Ill-posed problems can have a distribution of possible solutions rather than a unique one, where each solution incorporates significant features not present in the initial input. We investigate whether cycle-consistent generative neural network models based on generative adversarial networks (GANs) and variational autoencoders (VAEs) can properly sample from this distribution, testing on super-resolution of highly down-sampled images. We are able to produce diverse and plausible predictions, but, looking deeper, we find that the statistics of the generated distributions are substantially wrong. This is a critical flaw in applications that require any kind of uncertainty quantification. We trace this to the fact that these models cannot easily learn a bijective, invertible map between the latent space and the target distribution. Additionally, we describe a simple method for constraining the distribution of a deterministic encoder’s outputs via the Kullback-Leibler divergence without the reparameterization trick used in VAEs.

The VH1 protein encoded by the highly conserved H1 locus of orthopoxviruses is a dual-specificity phosphatase (DUSPs) that hydrolyzes phosphate groups from phosphorylated tyrosine, serine, and threonine residues of viral and host cell proteins. Because the DUSP activities are required for virus replication, VH1 is a prime target for the development of therapeutic inhibitors. However, the presentation of a shallow catalytic site has thwarted all drug development efforts. As an alternative to direct targeting of catalytic pockets, we describe surface contacts between VH1 and substrates that are essential for full activity and provide a new pathway for developing inhibitors of protein-protein interactions. Critical amino acid residues were manipulated by site-directed mutagenesis of VH1, and perturbation of peptide substrate interactions based on these mutations were assessed by high-throughput assays that employed surface plasmon resonance and phosphatase activities. Two positively-charged residues (Lys-20 and Lys-22) and the hydrophobic side chain of Met-60 appear to orient the polarity of the pTyr peptide on the VH1 surface, while additional amino acid residues that flank the catalytic site contribute to substrate recognition and productive dephosphorylation. We propose that the enzyme-substrate contact residues described here may serve as molecular targets for the development of inhibitors that specifically block VH1 catalytic activity and thus poxvirus replication.

Generation of 3-D tissue models from medical imagery is useful for surgical planning and computer simulations, but often requires some amount of manual effort. In this work, we use the clustering algorithm, DBSCAN, in concert with a 3-D buildup procedure to automatically generate 3-D surface models of the brain and lungs from computed tomography head and chest scans, respectively. Extensions to other tissue types such as heart and liver are demonstrated for contrast-enhanced imagery.

Applications such as grid-based real-space density functional theory (DFT) use the Poisson equation to compute electrostatics. However, the expected long tail of the electrostatic potential requires either the use of a large and costly outer domain or Dirichlet boundary conditions estimated via multipole expansion. We find that the oft-used single-center spherical multipole expansion is only appropriate for isotropic mesh domains such as spheres and cubes. In this work, we introduce a method suitable for high aspect ratio meshes whereby the charge density is partitioned into atomic domains and multipoles are computed for each domain. While this approach is moderately more expensive than a single-center expansion, it is numerically stable and still a small fraction of the overall cost of a DFT calculation. The net result is that when high aspect ratio systems are being studied, form-fitted meshes can now be used in lieu of cubic meshes to gain computational speedup.

Arthroscopic ankle arthrodesis provides an alternative to open techniques. Advancements in arthroscopic techniques and instrumentation have made the procedure easier to perform. Arthroscopic ankle arthrodesis has demonstrated faster rates of union, fewer complications, reduced postoperative pain, and shorter hospital stays. Sound surgical techniques, particularly with regard to joint preparation, are critical for success. Comorbidities, such as increased body mass index, history of smoking, malalignment, and posttraumatic arthritis should be considered when contemplating arthroscopic ankle arthrodesis. Although total ankle replacement continues to grow in popularity, arthroscopic ankle arthrodesis remains a viable alternative for management of the end-stage arthritic ankle.

Comprehensive, current and insightful, this well-illustrated text is devoted to the detailed management of common but often challenging complications that all foot and ankle surgeons encounter in thei

In this work, we investigate a block Jacobi-Davidson (J-D) variant suitable for sparse symmetric eigenproblems where a substantial number of extremal eigenvalues are desired (e.g., ground-state real-space quantum chemistry). Most J-D algorithm variations tend to slow down as the number of desired eigenpairs increases due to frequent orthogonalization against a growing list of solved eigenvectors. In our specification of block J-D, all of the steps of the algorithm are performed in clusters, including the linear solves, which allows us to greatly reduce computational effort with blocked matrix-vector multiplies. In addition, we move orthogonalization against locked eigenvectors and working eigenvectors outside of the inner loop but retain the single Ritz vector projection corresponding to the index of the correction vector. Furthermore, we minimize the computational effort by constraining the working subspace to the current vectors being updated and the latest set of corresponding correction vectors. Finally, we incorporate accuracy thresholds based on the precision required by the Fermi-Dirac distribution. The net result is a significant reduction in the computational effort against most previous block J-D implementations, especially as the number of wanted eigenpairs grows. We compare our approach with another robust implementation of block J-D (JDQMR) and the state-of-the-art Chebyshev filter subspace (CheFSI) method for various real-space density functional theory systems. Versus CheFSI, for first-row elements, our method yields competitive timings for valence-only systems and 4-6× speedups for all-electron systems with up to 10× reduced matrix-vector multiplies. For all-electron calculations on larger elements (e.g., gold) where the wanted spectrum is quite narrow compared to the full spectrum, we observe 60× speedup with 200× fewer matrix-vector multiples vs. CheFSI.

The modified Lapidus arthrodesis is a long-established surgical technique for management of hallux valgus that provides reproducible results and quality patient outcomes. The data from 367 consecutive patients undergoing unilateral modified Lapidus arthrodesis from January 1, 2007 to December 31, 2008 at participating centers were retrospectively evaluated. The included patients were categorized into early weightbearing (≤ 21 days) and delayed weightbearing (> 21 days) groups. A total of 24 nonunions (6.5%) were identified, with 13 (7.1%) in the early weightbearing group and 11 (6.0%) in the delayed weightbearing group. To date, the present study is the largest multicenter investigation to evaluate early weightbearing after modified Lapidus arthrodesis and the only large study to directly compare early and delayed weightbearing. The findings of the present study have shown that early weightbearing for modified Lapidus arthrodesis does not increase the risk of nonunion when evaluating various fixation constructs.

This is a retrospective, multicenter study examining the long-term results for the treatment of end-stage hallux rigidus using 3 different surgical procedures. A total of 158 subjects (105 females and 53 males) were included in the present study. They had undergone 1 of the following surgical procedures: arthrodesis, hemi-implant, or resectional arthroplasty. The long-term results for the subjective assessment of pain, function, and alignment, as well as objective radiographic and physical findings, were examined. The median interval to postoperative follow-up for the 3 procedure groups was 159 weeks. No statistically significant difference was found in age or the number of subjects included in the 3 treatment groups (p = .11 and p = .16, respectively). The body mass index was significantly different statistically among the 3 treatment groups, with the hemi-implant group representing a smaller body mass index compared with the other procedures (p = .007). No statistically significant difference was found in the subjective outcomes among the 3 treatment groups using the American College of Foot and Ankle Surgeons' First Metatarsophalangeal Joint and First Ray Scoring Scale (patient questionnaire) or the modified Hallux Metatarsophalangeal-Interphalangeal Scale of the American Orthopedic Foot and Ankle Society (p = .64 and p = .14, respectively). Furthermore, the correlation coefficient between the 2 subjective scoring scales was 0.78, statistically significant and reflecting a moderate to high correlation (p < .001). The results of the radiographic and clinical evaluation revealed that metatarsalgia was the most common finding for the arthrodesis group (9.8%), bony overgrowth into the joint for the hemi-implant group (28.3%), and floating hallux for the resectional arthroplasty group (30.9%). The results of our study suggest that all 3 surgical procedures are viable options for the treatment of end-stage hallux rigidus.

Despite a late start within the realm of arthroscopy, foot and ankle arthroscopy proves to be an important diagnostic and treatment tool for the foot and ankle specialist. As indication for arthroscopy increases, complications associated with foot and ankle arthroscopy must be revisited. We reviewed 405 foot and ankle arthroscopic procedures performed on 390 patients in 4 different facilities over a 3-year period extending from January 2005 to August 2008. Two-hundred-sixty foot and ankle arthroscopic procedures on 251 patients met our inclusion criteria. A total of 246 ankle and 14 posterior subtalar arthroscopic procedures were performed with a mean follow-up of 10.7 ± 3.5 months. Patient demographics, preoperative findings, intraoperative technique, and postoperative course were reviewed. We failed to identify statistically significant predictive factors for complications. Arthroscopy performed in combination with adjunctive procedures showed a trend toward higher complication rate, although statistical significance was not noted. Overall, 20 cases (7.69%) experienced arthroscopy-related complications, and this finding was comparable with previously published results. The most common complication was cutaneous nerve injury, which involved 9 cases (3.46%), and localized superficial infection, which involved 8 cases (3.08%). Injury to the superficial peroneal nerve accounted for 5 of the cutaneous nerve injuries. There were no cases of arthroscopy-related vascular injury. All cases of superficial postoperative infection resolved with antibiotic therapy, and none of the cases required return to the operating room. These results were also similar to published data.

Chronic ankle and subtalar joint (STJ) instability is a common clinical entity and the physician must be able to determine the exact pathologic condition for proper treatment. There are many diagnostic techniques that can be used to evaluate the ankle joint and STJ. These diagnostic techniques do not take the place of a proper patient history taking. Appropriate and aggressive rehabilitation should be attempted in all cases of chronic ankle and subtalar instability before electing surgical reconstruction.

Triple arthrodesis is commonly used to correct complex deformity with hindfoot valgus. The authors use an isolated medial incisional approach for subtalar and talonavicular joint arthrodesis to correct hindfoot deformity, including high degrees of hindfoot valgus. To assess outcomes achieved with this approach, we reviewed the records of 45 patients from the practices of 5 surgeons. Independent variables evaluated included patient age, primary pathology, use of biologic agents, operative time, time to union, and complications. The median patient age was 57 years (range, 14-78 years). Pathology leading to fusion included 27 (60%) posterior tibial tendon dysfunction, 6 (13.3%) tarsal coalition, 7 (5.5%) degenerative joint disease, 2 (4.4%) rheumatoid arthritis, and 1 (2.2%) each, with Charcot neuroarthropathy, multiple sclerosis, and poliomyelitis. Orthobiological materials were used in 27 (60.0%) of the patients. The median duration of surgery was 87 minutes (range, 65-164 minutes), and the median time to successful arthrodesis was 8 weeks (range, 6-20 weeks). A complication was observed in 6 (13.3%) of the patients, including 1 each of the following: painful calcaneal-cuboid joint, talar fracture, incision dehiscence, poor exposure that required abandonment of the procedure, elevated first ray, and painful fixation. None of the patients experienced a nonunion or an adverse event related to the medial neurovascular structures. Based on our experience with the procedure, the single medial-incision subtalar and talonavicular joint arthrodesis is a useful alternative to triple arthrodesis for the correction of hindfoot valgus deformity.

Distraction arthrodesis of the subtalar joint is often used for the correction of neglected calcaneal fractures. Although different techniques have been advocated, there remains some debate as to the optimal type of bone graft for this purpose. This study retrospectively reviewed one surgeon's results for distraction arthrodesis of the talocalcaneal joint for 15 consecutive feet in 15 patients using 12 frozen femoral head and 3 freeze-dried iliac crest allografts. Indications for distraction arthrodesis in this series included neglected calcaneal fracture (n = 10), failed open reduction with internal fixation (n = 3), malunion after ankle fusion (n = 1), and subtalar joint arthritis with deformity (n = 1). The mean patient age was 47.5 (range 29 to 66) years, and the mean duration of follow-up was 20.6 (range 13 to 31) months. Complete union was achieved in 14 (93.33%) feet. Orthobiological agents were used in every case, including 7 (46.67%) platelet-rich plasma, 5 (33.33%) demineralized bone matrix combined with platelet-rich plasma, 2 (13.33%) platelet-rich plasma combined with an implantable electrical bone growth stimulator, and 1 (6.67%) demineralized bone matrix only. One (6.67%) patient developed a nonunion with collapse of the allogeneic graft, requiring revision with autogenous iliac crest bone graft. There were 8 (53.33%) minor complications, including 4 (26.66%) cases with inferior heel irritation, 2 (13.33%) with sural nerve paresthesia, and 2 (13.33%) with wound dehiscence. In conclusion, the use of allograft for subtalar joint distraction arthrodesis results in similar union rates as autogenous iliac crest grafting previously reported in the literature.

Knowledge of the three-dimensional (3D) structures of protein complexes provides a fundamental understanding of biological systems, as well as novel insights for antimicrobial drug and vaccine design. Protein-protein docking is used to predict the 3D structures of protein complexes from their components in silico. In this study, we developed a protein-protein docking pipeline (PPDP) that integrates a variety of state-of-the-art protein docking and structure prediction techniques, providing a systematic platform to predict large-scale protein-protein interactions (PPIs). The PPDP is deployed on high performance computing (HPC) clusters, thus enabling Department of Defense scientists to harness HPC resources to investigate the PPIs of biowarfare agents for the development of countermeasures. We applied the PPDP to investigate the binding interactions of Yersinia effector proteins with their chaperone and the underlying specificity of chaperone/effector interactions.

Compared with traditional open arthrodesis, arthroscopic ankle arthrodesis has been associated with similar rates of fusion, decreased time to union, decreased pain, shorter hospital stay, earlier mobilization, reliable clinical results, and fewer complications. The aim of this case-control study was to analyze cost differences between outpatient arthroscopic and inpatient open ankle arthrodesis. To this end, the authors analyzed 20 ankle arthrodesis procedures: 10 performed by one surgeon on an inpatient basis using an open approach, and 10 performed by another surgeon on an outpatient basis arthroscopically. Patient age, body mass index, tourniquet time, length of stay, complications, days to clinical union, and insurance type, as well as charges and reimbursements for the surgeons and the hospital or surgery center were abstracted from the records. Statistically significant differences were observed between the outpatient arthroscopic and inpatient open arthrodesis groups for total site charges ($3898 +/- 0.00 versus $32,683 +/- $12,762, respectively, P < .0001), reimbursement to the surgeon ($1567 +/- $320 versus $1107 +/- $278, respectively, P = .003), and reimbursement to the hospital or ambulatory surgery center ($1110 +/- $287 versus $8432 +/- $2626, respectively); the ratio of hospital/surgery center charges to hospital/surgery center reimbursements was 28.48% for the inpatient arthroscopic group and 25.80% for the inpatient open arthrodesis group. Outpatient arthroscopic ankle arthrodesis, compared with inpatient open ankle arthrodesis, appears to be less expensive for third party payers, and surgeons are paid more, whereas hospitals and ambulatory surgical centers get paid a greater proportion of the charges that they bill.