Scan-Free and Fully Automatic Tracking of Native Knee Anatomy from Dynamic Stereo-Radiography with Statistical Shape and Intensity Models.

Kinematic tracking of native anatomy from stereo-radiography provides a quantitative basis for evaluating human movement. Conventional tracking procedures require significant manual effort and call for acquisition and annotation of subject-specific volumetric medical images. The current work introduces a framework for fully automatic tracking of native knee anatomy from dynamic stereo-radiography which forgoes reliance on volumetric scans. The method consists of three computational steps. First, captured radiographs are annotated with segmentation maps and anatomic landmarks using a convolutional neural network. Next, a non-convex polynomial optimization problem formulated from annotated landmarks is solved to acquire preliminary anatomy and pose estimates. Finally, a global optimization routine is performed for concurrent refinement of anatomy and pose. An objective function is maximized which quantifies similarities between masked radiographs and digitally reconstructed radiographs produced from statistical shape and intensity models. The proposed framework was evaluated against manually tracked trials comprising dynamic activities, and additional frames capturing a static knee phantom. Experiments revealed anatomic surface errors routinely below 1.0 mm in both evaluation cohorts. Median absolute errors of individual bone pose estimates were below 1.0 ∘ or mm for 15 out of 18 degrees of freedom in both evaluation cohorts. Results indicate that accurate pose estimation of native anatomy from stereo-radiography may be performed with significantly reduced manual effort, and without reliance on volumetric scans.

Fully automatic tracking of native knee kinematics from stereo-radiography with digitally reconstructed radiographs.

Precise measurement of joint-level motion from stereo-radiography facilitates understanding of human movement. Conventional procedures for kinematic tracking require significant manual effort and are time intensive. The current work introduces a method for fully automatic tracking of native knee kinematics from stereo-radiography sequences. The framework consists of three computational steps. First, biplanar radiograph frames are annotated with segmentation maps and key points using a convolutional neural network. Next, initial bone pose estimates are acquired by solving a polynomial optimization problem constructed from annotated key points and anatomic landmarks from digitized models. A semidefinite relaxation is formulated to realize the global minimum of the non-convex problem. Pose estimates are then refined by registering computed tomography-based digitally reconstructed radiographs to masked radiographs. A novel rendering method is also introduced which enables generating digitally reconstructed radiographs from computed tomography scans with inconsistent slice widths. The automatic tracking framework was evaluated with stereo-radiography trials manually tracked with model-image registration, and with frames which capture a synthetic leg phantom. The tracking method produced pose estimates which were consistently similar to manually tracked values; and demonstrated pose errors below 1.0 degree or millimeter for all femur and tibia degrees of freedom in phantom trials. Results indicate the described framework may benefit orthopaedics and biomechanics applications through acceleration of kinematic tracking.

Specimen-specific finite element representations of implanted hip capsules.

The hip capsule is a ligamentous structure that contributes to hip stability. This article developed specimen-specific finite element models that replicated internal-external (I-E) laxity for ten implanted hip capsules. Capsule properties were calibrated to minimize root mean square error (RMSE) between model and experimental torques. RMSE across specimens was 1.02 ± 0.21 Nm for I-E laxity and 0.78 ± 0.33 Nm and 1.10 ± 0.48 Nm during anterior and posterior dislocation, respectively. RMSE for the same models with average capsule properties was 2.39 ± 0.68 Nm. Specimen-specific models demonstrated the importance of capsule tensioning in hip stability and have relevance for surgical planning and evaluation of implant designs.

Treatment and Response Factors in Muscle Activation during Spinal Manipulation.

The forces applied during a spinal manipulation produce a neuromuscular response in the paraspinal muscles. A systematic evaluation of the factors involved in producing this muscle activity provides a clinical insight. The purpose of this study is to quantify the effect of treatment factors (manipulation sequence and manipulation site) and response factors (muscle layer, muscle location, and muscle side) on the neuromuscular response to spinal manipulation. The surface and indwelling electromyographies of 8 muscle sites were recorded during lumbar side-lying manipulations in 20 asymptomatic participants. The effects of the factors on the number of muscle responses and the muscle activity onset delays were compared using mixed-model linear regressions, effect sizes, and equivalence testing. The treatment factors did not reveal statistical differences between the manipulation sequences (first or second) or manipulation sites (L3 or SI) in the number of muscle responses (p = 0.11, p = 0.28, respectively), or in muscle activity onset delays (p = 0.35 p = 0.35, respectively). There were significantly shorter muscle activity onset delays in the multifidi compared to the superficial muscles (p = 0.02). A small effect size of side (d = 0.44) was observed with significantly greater number of responses (p = 0.02) and shorter muscle activity onset delays (p < 0.001) in the muscles on the left side compared to the right. The location, layer, and side of the neuromuscular responses revealed trends of decreasing muscle response rates and increasing muscle activity onset delays as the distance from the manipulation site increased. These results build on the body of work suggesting that the specificity of manipulation site may not play a role in the neuromuscular response to spinal manipulation-at least within the lumbar spine. In addition, these results demonstrate that multiple manipulations performed in similar areas (L3 and S1) do not change the response significantly, as well as contribute to the clinical understanding that the muscle response rate is higher and with a shorter delay, the closer it is to the manipulation.

Fully automatic tracking of native glenohumeral kinematics from stereo-radiography.

The current work introduces a system for fully automatic tracking of native glenohumeral kinematics in stereo-radiography sequences. The proposed method first applies convolutional neural networks to obtain segmentation and semantic key point predictions in biplanar radiograph frames. Preliminary bone pose estimates are computed by solving a non-convex optimization problem with semidefinite relaxations to register digitized bone landmarks to semantic key points. Initial poses are then refined by registering computed tomography-based digitally reconstructed radiographs to captured scenes, which are masked by segmentation maps to isolate the shoulder joint. A particular neural net architecture which exploits subject-specific geometry is also introduced to improve segmentation predictions and increase robustness of subsequent pose estimates. The method is evaluated by comparing predicted glenohumeral kinematics to manually tracked values from 17 trials capturing 4 dynamic activities. Median orientation differences between predicted and ground truth poses were 1.7∘ and 8.6∘ for the scapula and humerus, respectively. Joint-level kinematics differences were less than 2∘ in 65%, 13%, and 63% of frames for XYZ orientation DoFs based on Euler angle decompositions. Automation of kinematic tracking can increase scalability of tracking workflows in research, clinical, or surgical applications.

Modeling of active skeletal muscles: a 3D continuum approach incorporating multiple muscle interactions.

Skeletal muscles have a highly organized hierarchical structure, whose main function is to generate forces for movement and stability. To understand the complex heterogeneous behaviors of muscles, computational modeling has advanced as a non-invasive approach to evaluate relevant mechanical quantities. Aiming to improve musculoskeletal predictions, this paper presents a framework for modeling 3D deformable muscles that includes continuum constitutive representation, parametric determination, model validation, fiber distribution estimation, and integration of multiple muscles into a system level for joint motion simulation. The passive and active muscle properties were modeled based on the strain energy approach with Hill-type hyperelastic constitutive laws. A parametric study was conducted to validate the model using experimental datasets of passive and active rabbit leg muscles. The active muscle model with calibrated material parameters was then implemented to simulate knee bending during a squat with multiple quadriceps muscles. A computational fluid dynamics (CFD) fiber simulation approach was utilized to estimate the fiber arrangements for each muscle, and a cohesive contact approach was applied to simulate the interactions among muscles. The single muscle simulation results showed that both passive and active muscle elongation responses matched the range of the testing data. The dynamic simulation of knee flexion and extension showed the predictive capability of the model for estimating the active quadriceps responses, which indicates that the presented modeling pipeline is effective and stable for simulating multiple muscle configurations. This work provided an effective framework of a 3D continuum muscle model for complex muscle behavior simulation, which will facilitate additional computational and experimental studies of skeletal muscle mechanics. This study will offer valuable insight into the future development of multiscale neuromuscular models and applications of these models to a wide variety of relevant areas such as biomechanics and clinical research.

Evaluation of single-stage vision models for pose estimation of surgical instruments.

PURPOSE: Multiple applications in open surgical environments may benefit from adoption of markerless computer vision depending on associated speed and accuracy requirements. The current work evaluates vision models for 6-degree of freedom pose estimation of surgical instruments in RGB scenes. Potential use cases are discussed based on observed performance. METHODS: Convolutional neural nets were developed with simulated training data for 6-degree of freedom pose estimation of a representative surgical instrument in RGB scenes. Trained models were evaluated with simulated and real-world scenes. Real-world scenes were produced by using a robotic manipulator to procedurally generate a wide range of object poses. RESULTS: CNNs trained in simulation transferred to real-world evaluation scenes with a mild decrease in pose accuracy. Model performance was sensitive to input image resolution and orientation prediction format. The model with highest accuracy demonstrated mean in-plane translation error of 13 mm and mean long axis orientation error of 5[Formula: see text] in simulated evaluation scenes. Similar errors of 29 mm and 8[Formula: see text] were observed in real-world scenes. CONCLUSION: 6-DoF pose estimators can predict object pose in RGB scenes with real-time inference speed. Observed pose accuracy suggests that applications such as coarse-grained guidance, surgical skill evaluation, or instrument tracking for tray optimization may benefit from markerless pose estimation.

Knee pivot location in asymptomatic older adults.

Representative data of asymptomatic, native-knee kinematics is important when studying changes in knee function across the lifespan. High-speed stereo radiography (HSSR) provides a reliable measure of knee kinematics to <1 mm of translation and 1° of rotation, but studies often have limited statistical power to make comparisons between groups or measure the contribution of individual variability. The purpose of this study is to examine in vivo condylar kinematics to quantify the transverse center-of-rotation, or pivot, location across the flexion range and challenge the medial-pivot paradigm in asymptomatic knee kinematics. We quantified the pivot location during supine leg press, knee extension, standing lunge, and gait for 53 middle-aged and older adults (27 men; 26 women: 50.8 ± 7.0 yrs, 1.75 ± 0.1 m, 79.1 ± 15.4 kg). A central- to medial-pivot location was identified for all activities with increased knee flexion associated with posterior translation of the center-of-rotation. The association between knee angle and anterior-posterior center-of-rotation location was not as strong as the relation between medial-lateral and anterior-posterior location, excluding gait. The Pearson's correlation for gait was stronger between knee angle and anterior-posterior center-of-rotation location (P < 0.001) than medial-lateral and anterior-posterior location (P = 0.0122). Individual variability accounted for a measurable proportion in variance explained of center-of-rotation location. Unique to gait, the lateral translation of center-of-rotation location resulted in the anterior translation of center-of-rotation at <10° knee flexion. Furthermore, no association between vertical ground-reaction force and center-of-rotation was identified.

Instantaneous Generation of Subject-Specific Finite Element Models of the Hip Capsule.

Subject-specific hip capsule models could offer insights into impingement and dislocation risk when coupled with computer-aided surgery, but model calibration is time-consuming using traditional techniques. This study developed a framework for instantaneously generating subject-specific finite element (FE) capsule representations from regression models trained with a probabilistic approach. A validated FE model of the implanted hip capsule was evaluated probabilistically to generate a training dataset relating capsule geometry and material properties to hip laxity. Multivariate regression models were trained using 90% of trials to predict capsule properties based on hip laxity and attachment site information. The regression models were validated using the remaining 10% of the training set by comparing differences in hip laxity between the original trials and the regression-derived capsules. Root mean square errors (RMSEs) in laxity predictions ranged from 1.8° to 2.3°, depending on the type of laxity used in the training set. The RMSE, when predicting the laxity measured from five cadaveric specimens with total hip arthroplasty, was 4.5°. Model generation time was reduced from days to milliseconds. The results demonstrated the potential of regression-based training to instantaneously generate subject-specific FE models and have implications for integrating subject-specific capsule models into surgical planning software.

Impact of bone health on the mechanics of plate fixation for Vancouver B1 periprosthetic femoral fractures.

BACKGROUND: Condyle-spanning plate-screw constructs have shown potential to lower the risks of femoral refractures after the healing of a primary Vancouver type B1 periprosthetic femoral fracture. Limited information exists to show how osteoporosis (a risk factor for periprosthetic femoral fractures) may affect the plate fixation during activities of daily living. METHODS: Using total hip arthroplasty and plate-implanted finite element models of three osteoporotic femurs, this study simulated physiological loads of three activities of daily living, as well as osteoporosis associated muscle weakening, and compared the calculated stress/strain, load transfer and local stiffness with experimentally validated models of three healthy femurs. Two plating systems and two construct lengths (a diaphyseal construct and a condyle-spanning construct) were modeled. FINDINGS: Osteoporotic femurs showed higher bone strain (21.9%) and higher peak plate stress (144.3%) as compared with healthy femurs. Compared with shorter diaphyseal constructs, condyle-spanning constructs of two plating systems reduced bone strains in both healthy and osteoporotic femurs (both applying 'the normal' and 'the weakened muscle forces') around the most distal diaphyseal screw and in the distal metaphysis, both locations where secondary fractures are typically reported. The lowered resultant compressive force and the increased local compressive stiffness in the distal diaphysis and metaphysis may be associated with strain reductions via condyle-spanning constructs. INTERPRETATION: Strain reductions in condyle-spanning constructs agreed with the clinically reported lowered risks of femoral refractures in the distal diaphysis and metaphysis. Multiple condylar screws may mitigate the concentrated strains in the lateral condyle, especially in osteoporotic femurs.

Supine leg press as an alternative to standing lunge in high-speed stereo radiography.

The standing lunge is an activity commonly used to quantify in-vivo knee kinematics with fluoroscopy. The ability to perform the standing lunge varies between subjects and can necessitate movement accommodations to successfully complete the desired range of motion. We proposed a supine leg press as an alternative to the standing lunge that aimed to provide a similar evaluation of knee motion while increasing the measured range of motion. Tibiofemoral kinematics of 53 non-symptomatic adults (27 men, 26 women, 50.8 ± 7.0 yrs.) were calculated from the tracked high-speed stereo radiography (HSSR) images for supine leg press and standing lunge using CT-segmented bony geometries of the right lower limb. The supine leg press proved to be a useful alternative to the standing lunge while providing 46.2° greater range of motion in knee flexion. The difference in angle-matched kinematics across a 100° flexion range between the leg press and lunge was 0.70° in varus-valgus rotation, 1.5° in internal-external rotation, 1.0 mm in medial-lateral translation, 2.3 mm in anterior-posterior translation, and 0.46 mm in superior-inferior translation for men. The angle-matched difference for women across 100° was 0.58° in varus-valgus rotation, 2.4° internal-external rotation, 0.70 mm medial-lateral translation, 2.1 mm anterior-posterior translation, and 0.78 mm superior-inferior translation. The similar kinematics, while having a greater range of motion, and control of the applied load makes the supine leg press an alternative for quantifying in-vivo knee kinematics.

Simplified Mechanical Tests Can Simulate Physiological Mechanics of a Fixation Construct for Periprosthetic Femoral Fractures.

Plate fractures after fixation of a Vancouver Type B1 periprosthetic femoral fracture (PFF) are difficult to treat and could lead to severe disability. However, due to the lack of direct measurement of in vivo performance of the PFF fixation construct, it is unknown whether current standard mechanical tests or previous experimental and computational studies have appropriately reproduced the in vivo mechanics of the plate. To provide a basis for the evaluation and development of appropriate mechanical tests for assessment of plate fracture risk, this study applied loads of common activities of daily living (ADLs) to implanted femur finite element (FE) models with PFF fixation constructs with an existing or a healed PFF. Based on FE simulated plate mechanics, the standard four-point-bend test adequately matched the stress state and the resultant bending moment in the plate as compared with femur models with an existing PFF. In addition, the newly developed constrained three-point-bend tests were able to reproduce plate stresses in models with a healed PFF. Furthermore, a combined bending and compression cadaveric test was appropriate for risk assessment including both plate fracture and screw loosening after the complete healing of PFF. The result of this study provides the means for combined experimental and computational preclinical evaluation of PFF fixation constructs.

Effect of intraoperative treatment options on hip joint stability following total hip arthroplasty.

Dislocation remains the leading indication for revision of total hip arthroplasty (THA). The objective of this study was to use a computational model to compare the overall resistance to both anterior and posterior dislocation for the available THA constructs commonly considered by surgeons attempting to produce a stable joint. Patient-specific musculoskeletal models of THA patients performing activities consistent with anterior and posterior dislocation were developed to calculate joint contact forces and joint positions used for simulations of dislocation in a finite element model of the implanted hip that included an experimentally calibrated hip capsule representation. Dislocations were then performed with consideration of offset using +5 and +9 offset, iteratively with three lipped liner variations in jump distance (10°, 15°, and 20° lips), a size 40 head, and a dual-mobility construct. Dislocation resistance was quantified as the moment required to dislocate the hip and the integral of the moment-flexion angle (dislocation energy). Increasing head diameter increased resistive moment on average for anterior and posterior dislocation by 22% relative to a neutral configuration. A lipped liner resulted in increases in the resistive moment to posterior dislocation of 9%, 19%, and 47% for 10°, 15°, and 20° lips, a sensitivity of approximately 2.8 Nm/mm of additional jump distance. A dual-mobility acetabular design resulted in an average 38% increase in resistive moment and 92% increase in dislocation energy for anterior and posterior dislocation. A quantitative understanding of tradeoffs in the dislocation risk inherent to THA construct options is valuable in supporting surgical decision making.

Impact of periprosthetic femoral fracture fixation plating constructs on local stiffness, load transfer, and bone strains.

Secondary femoral fractures after the successful plate-screw fixation of a primary Vancouver type B1 periprosthetic femoral fracture (PFF) have been associated with the altered state of stress/strain in the femur as the result of plating. The laterally implanted condyle-spanning plate-screw constructs have shown promises clinically in avoiding secondary bone and implant failures as compared with shorter diaphyseal plates. Though the condyle-spanning plating has been hypothesized to avoid stress concentration in the femoral diaphysis through increasing the working length of the plate, biomechanical evidence is lacking on how plate length may impact the stress/strain state of the implanted femur. Through developing and experimentally validating finite element (FE) models of 3 cadaveric femurs, this study investigated the impact of plating on bone strains, load transfer and local stiffness, which were compared between FE models of 2 different plating systems that each had a diaphyseal configuration and a condyle-spanning configuration. Under simulated gait-loading, the condyle-spanning constructs of both plating systems were shown to lower the bone strains around the distal fixation screws (up to 24.8% reduction in maximum principal strain and 26.6% reduction in minimum principal strain) and in the distal metaphyseal shaft of the femur (up to 15.9% and 25.7% reductions in maximum and minimum principal strains, respectively), where secondary bone fractures have been typically reported. In the distal diaphyseal and metaphyseal shaft of femur, FE models of the condyle-spanning constructs were shown to increase the local compressive stiffness (up to 152.9% increases under simulated gait-loading) and decrease the transfer of compressive load (37.1% decreases under simulated gait-loading), which may be indicative of the lowered risks of bone damage.

Automatic tracking of healthy joint kinematics from stereo-radiography sequences.

Kinematic tracking of healthy joints in radiography sequences is frequently performed by maximizing similarities between computed perspective projections of 3D computer models and corresponding objects' appearances in radiographic images. Significant human effort associated with manual tracking presents a major bottleneck in biomechanics research methods and limits the scale of target applications. The current work introduces a method for fully-automatic tracking of tibiofemoral and patellofemoral kinematics in stereo-radiography sequences for subjects performing dynamic activities. The proposed method involves the application of convolutional neural networks for annotating radiographs and a multi-stage optimization pipeline for estimating bone pose based on information provided by neural net predictions. Predicted kinematics are evaluated by comparing against manually-tracked trends across 20 distinct trials. Median absolute differences below 1.5 millimeters or degrees for 6 tibiofemoral and 3 patellofemoral degrees of freedom demonstrate the utility of our approach, which improves upon previous semi-automatic methods by enabling end-to-end automation. Implementation of a fully-automatic pipeline for kinematic tracking will benefit evaluation of human movement by enabling large-scale studies of healthy knee kinematics.

The Use of Synthetic IMU Signals in the Training of Deep Learning Models Significantly Improves the Accuracy of Joint Kinematic Predictions.

Gait analysis based on inertial sensors has become an effective method of quantifying movement mechanics, such as joint kinematics and kinetics. Machine learning techniques are used to reliably predict joint mechanics directly from streams of IMU signals for various activities. These data-driven models require comprehensive and representative training datasets to be generalizable across the movement variability seen in the population at large. Bottlenecks in model development frequently occur due to the lack of sufficient training data and the significant time and resources necessary to acquire these datasets. Reliable methods to generate synthetic biomechanical training data could streamline model development and potentially improve model performance. In this study, we developed a methodology to generate synthetic kinematics and the associated predicted IMU signals using open source musculoskeletal modeling software. These synthetic data were used to train neural networks to predict three degree-of-freedom joint rotations at the hip and knee during gait either in lieu of or along with previously measured experimental gait data. The accuracy of the models' kinematic predictions was assessed using experimentally measured IMU signals and gait kinematics. Models trained using the synthetic data out-performed models using only the experimental data in five of the six rotational degrees of freedom at the hip and knee. On average, root mean square errors in joint angle predictions were improved by 38% at the hip (synthetic data RMSE: 2.3°, measured data RMSE: 4.5°) and 11% at the knee (synthetic data RMSE: 2.9°, measured data RMSE: 3.3°), when models trained solely on synthetic data were compared to measured data. When models were trained on both measured and synthetic data, root mean square errors were reduced by 54% at the hip (measured + synthetic data RMSE: 1.9°) and 45% at the knee (measured + synthetic data RMSE: 1.7°), compared to measured data alone. These findings enable future model development for different activities of clinical significance without the burden of generating large quantities of gait lab data for model training, streamlining model development, and ultimately improving model performance.

Machine learning for rapid estimation of lower extremity muscle and joint loading during activities of daily living.

Joint contact and muscle forces estimated with musculoskeletal modeling techniques offer useful metrics describing movement quality that benefit multiple research and clinical applications. The expensive processing of laboratory data associated with generating these outputs presents challenges to researchers and clinicians, including significant time and expertise requirements that limit the number of subjects typically evaluated. The objective of the current study was to develop and compare machine learning techniques for rapid, data-driven estimation of musculoskeletal metrics from derived gait lab data. OpenSim estimates of patient joint and muscle forces during activities of daily living were simulated using laboratory data from 70 total knee replacement patients and used to develop 4 different machine learning algorithms. Trained machine learning models predicted both trend and magnitude of estimated joint contact (mean correlation coefficients ranging from 0.93 to 0.94 during gait) and muscle forces (mean correlation coefficients ranging from 0.83 to 0.91 during gait) based on anthropometrics, ground reaction forces, and joint angle data. Patient mechanics were accurately predicted by recurrent neural networks, even after removing dependence on key subsets of predictor features. The ability to quickly estimate patient mechanics from derived measurements of movement has the potential to broaden the impact of musculoskeletal modeling by enabling faster assessment in both clinical and research settings.

Development of axial compression and combined axial compression and torque loading configurations to reproduce strain in the implanted femur during activities of daily living.

Femoral strain is indicative of the potential for bone remodeling (strain energy density, SED) and periprosthetic femoral fracture (magnitude of principal strains) after total hip arthroplasty (THA). Previous modeling studies have evaluated femoral strains in THA-implanted femurs under gait loads including both physiological hip contact force and femoral muscle forces. However, experimental replication of the complex muscle forces during activities of daily living (ADLs) is difficult for in vitro assessment of femoral implant or fixation hardware. Alternatively, cadaveric tests using simplified loading configurations have been developed to assess post-THA bone mechanics, although no current studies have demonstrated simplified loading configurations used in mechanical tests may simulate the physiological femoral strains under ADL loads. Using an optimization approach integrated with finite element analysis, this study developed axial compression and combined axial compression and torque testing configurations for three common ADLs (gait, stair-descent and sit-to-stand) via matching the SED profile of the femur in THA-implanted models of three specimens. The optimized simplified-loading models showed good agreement in predicting bone remodeling stimuli (post-THA change in SED per unit mass) and fatigue regions as compared with the ADL-loading models, as well as other modeling and clinical studies. The optimized simplified test configurations can provide a physiological-loading based pre-clinical platform for the evaluation of implant/fixation devices of the femur.

The Relationship Between and Correlates of Problematic Sexual Behavior and Major Mental Illness.

While research has consistently found that general distress and psychopathology are not predictive of sexual recidivism, examination of specific syndromes and their relationship to offending has revealed a potentially more complicated relationship. One proposed mechanism for the mixed findings with respect to major mental illness and sexual offending may be the confound of neurological injury. As identified in Mann et al. (2010) work on psychologically meaningful risk factors, mental illness represents an area in need of more study given the indirect influence it may exert on risk. To this end, the current paper summarizes the study of the relationship between neurological injury, psychosis and problematic sexual behavior among two Canadian samples of forensic and civil psychiatric patients. In the first study we observed higher than expected rates of sexually-themed psychotic symptoms (45%) and problematic sexual behavior (PSB; 40%) among a combined group of forensic and civil psychiatric patients (n = 109). Indeed 70 percent of those individuals who engaged in PSB endorsed sexually-themed psychotic symptoms. While comorbidity is common amongst this group, brain injury appeared to represent a specific liability. Compared to those who did not engage in PSB, those who did were almost 4x (OR = 3.83) more likely to have a documented history of brain injury (e.g., traumatic and acquired brain injury, including fetal alcohol syndrome). In the second study we sought to replicate this finding in a larger forensic sample of 1,240. However, the recorded rates of brain injury were significantly less, such that no relationship to PSB was observed. Based on the mixed findings to date, including our own data, questions remain about the nature of a potential shared vulnerability for psychosis and PSB previously postulated. Among psychiatrically complex individuals who engage in PSB, understanding etiology and links to risk are helpful, but perhaps more importantly is attention to the mechanisms through which symptoms confer risk (e.g., problem solving, sexual disinhibition, social/intimacy deficits) and how best to treat and manage them.

Deep Learning in Gait Parameter Prediction for OA and TKA Patients Wearing IMU Sensors.

Quantitative assessments of patient movement quality in osteoarthritis (OA), specifically spatiotemporal gait parameters (STGPs), can provide in-depth insight into gait patterns, activity types, and changes in mobility after total knee arthroplasty (TKA). A study was conducted to benchmark the ability of multiple deep neural network (DNN) architectures to predict 12 STGPs from inertial measurement unit (IMU) data and to identify an optimal sensor combination, which has yet to be studied for OA and TKA subjects. DNNs were trained using movement data from 29 subjects, walking at slow, normal, and fast paces and evaluated with cross-fold validation over the subjects. Optimal sensor locations were determined by comparing prediction accuracy with 15 IMU configurations (pelvis, thigh, shank, and feet). Percent error across the 12 STGPs ranged from 2.1% (stride time) to 73.7% (toe-out angle) and overall was more accurate in temporal parameters than spatial parameters. The most and least accurate sensor combinations were feet-thighs and singular pelvis, respectively. DNNs showed promising results in predicting STGPs for OA and TKA subjects based on signals from IMU sensors and overcomes the dependency on sensor locations that can hinder the design of patient monitoring systems for clinical application.