Chest
Physical DiagnosisMeasuring Cracklesa
Section snippets
METHODS
Crackles, classified as fine or coarse, were analyzed from a teaching tape by Murphy.7 The signals were digitized using a digital oscilloscope (Nicolet 4094) in the AC mode at a sampling rate of 10,000 points per second; 1.6-s segments were digitized at a time to yield approximately 16,000 values. These digital values were then transferred for further analysis to a Hewlett Packard 9836 desktop computer. Each 1.6-s signal was displayed as a whole (Fig 1), and one observer chose 50-ms segments
RESULTS
Mean values and standard deviations for the IDW, 2CD and LDW1 through LDW4 are shown in Table 1 for each of the two sets of measurements by each of the two observers. They are shown separately for the 64 fine crackles and for the 64 coarse crackles. Mean values for fine and coarse crackles differ significantly with respect to all measures shown.
Intraobserver and interobserver variation between means for groups of crackles can be noted in the data shown in Table 1. Frequency of disagreement
DISCUSSION
The demonstration by Murphy and his colleagues5 that machine analysis of individual crackles can provide useful information raises a number of questions. What are the best analytic approaches? What are the most useful measures? What instruments are available to make these measurements? How do we assess their value?
Desirable characteristics in any test include reproducibility, specificity, sensitivity and simplicity. This study concerns the reproducibility and specificity of measurements that
References (8)
Crackles and wheezes
Lancet
(1967)A treatise on the diseases of the chest, in which they are described according to their anatomical characters, and their diagnosis established on a new principle by means of acoustic instruments. Translated from the French with a preface and notes by John Forbes, M.D. First American edition
(1823)- et al.
Inspiratory crackles—early and late
Thorax
(1974) - et al.
Inspiratory crackles and mechanical events of breathing
Thorax
(1974)
Cited by (65)
Resonance based separation and energy based classification of lung sounds using tunable wavelet transform
2021, Computers in Biology and MedicineCitation Excerpt :However, the ICA based methods have some drawbacks degrading their performance such as the unintended changes in the sign (polarity) and the order of ICs, which result in higher RMS errors and possibly false detection generation. It is obvious that these distortions caused by ICA analysis would negatively affect the features extracted from LS signals and decrease the performance of pulmonary disease diagnosis approaches applied to detected crackles [4,25–27]. The EMD also has some drawbacks that degrade the overall system performance.
Iterative envelope mean fractal dimension filter for the separation of crackles from normal breath sounds
2021, Biomedical Signal Processing and ControlInstantaneous frequency based index to characterize respiratory crackles
2018, Computers in Biology and MedicineCitation Excerpt :They may also be seen as IF estimators based on zero-crossings measurements of different half cycle waves. Thus, their measurements have inter and intra-observer variability [51] since they are also affected by noise and filtering. Therefore, 2CDIF may be more useful than other time domain indices to characterize crackles.
Pulmonary crackles, what does the clinician hear?
2015, Revue des Maladies RespiratoiresIdentification of Velcro rales based on Hilbert-Huang transform
2014, Physica A: Statistical Mechanics and its ApplicationsCharacterization of crackles from patients with fibrosis, heart failure and pneumonia
2013, Medical Engineering and PhysicsCitation Excerpt :Holford [7] suggested the classification of the crackles as fine (high-pitched) or coarse (low-pitched) sounds based on two time intervals: initial deflection width (IDW) and two cycle duration (2CD). Hoevers and Loudon [8] proposed the same classification above using the largest deflection width (LDW). The 2CD has also been investigated to identify the different diseases.