Ultrasound Guidance Does Not Improve the Results of Shock Wave for Plantar Fasciitis or Calcific Achilles Tendinopathy: A Randomized Control Trial.

OBJECTIVE: To establish whether the use of ultrasound to direct shock waves to the area of greater calcification in calcaneal enthesopathies was more effective than the common procedure of directing shock waves to the point where the patient has the most tenderness. DESIGN: Two-armed nonblinded randomized control trial with allocation concealment. SETTING: The Sports Clinic at Sydney University. PATIENTS: Participants 18 years or older with symptomatic plantar fasciitis (PF) (with heel spur) or calcific Achilles tendinopathy (CAT). Seventy-four of 82 cases completed treatment protocol and 6-month follow-up. INTERVENTIONS: Patients were randomized to receive either ultrasound-guided (UG) or patient-guided (PG) shock wave at weekly intervals over 3 to 5 weeks. MAIN OUTCOME MEASURES: Reduced pain on visual analog scale (VAS) and improved functional score on Maryland Foot Score (MFS) (for PF) or Victorian Institute of Sport Assessment-Achilles (VISA-A) (for CAT). Follow-up was at 6 weeks and 3 and 6 months. RESULTS: Comparative 6-month improvements in MFS for the 47 PF cases were PG +20/100 and UG +14/100 (P = 0.20). Comparative 6-month improvement in VISA-A score for the 27 CAT cases were PG +35/100 and UG +27/100 (P = 0.37). Comparative (combined PF and CAT) 6-month improvement in VAS pain scores for all 38 PG cases were +38/100 with +37/100 for all 36 UG shock wave cases. CONCLUSIONS: Although both treatment groups had good clinical outcomes in this study, results for the 2 study groups were almost identical. CLINICAL RELEVANCE: This study shows that there is no major advantage in the addition of ultrasound for guiding shock waves when treating calcaneal enthesopathies (PF and CAT).

PreCoF: counterfactual explanations for fairness.

This paper studies how counterfactual explanations can be used to assess the fairness of a model. Using machine learning for high-stakes decisions is a threat to fairness as these models can amplify bias present in the dataset, and there is no consensus on a universal metric to detect this. The appropriate metric and method to tackle the bias in a dataset will be case-dependent, and it requires insight into the nature of the bias first. We aim to provide this insight by integrating explainable AI (XAI) research with the fairness domain. More specifically, apart from being able to use (Predictive) Counterfactual Explanations to detect explicit bias when the model is directly using the sensitive attribute, we show that it can also be used to detect implicit bias when the model does not use the sensitive attribute directly but does use other correlated attributes leading to a substantial disadvantage for a protected group. We call this metric PreCoF, or Predictive Counterfactual Fairness. Our experimental results show that our metric succeeds in detecting occurrences of implicit bias in the model by assessing which attributes are more present in the explanations of the protected group compared to the unprotected group. These results could help policymakers decide on whether this discrimination is justified or not.

Screening for Anxiety in Patients With Inflammatory Arthritis Using the Multidimensional Health Assessment Questionnaire.

OBJECTIVE: To analyze the Multidimensional Health Assessment Questionnaire (MDHAQ) in screening for anxiety in patients with rheumatoid arthritis (RA) and psoriatic arthritis (PsA), compared to the Hospital Anxiety and Depression Scale (HADS) as the reference standard. METHODS: Patients with a physician diagnosis of RA or PsA were invited to complete the MDHAQ and HADS at their routine rheumatology clinic visit. Sensitivity, specificity, percent agreement, and [Formula: see text] statistics were used to evaluate agreement between 2 MDHAQ items for anxiety and HADS subscale for Anxiety (HADS-A) score of ≥ 8. The first item is a question asked on a 4-point scale (0-3.3), and the second is a yes or no (blank) question asked within a 60-item review of symptoms (ROS) checklist. RESULTS: The study included 183 participants, of whom 126 (68.9%) had RA and 57 (31.1%) had PsA. The mean age was 57.3 years and 66.7% were female. Positive screening for anxiety according to a HADS-A score of ≥ 8 was seen in 39.3% of patients. Compared to those with a HADS-A score of ≥ 8, patients with an MDHAQ score of ≥ 2.2 or a positive on ROS had a sensitivity of 69.9%, specificity of 73.6% and substantial agreement (agreement 80.9%, [Formula: see text] 0.59). CONCLUSION: The MDHAQ provides information similar to the HADS in screening for anxiety in patients with RA and PsA. The use of this single questionnaire, which can also be used to monitor clinical status and to screen for fibromyalgia and depression without requiring multiple questionnaires, may prove a valuable tool in routine clinical practice.

Hemorrhagic Meningitis Due to Varicella-Zoster Virus: A Case Report With Clinical, Cytologic, and Radiologic Correlation.

Although varicella-zoster virus (VZV) is known to affect the central nervous system in a protean manner, hemorrhagic VZV meningitis has not been well documented in the literature. Here, we correlate the clinical, cytologic, and radiologic findings in an immunocompromised patient presenting with subarachnoid hemorrhage associated with VZV meningitis. Clinical findings included multidermatomal zoster, myelitis, and neurapraxia. Magnetic resonance imaging findings included superficial siderosis and diffused linear and nodular leptomeningeal enhancement. Cerebrospinal fluid cytology revealed hemorrhage and lymphocytic pleocytosis. This case report adds hemorrhagic meningitis to the spectrum of complications associated with disseminated VZV infection.

Feasibility of delivering and evaluating stratified care integrated with telehealth ('Rapid Stratified Telehealth') for patients with low back pain: protocol for a feasibility and pilot randomised controlled trial.

INTRODUCTION: Long waiting time is an important barrier to accessing recommended care for low back pain (LBP) in Australia's public health system. This study describes the protocol for a randomised controlled trial (RCT) that aims to establish the feasibility of delivering and evaluating stratified care integrated with telehealth ('Rapid Stratified Telehealth'), which aims to reduce waiting times for LBP. METHODS AND ANALYSIS: We will conduct a single-centre feasibility and pilot RCT with nested qualitative interviews. Sixty participants with LBP newly referred to a hospital outpatient clinic will be randomised to receive Rapid Stratified Telehealth or usual care. Rapid Stratified Telehealth involves matching the mode and type of care to participants' risk of persistent disabling pain (using the Keele STarT MSK Tool) and presence of potential radiculopathy. 'Low risk' patients are matched to one session of advice over the telephone, 'medium risk' to telehealth physiotherapy plus App-based exercises, 'high risk' to telehealth physiotherapy, App-based exercises, and an online pain education programme, and 'potential radiculopathy' fast tracked to usual in-person care. Primary outcomes include the feasibility of delivering Rapid Stratified Telehealth (ie, acceptability assessed through interviews with clinicians and patients, intervention fidelity, appointment duration, App useability and online pain education programme usage) and evaluating Rapid Stratified Telehealth in a future trial (ie, recruitment rates, consent rates, lost to follow-up and missing data). Secondary outcomes include waiting times, number of appointments, intervention and healthcare costs, clinical outcomes (pain, function, quality of life, satisfaction), healthcare use and adverse events (AEs). Quantitative analyses will be descriptive and inform a future adequately-powered RCT. Interview data will be analysed using thematic analysis. ETHICS AND DISSEMINATION: This study has received approval from the Ethics Review Committee (RPAH Zone: X21-0221). Results will be published in peer-reviewed journals and presented at conferences. TRIAL REGISTRATION NUMBER: ACTRN12621001104842.

Deep Learning on Big, Sparse, Behavioral Data.

The outstanding performance of deep learning (DL) for computer vision and natural language processing has fueled increased interest in applying these algorithms more broadly in both research and practice. This study investigates the application of DL techniques to classification of large sparse behavioral data-which has become ubiquitous in the age of big data collection. We report on an extensive search through DL architecture variants and compare the predictive performance of DL with that of carefully regularized logistic regression (LR), which previously (and repeatedly) has been found to be the most accurate machine learning technique generally for sparse behavioral data. At a high level, we demonstrate that by following recommendations from the literature, researchers and practitioners who are not DL experts can achieve world-class performance using DL. More specifically, we report several findings. As a main result, applying DL on 39 big sparse behavioral classification tasks demonstrates a significant performance improvement compared with LR. A follow-up result suggests that if one were to choose the best shallow technique (rather than just LR), there still would often be an improvement from using DL, but that in this case the magnitude of the improvement might not justify the high cost. Investigating when DL performs better, we find that worse performance is obtained for data sets with low signal-from-noise separability-in line with prior results comparing linear and nonlinear classifiers. Exploring why the deep architectures work well, we show that using the first-layer features learned by DL yields better generalization performance for a linear model than do unsupervised feature-reduction methods (e.g., singular-value decomposition). However, to do well enough to beat well-regularized LR with the original sparse representation, more layers from the deep distributed architecture are needed. With respect to interpreting how deep models come to their decisions, we demonstrate how the neurons on the lowest layer of the deep architecture capture nuances from the raw fine-grained features and allow intuitive interpretation. Looking forward, we propose the use of instance-level counterfactual explanations to gain insight into why deep models classify individual data instances the way they do.

Wearable Multimodal Stethoscope Patch for Wireless Biosignal Acquisition and Long-Term Auscultation.

Detecting critical events in postoperative care and improving comfort, costs and availability in sleep assessment are two of many areas in which wearable biosignal acquisition can be a viable tool. Modern sensors as well as patch and textile integration facilitate unobtrusive biosignal acquisition, yet placing sensors at different locations across the body is still prevailing. Actigraphy and the electrocardiogram (ECG) are commonly integrated modalities. The stethoscope however, despite its wide range of applications, has been neglected from these developments. The introduction of digital stethoscopes, recently led to an objectification and increased interest in the field. We present the prototype of a wearable, Bluetooth 5.0 LE enabled multimodal sensor patch combining five modalities: MEMS stethoscope, ambient noise sensing, ECG, impedance pneumography (IP) and 9-axial actigraphy. The system alleviates the need for sensors at different body positions and enables long-term auscultation. Using high sampling rates and online synchronization, multimodal sensor fusion becomes feasible. The patch measures 70 mm x 60 mm and is attached using three 24 mm Ag/AgCl electrodes. High quality cardiac and pulmonary auscultation as well as ECG and IP acquisition are demonstrated. We derived respiration surrogates with linear correlations to a reference exceeding 0.91 and conclude that the system can be utilized in fields requiring unobtrusive yet high quality signal acquisition. Future research will include the integration of additional sensors and further size reduction.

Sex differences in righting from supine to prone in rats (Rattus norvegicus): a masculinized skeletomusculature is not required.

Previous research has shown that sex differences exist in the composition of lateral movements (E. F. Field, I. Q. Whishaw, & S. M. Pellis, 1996, 1997a, 1997b; see also records 1996-06132-009, 1997-05322-015, and 1997-04722-005). An unresolved question is whether sex differences are present in other movements, such as rotation around the longitudinal axis, and whether this difference is dependent on a feminine or masculine skeletomusculature. Female rats (Rattus norvegicus) first rotate their forequarters and then their hindquarters in the same direction. Male rats exhibit rotation of the hindquarters counter to the direction of forequarter rotation. Males with the testicular feminized mutation, who have a feminized skeletomusculature and masculinized central nervous system, are similar to male controls. This study provides evidence that sex differences in movement integration are not restricted to the lateral plane, are not solely due to sex differences in skeletomusculature, and thus are likely mediated by the central nervous system.

Active Learning-Based Pedagogical Rule Extraction.

Many of the state-of-the-art data mining techniques introduce nonlinearities in their models to cope with complex data relationships effectively. Although such techniques are consistently included among the top classification techniques in terms of predictive power, their lack of transparency renders them useless in any domain where comprehensibility is of importance. Rule-extraction algorithms remedy this by distilling comprehensible rule sets from complex models that explain how the classifications are made. This paper considers a new rule extraction technique, based on active learning. The technique generates artificial data points around training data with low confidence in the output score, after which these are labeled by the black-box model. The main novelty of the proposed method is that it uses a pedagogical approach without making any architectural assumptions of the underlying model. It can therefore be applied to any black-box technique. Furthermore, it can generate any rule format, depending on the chosen underlying rule induction technique. In a large-scale empirical study, we demonstrate the validity of our technique to extract trees and rules from artificial neural networks, support vector machines, and random forests, on 25 data sets of varying size and dimensionality. Our results show that not only do the generated rules explain the black-box models well (thereby facilitating the acceptance of such models), the proposed algorithm also performs significantly better than traditional rule induction techniques in terms of accuracy as well as fidelity.

Predictive Modeling With Big Data: Is Bigger Really Better?

With the increasingly widespread collection and processing of "big data," there is natural interest in using these data assets to improve decision making. One of the best understood ways to use data to improve decision making is via predictive analytics. An important, open question is: to what extent do larger data actually lead to better predictive models? In this article we empirically demonstrate that when predictive models are built from sparse, fine-grained data-such as data on low-level human behavior-we continue to see marginal increases in predictive performance even to very large scale. The empirical results are based on data drawn from nine different predictive modeling applications, from book reviews to banking transactions. This study provides a clear illustration that larger data indeed can be more valuable assets for predictive analytics. This implies that institutions with larger data assets-plus the skill to take advantage of them-potentially can obtain substantial competitive advantage over institutions without such access or skill. Moreover, the results suggest that it is worthwhile for companies with access to such fine-grained data, in the context of a key predictive task, to gather both more data instances and more possible data features. As an additional contribution, we introduce an implementation of the multivariate Bernoulli Naïve Bayes algorithm that can scale to massive, sparse data.

p21 loss compromises the relative quiescence of forebrain stem cell proliferation leading to exhaustion of their proliferation capacity.

Adult stem cells in various tissues are relatively quiescent. The cell cycle inhibitor p21cip1/waf1 (p21) has been shown to be important for maintaining hematopoietic stem cell quiescence and self-renewal. We examined the role of p21 in the regulation of adult mammalian forebrain neural stem cells (NSCs). We found that p21-/- mice between post-natal age 60-240 d have more NSCs than wild-type (+/+) controls due to higher proliferation rates of p21-/- NSCs. Thereafter, NSCs in p21-/- mice decline and are reduced in number at 16 mo relative to p21+/+ mice. Similarly, both p21-/- and p21+/+ NSCs display self-renewal in vitro; however, p21-/- NSCs display limited in vitro self-renewal (surviving a few passages, then exhausting). Thus, p21 contributes to adult NSC relative quiescence, which we propose is necessary for the life-long maintenance of NSC self-renewal because NSCs may be limited to a finite number of divisions.

In vivo infusions of exogenous growth factors into the fourth ventricle of the adult mouse brain increase the proliferation of neural progenitors around the fourth ventricle and the central canal of the spinal cord.

Stem cells isolated from the fourth ventricle and spinal cord form neurospheres in vitro in response to basic fibroblast growth factor (FGF2)+heparin (H) or epidermal growth factor (EGF)+FGF2 together. To determine whether these growth factor conditions are sufficient to induce stem cells within the fourth ventricle and spinal cord to proliferate and expand their progeny in vivo, we infused EGF and FGF2, alone or together, with or without H, into the fourth ventricle for 6 days via osmotic minipumps. Animals were injected with bromodeoxyuridine (BrdU) on days 4, 5 and 6 of infusion in order to label cells proliferating in response to the growth factors. Infusions of EGF+FGF2+H into the fourth ventricle resulted in the largest proliferative effect, a 10.8-fold increase in the number of BrdU+ cells around the fourth ventricle, and a 33.5-fold increase in the number of BrdU+ cells around the central canal of the spinal cord, as compared to vehicle infused controls. The majority of the cells were nestin+ after 6 days of infusion. Seven weeks post-infusion, 22 and 30% of the number of BrdU+ cells induced to proliferate after 6 days of EGF+FGF2+H infusions were still detected around the fourth ventricle and central canal of the spinal cord, respectively. Analysis of the fates of the remaining cells showed that a small percentage of BrdU+ cells around the fourth ventricle and in the white matter of the spinal cord differentiated into astrocytes and oligodendrocytes. BrdU+ neurons were not found in the brainstem or in the grey matter of the cervical spinal cord 7 weeks post-infusion. These results show that endogenous stem cells and progenitors around the fourth ventricle and central canal of the spinal cord proliferate in response to exogenously applied growth factors, but unlike in the lateral ventricle where they generate some new neurons, they only produce new astrocytes and oligodendrocytes at 7 weeks post-infusion.