ABSTRACT
OBJECTIVE: Globally, tuberculosis (TB) remains one of the most deadly diseases. Although several effective diagnosis methods exist, in lower income countries clinics may not be in a position to afford expensive equipment and employ the trained experts needed to interpret results. In these situations, symptoms including cough are commonly used to identify patients for testing. However, self-reported cough has suboptimal sensitivity and specificity, which may be improved by digital detection. APPROACH: This study investigates a simple and easily applied method for TB screening based on the automatic analysis of coughing sounds. A database of cough audio recordings was collected and used to develop statistical classifiers. MAIN RESULTS: These classifiers use short-term spectral information to automatically distinguish between the coughs of TB positive patients and healthy controls with an accuracy of 78% and an AUC of 0.95. When a set of five clinical measurements is available in addition to the audio, this accuracy improves to 82%. By choosing an appropriate decision threshold, the system can achieve a sensitivity of 95% at a specificity of approximately 72%. The experiments suggest that the classifiers are using some spectral information that is not perceivable by the human auditory system, and that certain frequencies are more useful for classification than others. SIGNIFICANCE: We conclude that automatic classification of coughing sounds may represent a viable low-cost and low-complexity screening method for TB.
Subject(s)
Cough/complications , Mass Screening/methods , Sound , Tuberculosis/complications , Tuberculosis/diagnosis , Automation , Female , Humans , MaleABSTRACT
In experiments on 11 closed chest dogs the behaviour of pulmonary gas exchange and haemodynamics during isovolaemic haemodilution with 6% dextran was studied. The dogs were ventilated artificially (IPPB, PEEP = 0) with room air. After haemodilution a slight increase of arterial PO2 from 86 to 92 mm Hg was found. In another series of experiments an inspiratory O2-concentration of 25% was applied resulting in an increase of arterial PO2 from 106 to 113 mm Hg. In both series a decrease of alveolararterial PO2 gradients was observed. Effective pulmonary capillary blood flow varied in accordance with changes of cardiac output. Thus intrapulmonary shunt is supposed to have remained constant. The changes of pulmonary O2 diffusing capacity could be explained by the effect of haemodilution per se. At the end of the experiments ventilation was changed by adding a positive endexpiratory pressure of 8 cm H2O resulting in a decrease of arterial PO2 and a steep fall of cardiac output. In conclusion, isovolaemic haemodilution leads to only negligible variations of pulmonary gas exchange which should not be of any clinical importance.
Subject(s)
Hemodilution , Hemodynamics , Respiration , Animals , Dextrans , Dogs , Oxygen Consumption , Pulmonary Diffusing CapacityABSTRACT
In experiments on 6 closed chest dogs the behaviour of pulmonary gas exchange during isovolaemic haemodilution with 6% dextran was studied. The dogs were ventilated artificially using an inspiratory gas mixture containing 25% O2. A slight increase of arterial Po2 values was found. This change was accompanied by a decrease in alveolar-arterial Po2 and Pco2 gradients. The size of effective pulmonary capillary blood flow varied in accordance with the size of cardiac output. The changes of pulmonary O2 diffusing capacity could be explained by the effect of haemodilution per se.