were small compared to the dominant motion in the sagittal plane. The demonstrated success in capturing continuous 3D in vivo lumbar intervertebral kinematics during functional tasks affords the possibility to create a baseline data set for evaluating the lumbar spinal function. The technique can be used to address the gaps in knowledge of lumbar kinematics, to improve the accuracy of the kinematic input into biomechanical models, and to support development of new disk replacement designs more closely replicating the natural lumbar biomechanics.

In this paper, we present a new methodology for subject-specific finite element modeling of the tibiofemoral joint based on in vivo computed tomography (CT), magnetic resonance imaging (MRI), and dynamic stereo-radiography (DSX) data. We implemented and compared two techniques to incorporate in vivo skeletal kinematics as boundary conditions: one used MRI-measured tibiofemoral kinematics in a nonweight-bearing supine position and allowed five degrees of freedom (excluding flexion-extension) at the joint in response to an axially applied force; the other used DSX-measured tibiofemoral kinematics in a weight-bearing standing position and permitted only axial translation in response to the same force. Verification and comparison of the model predictions employed data from a meniscus transplantation study subject with a meniscectomized and an intact knee. The model-predicted cartilage-cartilage contact areas were examined against "benchmarks" from a novel in situ contact area analysis (ISCAA) in which the intersection volume between nondeformed femoral and tibial cartilage was characterized to determine the contact. The results showed that the DSX-based model predicted contact areas in close alignment with the benchmarks, and outperformed the MRI-based model: the contact centroid predicted by the former was on average 85% closer to the benchmark location. The DSX-based FE model predictions also indicated that the (lateral) meniscectomy increased the contact area in the lateral compartment and increased the maximum contact pressure and maximum compressive stress in both compartments. We discuss the importance of accurate, task-specific skeletal kinematics in subject-specific FE modeling, along with the effects of simplifying assumptions and limitations.

Background:Anterior cruciate ligament (ACL) injury increases the risk of meniscus and articular cartilage damage, but the causes are not well understood. Previous in vitro studies were static, required extensive knee dissection, and likely altered meniscal and cartilage contact due to the insertion of pressure sensing devices.Hypothesis:ACL deficiency will lead to increased translation of the lateral meniscus and increased deformation of the medial meniscus as well as alter cartilage contact location, strain, and area.Study Design:Descriptive laboratory study.Methods:With minimally invasive techniques, six 1.0-mm tantalum beads were implanted into the medial and lateral menisci of 6 fresh-frozen cadaveric knees. Dynamic stereo x-rays (DSXs) were obtained during dynamic knee flexion (from 15° to 60°, simulating a standing squat) with a 46-kg load in intact and ACL-deficient states. Knee kinematics, meniscal movement and deformation, and cartilage contact were compared by novel imaging coregistration.Results:During dynamic knee flexion from 15° to 60°, the tibia translated 2.6 mm (P = .05) more anteriorly, with 2.3° more internal rotation (P = .04) with ACL deficiency. The medial and lateral menisci, respectively, translated posteriorly an additional 0.7 mm (P = .05) and 1.0 mm (P = .03). Medial and lateral compartment cartilage contact location moved posteriorly (2.0 mm [P = .05] and 2.0 mm [P = .04], respectively).Conclusion:The lateral meniscus showed greater translation with ACL deficiency compared with the medial meniscus, which may explain the greater incidences of acute lateral meniscus tears and chronic medial meniscus tears. Furthermore, cartilage contact location moved further posteriorly than that of the meniscus in both compartments, possibly imparting more meniscal stresses that may lead to early degeneration. This new, minimally invasive, dynamic in vitro model allows the study of meniscus function and cartilage contact and can be applied to evaluate different pathologies and surgical techniques.Clinical Relevance:This novel model illustrates that ACL injury may lead to significant meniscus and cartilage abnormalities acutely, and these parameters are dynamically measurable while maintaining native anatomy.

Biomechanical analyses of the head and neck system require knowledge of neck muscle forces, which are often estimated from neck muscle volumes. Here we use magnetic resonance images (MRIs) of 17 subjects (6 females, 11 males) to develop a method to predict the volumes of 16 neck muscles by first predicting the total neck muscle volume (TMV) from subject sex and anthropometry, and then predicting individual neck muscle volumes using fixed volume proportions for each neck muscle. We hypothesized that the regression equations for total muscle volume as well as individual muscle volume proportions would be sex specific. We found that females have 59% lower TMV compared to males (females: 510±43 cm3, males: 814±64 cm3; p<0.0001) and that TMV (in cm3) was best predicted by a regression equation that included sex (male=0, female=1) and neck circumference (NC, in cm): TMV=269+13.7NC−233 Sex (adjusted R2=0.868; p<0.01). Individual muscle volume proportions were not sex specific for most neck muscles, although small sex differences existed for three neck muscles (obliqus capitis inferior, longus capitis, and sternocleidomastoid). When predicting individual muscle volumes in subjects not used to develop the model, coefficients of concordance ranged from 0.91 to 0.99. This method of predicting individual neck muscle volumes has the advantage of using only one sex-specific regression equation and one set of sex-specific volume proportions. These data can be used in biomechanical models to estimate muscle forces and tissue loads in the cervical spine.

No abstract

No abstract

Details of small anatomical landmarks and pathologies, such as small changes of the microvasculature and soft exudates, are critical to accurate disease analysis. However, actual medical images always suffer from limited spatial resolution, due to imaging equipment and imaging parameters (e.g. scanning time of CT images). Recently, machine learning, especially deep learning techniques, have brought revolution to image super resolution reconstruction. Motivated by these achievements, in this paper, we propose a novel super resolution method for medical images based on an improved generative adversarial networks. To obtain useful image details as much as possible while avoiding the fake information in high frequency, the original squeeze and excitation block is improved by strengthening important features while weakening non-important ones. Then, by embedding the improved squeeze and excitation block in a simplified EDSR model, we build a new image super resolution network. Finally, a new fusion loss that can further strengthen the constraints on low-level features is designed for training our model. The proposed image super resolution model has been validated on the public medical images, and the results show that visual effects of the reconstructed images by our method, especially in the case of high upscaling factors, outperform state-ofthe-art deep learning-based methods such as SRGAN, EDSR, VDSR and D-DBPN. INDEX TERMSGenerative adversarial network, medical image reconstruction, squeeze and excitation block, super resolution.