Effect of air flow and flow transducer on tracheal breath sounds

Abstract

Tracheal breath sounds (TBSs) were analysed in 12 normal adult subjects at the air flow levels of 1.6, 2.1 and 2.6 litres-1, using a low-resistance flow transducer, and at an estimated flow of 2.1 litres-1 without the flow transducer. The major findings were that the TBSs were significantly distorted by the flow transducer, but were independent of air flow. We therefore conclude that true TBSs can only really be recorded when an oral flow transducer is not used. The within-subject reproducibility of TBS was relatively good, while the cross-subject comparison of TBS showed that each subject produced their own unique spectral pattern, although the TBSs of normal subjects fall in a common frequency range. Also, inspiratory sounds were significantly lower in peak frequency than expiratory TBSs.

Acoustic properties of the normal chest

“…By the mid-1970s, other types of microphone were being adopted; condenser and externally polarized microphones [6,12,[26][27][28][29][30], or electret microphones [23,[31][32][33][34]. A few research teams experimented with dynamic microphones [30], accelerometers, and various piezoelectric contact microphones [35][36][37][38], but the most suitable type was generally considered to be the condenser microphone [23].…”

“…Coupling devices have included: 1) the circular plastic bell-endpiece from a stethoscope [27,41]; 2) a cylindrical probe made of polyvinylchloride (PVC) [29]; 3) a plastic cone [22,23]; 4) a rubber diaphragm fitted over a metal alloy tube [32]; and 5) a small plastic tube [21,34].…”

“…Mention is sometimes made of difficulties arising from the limited storage capacities for digitized data [29,34,39,[49][50][51], but the situation has changed radically over the last 4 yrs due to the development of efficient data handling software equipping widely available computers with clock speeds of over 33 MHz, capable of writing directly to large-capacity hard discs of 100 Mb or more.…”

“…Digitized microphone signals are normally broken up into blocks of data points and spectral components determined by subjecting the block of data to a FFT. Investigators have used various data blocks, whose numbers are powers of 2, such as: 128 points [13]; 512 points [28]; 1,024 points [21,29,34]; 2,048 points [40,41]; and 4,096 points [22]. Each FFT treated block of data is then used to provide a discrete plot of amplitude versus frequency, where the quality of spectral resolution depends directly upon the number of FFT points per block and is inversely proportional to the sampling frequency.…”

Automatic wheeze detection based on auditory modelling

“…Breath sounds have a strong relationship with flow rate [20][21][22][23] . That is, all frequency components of spectral power are changed with flow rate: the higher the flow rate, the higher the spectral power.…”