Many hemodialysis patients undergo plasitc surgery to form the arterio-venous fistula (AVF) in their forearm to improve the vascular access by shunting blood flows. The issue of AVF is the stenosis caused by the disturbance of blood flows; therefore the auscultation system to assist the stenosis diagnosis has been developed. Although the system is intended to be used as a steady monitoring for stenosis assessment, its efficiency was not always high because it cannot estimate where the stenosis locates. In this study, for extracting and estimating the stenosis signal, the shunt murmurs captured by many microphones were decomposed by the principal component analysis (PCA). Furthermore, applying the hierarchical categorization of the recursive subdivision self-organizing map (rs-SOM), the modelling of the stenosis signal was proposed to realise the effective stenosis assessment. The false-positive rate of the stenosis assessment was significantly reduced by using the improved auscultation system.

Blood pressure measurement by auscultatory method is a compulsory skill that is required by all healthcare practitioners. During the measurement, they must concentrate on recognizing the Korotkoff sounds, looking at the sphygmomanometer scale, and constantly deflating the cuff pressure simultaneously. This complex operation is difficult for the new learners and they need a lot of practice with the supervisor in order to guide them on their measurements. However, the supervisor is not always available and consequently, they always face the problem of lack of enough training. In order to help them mastering the skill of measuring blood pressure by auscultatory method more efficiently and effectively, we propose using a sensor device to capture the signals of Korotkoff sounds and cuff pressure during the measurement, and display the signal changes on a visualization tool through wireless connection. At the end of the measurement, the learners can verify their skill on deflation speed and recognition of Korotkoff sounds using the graphical view, and compare their measurements with the machine instantly. By using this device, the new learners do not need to wait for their supervisor for training but can practice with their colleagues more frequently. As a result, they will be able to acquire the skill in a shorter time and be more confident with their measurements.

In this paper, a single-channel adaptive noise canceller (SCANC) is proposed to enhance heart sounds during auscultation. Heart sounds provide important information about the condition of the heart, but other sounds interfere with heart sounds during auscultation. The adaptive noise canceller (ANC) is widely used to reduce noises from biomedical signals, but it is not suitable for enhancing auscultatory sounds acquired by a stethoscope. While the ANC needs two inputs, a stethoscope provides only one input. Other approaches, such as ECG gating and wavelet de-noising, are rather complex and difficult to implement as real-time systems. The proposed SCANC uses a single-channel input based on Heart Sound Inherency Indicator and reference generator. The architecture is simple, so it can be easily implemented in real-time systems. It was experimentally confirmed that the proposed SCANC is efficient for heart sound enhancement and is robust against the heart rate variations.

A remote auscultation support system was developed that compresses and records in real time the patient's breath sound and heart sound, obtained using a stethoscope, and sends this data to an attending doctor at a hospital via network. For real-time recording of the breath sound and heart sound, special-purpose, high-quality sound coding technology was developed and incorporated in the system. This sound coding technology enables the amount of data to be reduced to about 1/18 with virtually no deterioration of the properties of the auscultation sound, high-speed transmission of this data using network, and remote diagnosis of the auscultation sound by a medical specialist. The auscultation locations of each patient, together with the doctor, stethoscoper, and patient database are input into the system in advance at the hospital. At the patient's home or sanatorium, the auscultation sound is recorded according to a human body display that shows auscultation locations, and then sent to the hospital. To ensure patient confidentiality when the auscultation data is transmitted via network, the system scrambles the auscultation data and allows only the attending doctor to play and diagnose the auscultation sound. These features not only support an understanding of the condition of patients being treated at home, but they also enable the construction of an auscultation database for electronic charts that allows auscultation results to be shared within the hospital. When this remote auscultation support system was manufactured and its performance was assessed, virtually the same waveform was obtained for the recorded and played breath sound as for the original breath sound. Results showed that even at a sampling frequency of 11 kHz, remote diagnosis by a medical specialist was in fact possible. Furthermore, if auscultation data of 10 seconds per location for 10 locations is sent, the amount of data sent is only about 120 Kbytes. Since this amount of data converts to only about 25 pages of electronic mail text, even via the existing mobile network the auscultation sounds of many patients can be sent efficiently.