Validation of an Automated Cough Detection Algorithm for Tracking Recovery of Pulmonary Tuberculosis Patients

Background

A laboratory-free test for assessing recovery from pulmonary tuberculosis (TB) would be extremely beneficial in regions of the world where laboratory facilities are lacking. Our hypothesis is that analysis of cough sound recordings may provide such a test. In the current paper, we present validation of a cough analysis tool.

Methodology/Principal Findings

Cough data was collected from a cohort of TB patients in Lima, Peru and 25.5 hours of recordings were manually annotated by clinical staff. Analysis software was developed and validated by comparison to manual scoring. Because many patients cough in bursts, coughing was characterized in terms of cough epochs. Our software correctly detects 75.5% of cough episodes with a specificity of 99.6% (comparable to past results using the same definition) and a median false positive rate of 4 false positives/hour, due to the noisy, real-world nature of our dataset. We then manually review detected coughs to eliminate false positives, in effect using the algorithm as a pre-screening tool that reduces reviewing time to roughly 5% of the recording length. This cough analysis approach provides a foundation to support larger-scale studies of coughing rates over time for TB patients undergoing treatment.     

Prediction and course of symptoms and lung function around an exacerbation in chronic obstructive pulmonary disease

Background

Frequent exacerbations induce a high burden to Chronic Obstructive Pulmonary Disease (COPD). We investigated the course of exacerbations in the published COSMIC study that investigated the effects of 1-year withdrawal of fluticasone after a 3-month run-in treatment period with salmeterol/fluticasone in patients with COPD.

Methods

In 373 patients, we evaluated diary cards for symptoms, Peak Expiratory Flow (PEF), and salbutamol use and assessed their course during exacerbations.

Results

There were 492 exacerbations in 224 patients. The level of symptoms of cough, sputum, dyspnea and nocturnal awakening steadily increased from 2 weeks prior to exacerbation, with a sharp rise during the last week. Symptoms of cough, sputum, and dyspnea reverted to baseline values at different rates (after 4, 4, and 7 weeks respectively), whereas symptoms of nocturnal awakening were still increased after eight weeks. The course of symptoms was similar around a first and second exacerbation. Increases in symptoms and salbutamol use and decreases in PEF were associated with a higher risk to develop an exacerbation, but with moderate predictive values, the areas under the receiver operating curves ranging from 0.63 to 0.70.

Conclusions

Exacerbations of COPD are associated with increased symptoms that persist for weeks and the course is very similar between a first and second exacerbation. COPD exacerbations are preceded by increased symptoms and salbutamol use and lower PEF, yet predictive values are too low to warrant daily use in clinical practice.     

Communicative interactions improve visual detection of biological motion

Background

In the context of interacting activities requiring close-body contact such as fighting or dancing, the actions of one agent can be used to predict the actions of the second agent [1]. In the present study, we investigated whether interpersonal predictive coding extends to interactive activities – such as communicative interactions - in which no physical contingency is implied between the movements of the interacting individuals.

Methodology/Principal Findings

Participants observed point-light displays of two agents (A and B) performing separate actions. In the communicative condition, the action performed by agent B responded to a communicative gesture performed by agent A. In the individual condition, agent A''s communicative action was substituted with a non-communicative action. Using a simultaneous masking detection task, we demonstrate that observing the communicative gesture performed by agent A enhanced visual discrimination of agent B.

Conclusions/Significance

Our finding complements and extends previous evidence for interpersonal predictive coding, suggesting that the communicative gestures of one agent can serve as a predictor for the expected actions of the respondent, even if no physical contact between agents is implied.     

Facial and Prosodic Emotion Recognition Deficits Associate with Specific Clusters of Psychotic Symptoms in Schizophrenia

Background

Patients with schizophrenia perform significantly worse on emotion recognition tasks than healthy participants across several sensory modalities. Emotion recognition abilities are correlated with the severity of clinical symptoms, particularly negative symptoms. However, the relationships between specific deficits of emotion recognition across sensory modalities and the presentation of psychotic symptoms remain unclear. The current study aims to explore how emotion recognition ability across modalities and neurocognitive function correlate with clusters of psychotic symptoms in patients with schizophrenia.

Methods

111 participants who met the DSM-IV diagnostic criteria for schizophrenia and 70 healthy participants performed on a dual-modality emotion recognition task, the Diagnostic Analysis of Nonverbal Accuracy 2-Taiwan version (DANVA-2-TW), and selected subscales of WAIS-III. Of all, 92 patients received neurocognitive evaluations, including CPT and WCST. These patients also received the PANSS for clinical evaluation of symptomatology.

Results

The emotion recognition ability of patients with schizophrenia was significantly worse than healthy participants in both facial and vocal modalities, particularly fearful emotion. An inverse correlation was noted between PANSS total score and recognition accuracy for happy emotion. The difficulty of happy emotion recognition and earlier age of onset, together with the perseveration error in WCST predicted total PANSS score. Furthermore, accuracy of happy emotion and the age of onset were the only two significant predictors of delusion/hallucination. All the associations with happy emotion recognition primarily concerned happy prosody.

Discussion

Deficits in emotional processing in specific categories, i.e. in happy emotion, together with deficit in executive function, may reflect dysfunction of brain systems underlying severity of psychotic symptoms, in particular the positive dimension.     

When the Sun Prickles Your Nose: An EEG Study Identifying Neural Bases of Photic Sneezing

Background

Exposure to bright light such as sunlight elicits a sneeze or prickling sensation in about one of every four individuals. This study presents the first scientific examination of this phenomenon, called ‘the photic sneeze reflex’.

Methodology and Principal Findings

In the present experiment, ‘photic sneezers’ and controls were exposed to a standard checkerboard stimulus (block 1) and bright flashing lights (block 2) while their EEG (electro-encephalogram) was recorded. Remarkably, we found a generally enhanced excitability of the visual cortex (mainly in the cuneus) to visual stimuli in ‘photic sneezers’ compared with control subjects. In addition, a stronger prickling sensation in the nose of photic sneezers was found to be associated with activation in the insula and stronger activation in the secondary somatosensory cortex.

Conclusion

We propose that the photic sneeze phenomenon might be the consequence of higher sensitivity to visual stimuli in the visual cortex and of co-activation of somatosensory areas. The ‘photic sneeze reflex’ is therefore not a classical reflex that occurs only at a brainstem or spinal cord level but, in stark contrast to many theories, involves also specific cortical areas.     

Are Lay People Good at Recognising the Symptoms of Schizophrenia?

Aim

The aim of this study was to explore the general public’s perception of schizophrenia symptoms and the need to seek-help for symptoms. The recognition (or ‘labelling’) of schizophrenia symptoms, help-seeking behaviours and public awareness of schizophrenia have been suggested as potentially important factors relating to untreated psychosis.

Method

Participants were asked to rate to what extent they believe vignettes describing classic symptoms (positive and negative) of schizophrenia indicate mental illness. They were also asked if the individuals depicted in the vignettes required help or treatment and asked to suggest what kind of help or treatment.

Results

Only three positive symptoms (i.e., Hallucinatory behaviour, Unusual thought content and Suspiciousness) of schizophrenia were reasonably well perceived (above 70%) as indicating mental illness more than the other positive or negative symptoms. Even when the participants recognised that the symptoms indicated mental illness, not everyone recommended professional help.

Conclusion

There may be a need to improve public awareness of schizophrenia and psychosis symptoms, particularly regarding an awareness of the importance of early intervention for psychosis.     

Exposure to Pornographic Videos and Its Effect on HIV-Related Sexual Risk Behaviours among Male Migrant Workers in Southern India

Objective

Research on pornography and its association with HIV-related sexual behaviours is limited in India. This study aims to examine the prevalence and correlates of viewing pornographic videos and examine its associations with HIV-related sexual risk behaviours among male migrant workers in India.

Methods

Data were drawn from a cross-sectional survey conducted in 2007–08 across 21 districts in four states of India. Respondents included 11,219 male migrants aged 18 years or older, who had migrated to at least two places in the past two years for work. Bivariate and multivariate methods were used to examine the association between viewing pornography and HIV-related sexual risk behaviours.

Results

Two-fifths (40%) of the migrants had viewed pornographic videos in one month prior to the survey. Migrants aged 25–29 years, literate, unmarried and away from native village for more than five years were more likely to view pornography than their counterparts. Migrants who viewed pornographic videos were more likely to engage in paid (Adjusted odds ratio [AOR]: 4.2, 95% confidence interval [CI]: 3.7–4.8) and unpaid sex (AOR: 4.2, 95% CI: 3.7–4.7), report inconsistent condom use in paid sex (AOR: 2.3, 95% CI: 1.7–3.0) and experience STI-like symptoms (AOR: 1.7, 95% CI: 1.5–1.8) than their counterparts.

Conclusions

The findings regarding migrants'' exposure to pornography and its linkage with high HIV risk behaviour suggest that the HIV prevention programmes for migrants need to be more innovative to communicate on the negative-effects of viewing pornography. More importantly, programmes need to find alternative ways to engage migrants in infotainment activities during their leisure time in an effort to reduce their exposure to pornographic videos as well as risky sexual behaviours.     

Clinical and Radiographic Factors Do Not Accurately Diagnose Smear-Negative Tuberculosis in HIV-infected Inpatients in Uganda: A Cross-Sectional Study

Background

Although World Health Organization guidelines recommend clinical judgment and chest radiography for diagnosing tuberculosis in HIV-infected adults with unexplained cough and negative sputum smears for acid-fast bacilli, the diagnostic performance of this approach is unknown. Therefore, we sought to assess the accuracy of symptoms, physical signs, and radiographic findings for diagnosing tuberculosis in this population in a low-income country with a high incidence of tuberculosis.

Methodology

We performed a cross-sectional study enrolling consecutive HIV-infected inpatients with unexplained cough and negative sputum smears for acid-fast bacilli at Mulago Hospital in Kampala, Uganda. Trained medical officers prospectively collected data on standard symptoms and signs of systemic respiratory illness, and two radiologists interpreted chest radiographs in a standardized fashion. We calculated positive- and negative-likelihood ratios of these factors for diagnosing pulmonary tuberculosis (defined when mycobacterial cultures of sputum or bronchoalveolar lavage fluid were positive). We used both conventional and novel regression techniques to develop multivariable prediction models for pulmonary tuberculosis.

Principal Findings

Among 202 enrolled HIV-infected adults with negative sputum smears for acid-fast bacilli, 72 (36%) had culture-positive pulmonary tuberculosis. No single factor, including respiratory symptoms, physical findings, CD4+ T-cell count, or chest radiographic abnormalities, substantially increased or decreased the likelihood of pulmonary tuberculosis. After exhaustive testing, we were also unable to identify any combination of factors which reliably predicted bacteriologically confirmed tuberculosis.

Conclusions and Significance

Clinical and radiographic criteria did not help diagnose smear-negative pulmonary tuberculosis among HIV-infected patients with unexplained cough in a low-income setting. Enhanced diagnostic methods for smear-negative tuberculosis are urgently needed.     

Risk Factors for Acute Respiratory Infection in the Australian Community

Objectives

The objective of this study was to identify the risk factors for ARI in the Australian community.

Methods

We used a national survey of 7578 randomly selected respondents in 2008–2009 to identify the risk factors of ARI. A case was defined as a person experiencing cold or flu with one or more symptoms of: fever, chills, sore throat, runny nose, or cough in the previous four weeks.

Results

There were 19.8% (1505/7578) of respondents who reported ARI in the four weeks prior to the survey. Age was an independent risk factor for ARI, with the risk of acquiring ARI decreasing as age increased. Respondents reporting asthma (OR 1.4, 95%CI: 1.2–1.7) or having someone in their house attending childcare (OR 1.6, 95%CI: 1.2–2.1) were more likely to report ARI.

Conclusions

It is important to identify ways of interrupting transmission of ARI amongst children. Improving identification of risk factors will enable targeted interventions for this exceedingly common syndrome.     

Effect before Cause: Supramodal Recalibration of Sensorimotor Timing

Background

Our motor actions normally generate sensory events, but how do we know which events were self generated and which have external causes? Here we use temporal adaptation to investigate the processing stage and generality of our sensorimotor timing estimates.

Methodology/Principal Findings

Adaptation to artificially-induced delays between action and event can produce a startling percept—upon removal of the delay it feels as if the sensory event precedes its causative action. This temporal recalibration of action and event occurs in a quantitatively similar manner across the sensory modalities. Critically, it is robust to the replacement of one sense during the adaptation phase with another sense during the test judgment.

Conclusions/Significance

Our findings suggest a high-level, supramodal recalibration mechanism. The effects are well described by a simple model which attempts to preserve the expected synchrony between action and event, but only when causality indicates it is reasonable to do so. We further demonstrate that this model successfully characterises related adaptation data from outside the sensorimotor domain.     

Task Complexity Differentially Affects Executed and Imagined Movement Preparation: Evidence from Movement-Related Potentials

Background

The neural simulation theory predicts similarity for the neural mechanisms subserving overt (motor execution) and covert (movement imagination) actions. Here we tested this prediction for movement preparation, a key characteristic of motor cognition.

Methodology/Principal Findings

High-density electroencephalogram (EEG) was recorded during covert and overt actions. Movement preparation was studied with a motor priming paradigm, which varied task complexity and amount of advance information. Participants performed simple or complex sequential finger movements either overtly or covertly. Advance information was either fully predictive or partially predictive. Stimulus-locked event-related potential (ERP) data showed the typical pattern of foreperiod activation for overt and covert movements. The foreperiod contingent negative variation (CNV) differed between simple and complex movements only in the execution task. ERP topographies differed between execution and imagination only when advance information was fully predictive.

Conclusions/Significance

Results suggest a differential contribution of the movement preparation network to action imagination and execution. Overt and covert actions seem to involve similar though not identical mechanisms, where overt actions engage a more fine-grained modulation of covert preparatory states.     

The magic grasp: motor expertise in deception

Background

Most of us are poor at faking actions. Kinematic studies have shown that when pretending to pick up imagined objects (pantomimed actions), we move and shape our hands quite differently from when grasping real ones. These differences between real and pantomimed actions have been linked to separate brain pathways specialized for different kinds of visuomotor guidance. Yet professional magicians regularly use pantomimed actions to deceive audiences.

Methodology and Principal Findings

In this study, we tested whether, despite their skill, magicians might still show kinematic differences between grasping actions made toward real versus imagined objects. We found that their pantomimed actions in fact closely resembled real grasps when the object was visible (but displaced) (Experiment 1), but failed to do so when the object was absent (Experiment 2).

Conclusions and Significance

We suggest that although the occipito-parietal visuomotor system in the dorsal stream is designed to guide goal-directed actions, prolonged practice may enable it to calibrate actions based on visual inputs displaced from the action.     

Extracting diagnoses and investigation results from unstructured text in electronic health records by semi-supervised machine learning

Background

Electronic health records are invaluable for medical research, but much of the information is recorded as unstructured free text which is time-consuming to review manually.

Aim

To develop an algorithm to identify relevant free texts automatically based on labelled examples.

Methods

We developed a novel machine learning algorithm, the ‘Semi-supervised Set Covering Machine’ (S3CM), and tested its ability to detect the presence of coronary angiogram results and ovarian cancer diagnoses in free text in the General Practice Research Database. For training the algorithm, we used texts classified as positive and negative according to their associated Read diagnostic codes, rather than by manual annotation. We evaluated the precision (positive predictive value) and recall (sensitivity) of S3CM in classifying unlabelled texts against the gold standard of manual review. We compared the performance of S3CM with the Transductive Vector Support Machine (TVSM), the original fully-supervised Set Covering Machine (SCM) and our ‘Freetext Matching Algorithm’ natural language processor.

Results

Only 60% of texts with Read codes for angiogram actually contained angiogram results. However, the S3CM algorithm achieved 87% recall with 64% precision on detecting coronary angiogram results, outperforming the fully-supervised SCM (recall 78%, precision 60%) and TSVM (recall 2%, precision 3%). For ovarian cancer diagnoses, S3CM had higher recall than the other algorithms tested (86%). The Freetext Matching Algorithm had better precision than S3CM (85% versus 74%) but lower recall (62%).

Conclusions

Our novel S3CM machine learning algorithm effectively detected free texts in primary care records associated with angiogram results and ovarian cancer diagnoses, after training on pre-classified test sets. It should be easy to adapt to other disease areas as it does not rely on linguistic rules, but needs further testing in other electronic health record datasets.     

CBFS: high performance feature selection algorithm based on feature clearness

Background

The goal of feature selection is to select useful features and simultaneously exclude garbage features from a given dataset for classification purposes. This is expected to bring reduction of processing time and improvement of classification accuracy.

Methodology

In this study, we devised a new feature selection algorithm (CBFS) based on clearness of features. Feature clearness expresses separability among classes in a feature. Highly clear features contribute towards obtaining high classification accuracy. CScore is a measure to score clearness of each feature and is based on clustered samples to centroid of classes in a feature. We also suggest combining CBFS and other algorithms to improve classification accuracy.

Conclusions/Significance

From the experiment we confirm that CBFS is more excellent than up-to-date feature selection algorithms including FeaLect. CBFS can be applied to microarray gene selection, text categorization, and image classification.     

The distribution and mutagenesis of short coding INDELs from 1,128 whole exomes

Background

Identifying insertion/deletion polymorphisms (INDELs) with high confidence has been intrinsically challenging in short-read sequencing data. Here we report our approach for improving INDEL calling accuracy by using a machine learning algorithm to combine call sets generated with three independent methods, and by leveraging the strengths of each individual pipeline. Utilizing this approach, we generated a consensus exome INDEL call set from a large dataset generated by the 1000 Genomes Project (1000G), maximizing both the sensitivity and the specificity of the calls.

Results

This consensus exome INDEL call set features 7,210 INDELs, from 1,128 individuals across 13 populations included in the 1000 Genomes Phase 1 dataset, with a false discovery rate (FDR) of about 7.0%.

Conclusions

In our study we further characterize the patterns and distributions of these exonic INDELs with respect to density, allele length, and site frequency spectrum, as well as the potential mutagenic mechanisms of coding INDELs in humans.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1333-7) contains supplementary material, which is available to authorized users.     

Exploratory analysis of a phase III trial of pirfenidone identifies a subpopulation of patients with idiopathic pulmonary fibrosis as benefiting from treatment

Background

A phase III trial in Japan showed that pirfenidone is effective for idiopathic pulmonary fibrosis (IPF). To find out which patients specifically benefit from pirfenidone, we analyzed in an exploratory manner the data from the phase III trial.

Methods

The patients in the phase III trial were stratified by baseline percentage predicted vital capacity (%VC), arterial oxygen partial pressure (PaO₂), and the lowest oxygen saturation by pulse oximetry (SpO₂) during the 6-minute steady-state exercise test (6MET). In the subpopulations, changes in VC and subjective symptoms (cough and dyspnea on the Fletcher, Hugh-Jones [F, H-J] Classification scale) were evaluated in patients treated with high-dose (1800 mg/day) pirfenidone, low-dose (1200 mg/day) pirfenidone, and placebo at week 52.

Results

Significant efficacy of pirfenidone in reducing the decline in VC could be seen in a subpopulation having %VC ≥ 70% and SpO₂ < 90% at baseline. This favorable effect was accompanied by categorical change in VC and progression-free survival time. In the subpopulation, pirfenidone significantly suppressed cough and dyspnea.

Conclusions

IPF patients having %VC ≥ 70% and SpO₂ < 90% at baseline will most likely benefit from pirfenidone when evaluated using changes in VC (and %VC), and cough and dyspnea symptoms. This subpopulation could expect to benefit most from pirfenidone treatment.

Trial Registration

This clinical trial was registered with the Japan Pharmaceutical Information Center (JAPIC) on September 13th, 2005 (Registration Number: JAPICCTI-050121).     

Untrained Chimpanzees (Pan troglodytes schweinfurthii) Fail to Imitate Novel Actions

Background

Social learning research in apes has focused on social learning in the technical (problem solving) domain - an approach that confounds action and physical information. Successful subjects in such studies may have been able to perform target actions not as a result of imitation learning but because they had learnt some technical aspect, for example, copying the movements of an apparatus (i.e., different forms of emulation learning).

Methods

Here we present data on action copying by non-enculturated and untrained chimpanzees when physical information is removed from demonstrations. To date, only one such study (on gesture copying in a begging context) has been conducted – with negative results. Here we have improved this methodology and have also added non-begging test situations (a possible confound of the earlier study). Both familiar and novel actions were used as targets. Prior to testing, a trained conspecific demonstrator was rewarded for performing target actions in view of observers. All but one of the tested chimpanzees already failed to copy familiar actions. When retested with a novel target action, also the previously successful subject failed to copy – and he did so across several contexts.

Conclusion

Chimpanzees do not seem to copy novel actions, and only some ever copy familiar ones. Due to our having tested only non-enculturated and untrained chimpanzees, the performance of our test subjects speak more than most other studies of the general (dis-)ability of chimpanzees to copy actions, and especially novel actions.