Application of Laser Doppler Vibrometery for human heart auscultation.

In this study the potential of a Laser Doppler Vibrometer (LDV) was tested as a non-contact sensor for the classification of heart sounds. Of the twenty participants recorded using the LDV, five presented with Aortic Stenosis (AS), three were healthy and twelve presented with other pathologies. The recorded heart sounds were denoised and segmented using a combination of the Electrocardiogram (ECG) data and the complexity of the signal. Frequency domain features were extracted from the segmented heart sound cycles and used to train a K-nearest neighbor classifier. Due to the small number of participants, the classifier could not be trained to differentiate between normal and abnormal participants, but could successfully distinguish between participants who presented with AS and those who did not. A sensitivity of 80 % and a specificity of 100 % were achieved a test dataset.

Analysis of adventitious lung sounds originating from pulmonary tuberculosis.

Tuberculosis is a common and potentially deadly infectious disease, usually affecting the respiratory system and causing the sound properties of symptomatic infected lungs to differ from non-infected lungs. Auscultation is often ruled out as a reliable diagnostic technique for TB due to the random distribution of the infection and the varying severity of damage to the lungs. However, advancements in signal processing techniques for respiratory sounds can improve the potential of auscultation far beyond the capabilities of the conventional mechanical stethoscope. Though computer-based signal analysis of respiratory sounds has produced a significant body of research, there have not been any recent investigations into the computer-aided analysis of lung sounds associated with pulmonary Tuberculosis (TB), despite the severity of the disease in many countries. In this paper, respiratory sounds were recorded from 14 locations around the posterior and anterior chest walls of healthy volunteers and patients infected with pulmonary TB. The most significant signal features in both the time and frequency domains associated with the presence of TB, were identified by using the statistical overlap factor (SOF). These features were then employed to train a neural network to automatically classify the auscultation recordings into their respective healthy or TB-origin categories. The neural network yielded a diagnostic accuracy of 73%, but it is believed that automated filtering of the noise in the clinics, more training samples and perhaps other signal processing methods can improve the results of future studies. This work demonstrates the potential of computer-aided auscultation as an aid for the diagnosis and treatment of TB.

Wireless physiological monitoring system for psychiatric patients.

Patients in psychiatric hospitals that are sedated or secluded are at risk of death or injury if they are not continuously monitored. Some psychiatric patients are restless and aggressive, and hence the monitoring device should be robust and must transmit the data wirelessly. Two devices, a glove that measures oxygen saturation and a dorsally-mounted device that measures heart rate, skin temperature and respiratory rate were designed and tested. Both devices connect to one central monitoring station using two separate Bluetooth connections, ensuring a completely wireless setup. A Matlab graphical user interface (GUI) was developed for signal processing and monitoring of the vital signs of the psychiatric patient. Detection algorithms were implemented to detect ECG arrhythmias such as premature ventricular contraction and atrial fibrillation. The prototypes were manufactured and tested in a laboratory setting on healthy volunteers.