A Feasibility Study for Usefulness of Preoperative Simulation of Medial Open-wedge High Tibial Osteotomy for Predicting Postoperative Realignment

Three-dimensional preoperative surgical realignment simulation of medial open-wedge high tibial osteotomy (OWHTO), in which simplified as the rigid rotation around the hinge axis, has been performed to predict the postoperative change and to develop a patient specific instrument for accurate osteotomy. However, the realistic practicality of this extremely simplified simulation method has not been verified. The purpose of this study was to investigate the usefulness of realignment simulation, in which medial OWHTO is simplified as a rotation around a hinge axis, in comparison with a postoperative CT model. Three-dimensional surface model of the tibia and femur was created from preoperative computed-tomography (CT) images (preoperative model) of three patients. Sixty computer simulation models of the medial OWHTO in each patient were created by realignment simulation, in which medial OWHTO is simplified as the rigid rotation of proximal part of tibia relative to the distal part from 1 degree to 20 degrees around three type of hinge axes. The simulation models were compared with the actual postoperative model created from postoperative CT images to assess the reality of the simulation model. After the distal parts of the tibia between each simulation model and postoperative CT model were aligned by a surface registration, average surface distance between two models was calculated as an index representing the similarity of the simulation model to the postoperative model. The minimum average surface distance between the simulation and postoperative CT models were almost 1mm in each patient. The rotation angles at which the minimum average surface distance was represented were almost identical to the actual correction angles. Overlaying the simulation and the postoperative CT models, we found that the posterior tibial tilt and the axial rotation of the proximal tibia of the simulation model well represented that of the postoperative CT model as well as the valgus correction. Therefore, the realignment simulation of medial OWHTO simplified as the rigid rotation around the hinge axis can generate the realistic candidates of postoperative realignment that includes the actual postoperative realignment, suggesting the usefulness for the preoperative simulation method.

Anatomic Study and Assembled Navigation Template-Assisted Implantation for Anterograde Transpubic Screws

Objective: Research the anatomical parameters of the anterograde transpubic screw corridor and evaluate the safety of anterograde transpubic screw placement assisted by the assembled navigation template.Methods: A total of 50 normal subjects, including 25 males and 25 females, underwent pelvic CT scanning in our hospital from January 2020 to September 2020. A 3D model of the ilium was established. The ilium was divided into zone Ⅰ, Ⅱ and Ⅲ according to Nakatani classification. The anterograde transpubic screw channel completely passes through zone Ⅰ and Ⅱ to form corridor A. The anterograde screw channel completely passes through zone Ⅰ, Ⅱ and Ⅲ to form corridor B. The diameter and length of the inner circle, the distance from the center of the inner circle to the posterior superior and to the inferior iliac spine of corridor A and corridor B were measured, respectively. A total of 9 patients underwent anterograde transpubic screw and transverse sacroiliac screw placement assisted by the assembled navigation template in our hospital, including 5 males and 4 females, were retrospectively analyzed. Operative time, blood loss, incision length and fluoroscopy times were recorded. Grading score and Matta score were evaluated after surgery.Results: In 50 normal subjects, the diameter of corridor A was 11.16±2.13 mm, and that of corridor B was 8.54±1.52mm, and the difference between the two corridors was statistically significant (P=0.000). The length of corridor A was 86.39±9.35 mm, and that of corridor B was 117.05±5.91 mm, with significant difference between the two corridors (P=0.000). The surface distance from the screw entry point to the posterior superior iliac spine in corridor A was 109.31± 11.06mm, and that in corridor B was 127.86± 8.23mm, the difference between the two corridors was statistically significant (P=0.000). The surface distance from the screw entry point to the posterior inferior iliac spine in corridor A was 91.16±10.34 mm, and that in corridor B was 106.92±7.91 mm. The difference between the two corridors was statistically significant (P=0.000). Nine patients successfully completed surgery, and a total of 18 sacroiliac transverse screws and 11 anterograde transpubic screws were inserted with the assistance of assembled navigation templates. The mean operation time of the 9 patients was 108.75±25.71 min, the blood loss was 141.11±50.21 ml, the incision length was 14±4.62 cm, and the intraoperative fluoroscopy was 17.89±4.01 times. Matta scores were excellent in 5 patients and good in 4 patients. One of the anterograde transpubic screw was in Grade 1, and 10 were in Grade 0. One S1 screw was in Grade 1, and 8 S1 screws were in Grade 0. Nine S2 screws were in Grade 0.Conclusions: Majority of the patients can accommodate anterograde transpubic screw s with diameter of 6.5 mm. Anterograde transpubic screw placement assisted by an assembled navigation template is clinically feasible, and with low cortical breaches.

Mixup (Sample Pairing) Can Improve the Performance of Deep Segmentation Networks

Researchers address the generalization problem of deep image processing networks mainly through extensive use of data augmentation techniques such as random flips, rotations, and deformations. A data augmentation technique called mixup, which constructs virtual training samples from convex combinations of inputs, was recently proposed for deep classification networks. The algorithm contributed to increased performance on classification in a variety of datasets, but so far has not been evaluated for image segmentation tasks. In this paper, we tested whether the mixup algorithm can improve the generalization performance of deep segmentation networks for medical image data. We trained a standard U-net architecture to segment the prostate in 100 T2-weighted 3D magnetic resonance images from prostate cancer patients, and compared the results with and without mixup in terms of Dice similarity coefficient and mean surface distance from a reference segmentation made by an experienced radiologist. Our results suggest that mixup offers a statistically significant boost in performance compared to non-mixup training, leading to up to 1.9% increase in Dice and a 10.9% decrease in surface distance. The mixup algorithm may thus offer an important aid for medical image segmentation applications, which are typically limited by severe data scarcity.

Validation of Automated Ultrasound-based Registration for Navigated Scaphoid Fixation

Fractures of the scaphoid bone may be treated in a minimally-invasive fashion. Conventionally, fluoroscopy is required to guide the placement of an osteosynthesis screw. In this work, an alternative method based on volumetric ultrasound is validated. Methods: The fully automatic and fast image processing pipeline involves two machine learning architectures for segmentation and registration. A pre-operatively acquired 3D bone model is registered to the 3D bone surface segmented from the intra-operative ultrasound. Screw positioning is planned in an automated fashion and evaluated in an in-vitro setting: Volumetric ultrasound images of a 3D-printed phantom of a human wrist are acquired for 22 different probe poses. For 220 test runs with different initial displacements, the resulting screw placement within a defined safe zone is evaluated. If the screw lies within the safe zone, its placement is assumed to be successful. Results: An isolated analysis of the registration results in a surface distance error of the registered meshes of 0.49 ± 0.01mm, with successful screw placement in all of the evaluated 220 test runs. The full pipeline, combining segmentation and registration, achieves a mean surface distance error of 0.79 ± 0.37mm, leading to successful screw placements for 149 out of 220 test runs. Poses not suited for the registration could be determined. Excluding these from the analysis, 139 out of 160 test runs are successful. Conclusion: The method proves to be promising when evaluating the registration alone, even given the challenging setup of sub-optimal probe positions. The experiments also demonstrate that further improvement regarding the segmentation is necessary.

Learning by Doing: The Use of Distance, Corners and Length in Rewarded Geometric Tasks by Zebrafish (Danio rerio)

Zebrafish spontaneously use distance and directional relationships among three-dimensional extended surfaces to reorient within a rectangular arena. However, they fail to take advantage of either an array of freestanding corners or an array of unequal-length surfaces to search for a no-longer-present goal under a spontaneous cued memory procedure, being unable to use the information supplied by corners and length without some kind of rewarded training. The present study aimed to tease apart the geometric components characterizing a rectangular enclosure under a procedure recruiting the reference memory, thus training zebrafish in fragmented layouts that provided differences in surface distance, corners, and length. Results showed that fish, besides the distance, easily learned to use both corners and length if subjected to a rewarded exit task over time, suggesting that they can represent all the geometrically informative parts of a rectangular arena when consistently exposed to them. Altogether, these findings highlight crucially important issues apropos the employment of different behavioral protocols (spontaneous choice versus training over time) to assess spatial abilities of zebrafish, further paving the way to deepen the role of visual and nonvisual encodings of isolated geometric components in relation to macrostructural boundaries.

Numerical Optimization of Drying Energy Consumption from Multiple Jets Impinging on a Moving Curved Surface

Due to the increasing energy cost, the efficiency of the industrial dryer as the energy-intensive processes should be improved. The designer should optimize the design parameters of industrial drying equipment to achieve the minimum drying energy consumption. SST k-ω turbulence model is used to simulate a real geometry for industrial drying applications. For the optimization of the impinging round jet, the specific drying energy consumption is set as the objective function to be minimized. The jet to surface distance, jet to jet spacing, jet inlet velocity, jet angle, and surface velocity are chosen as the design parameters. The SHERPA search algorithm is used to search for the optimal point from the weighted sum of all objectives method. One correlation is developed and validated for the specific drying energy consumption. It is found that the SST k-ω turbulence model succeeded with reasonable accuracy in reproducing the experimental results. The minimum specific energy consumption correlates with high values of the jet to jet spacing, jet angle, and surface velocity and low values of the nozzle to surface distance and jet inlet velocity. The agreement in the prediction of the specific drying energy consumption between the numerical simulation and correlation is found to be reasonable and all the data points deviate from the correlation by less than 7%.

Large Scale Membrane Movement Induced by a Cation Switch

<p></p><p>A biomembrane sample system where millimolar changes of cations induce reversible large scale (≥ 200 Å) changes in the membrane-to-surface distance is described. The system composes of a free-floating bilayer (FFB), formed adjacent to a self-assembled monolayer (SAM). To examine the membrane movements, differently charged FFBs in the presence and absence of Ca2+ and Na+, respectively, were examined using neutron reflectivity (NR) and quartz crystal microbalance (QCM) measurements, alongside molecular dynamics (MD) simulations. In NR the variation of Ca2+ and Na+ concentration enabled precision manipulation of the FFB-to-surface distance. Simulations suggest that Ca2+ ions bridge between SAM and bilayer whereas the more diffuse binding of Na+, especially to bilayers, is unable to fully overcome the repulsion between anionic FFB and anionic SAM. Reproduced NR results with QCM demonstrate the potential of this easily producible sample system to become a standard analysis tool for e.g. investigating membrane binding effects, endocytosis and cell signalling.<br></p><p></p>