RU2528653C2 - Acoustic diagnostic technique for individual's pulmonary focal lesions - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: invention refers to medicine, namely to pulmonology, and can be used for the acoustic diagnosis of individual's pulmonary focal lesions. That is ensured by recording respiratory murmurs of the chest surface in classical auscultation areas. Inhalation respiratory murmurs are recorded in the A mode. Their spectra are calculated. The acoustic parameters are measured. Spectral background values are subtracted from the spectral inhalation murmurs. That is followed by determining lower and upper frequencies for the difference spectrum accompanied by an analysed spectrum increase over the background by 10 dB. An upper cut-off frequency of -3 dB is measured by a lower frequency of a 10dB excess of the analysed spectrum over the background. The upper cut-off frequency of -20 dB from the average noise amplitude between the lower frequency of the 10 dB excess of the analysed spectrum over the background and the spectral cut-off frequency of -3 dB. The local pathological areas are diagnosed by comparing the acoustic parameters of the cut-off frequency of -3 dB and -20 dB to the respective threshold values in each analysis point.

EFFECT: technique provides more effective diagnosis of the pulmonary focal lesions and enables intra-roentgen monitoring of the pulmonary focal mass lesions.

1 ex, 2 tbl, 1 dwg

Description

The invention relates to medicine, in particular to pulmonology, as well as to medical devices, and can be used for acoustic diagnosis of focal lesions in the human lungs.

In clinical practice, acoustic methods for the diagnosis of focal masses in the human lungs are known and widely used, based on subjective listening to sound phenomena occurring in the lungs - auscultation.

Auscultation of the lungs is carried out by putting instruments on the chest surface for listening to a phonendoscope (stethoscope) (R. Laennec, 1817).

The disadvantage of this method of auscultation of the lungs is the subjectivity of the evaluation of the obtained data, the variability of the acoustic parameters of the device, the dependence of the assessment of respiratory noise on the experience of the doctor and the properties of his hearing aid. Currently, the described shortcomings turn auscultation of the lungs into an auxiliary method in relation to modern objective non-acoustic methods for diagnosing focal lesions in the human lungs: X-ray, magnetic resonance, radiographic and others. (“Respiratory diseases: A guide for doctors”: in 4 volumes under the general editorship of N.R. Paleev. Vol. 1 “General pulmonology” / N.I. Aleksandrova, A.G. Bobkov, N.A. Bogdanov and others; under the editorship of N.V. Putov. - M .: Medicine, 1989, p. 253).

Known phonopneumography - a method for an objective study of respiratory noise with the determination of the amplitude and frequency of the spectrum based on the use of a computer through the device. (Murphy RL // Thorax. - 1981 - Vol. 36, p. 99-104. Banaszak, EF, RCKory and GLSnider. Phonopneumpgraphy. Am. Rev. Respir. Dis. 1973 - 107: 449-455. Kraman SS Determination of the site of production of respiratory sounds by subtraction phonopneumography. Am. Rev. Respir. Dis. - 1980 .-- 122: 303-309. Underner M., F. Boita, D. Tete and F. Patte. Auscultation of pulmonary rales, from Laennec to phonopneumography.Rev.Pneumol.Clin. - 1985 - 41: 331-335. Tinkelman DG, C. Lutz, B. Conner. Analysis of breath sounds in normal and asthmatic children and adults using computer digitized airway phonopneumography ( CDAP). Respir. Med. - 1991 - 85: 125-131).

The disadvantage of phonopneumography is that it does not take into account the difference in the speed of the breathing maneuvers being examined by inspiration, which affects the quality of diagnosis and does not allow standardizing the forced air flow.

There is also a method of standardization of respiratory flows during phonopneumography - (N. Gavriely, Palty Y., Alroy G. Spectral characteristics of normal breath sounds. J. Appl. Physiol. 1981, 50, p.307-14), which consists in the flows are standardized for a volume velocity of 1-2 l / s during the inspiratory phase; the frequency of respiratory noises in the frequency band 75-2000 Hz is measured. The maximal inspiratory ODS frequencies determined in this way averaged 434–475 Hz. The disadvantage of this method is the low efficiency of detecting focal formations in the lungs of a person.

There is a method of acoustic diagnosis of focal lesions in the human lung (patent RU No. 2314751, class A61B 5/08, 2008, bull. No. 2), taken as a prototype, the essence of which is to register the main respiratory sounds on the surface of the chest, calculating them spectra, measuring the acoustic parameters characterizing respiratory noise, mapping the acoustic parameters of respiratory noise along the surface of the chest, comparing the acoustic parameters of respiratory noise with a threshold that separates the norm from the pathology, identifying local pa ologicheskih sites. The prototype method is implemented in the multi-channel measuring system VRI-XP (Deep Breeze Ltd, Or Akiva, Israel). To register the main respiratory noise in the prototype, a narrow frequency band of 100-200 Hz is used.

The disadvantage of the prototype is the low efficiency of detecting focal formations in the human lungs.

The objective of the proposed method is to increase the efficiency of detecting focal formations in the human lungs.

The problem is achieved by the method of acoustic diagnosis of focal formations in the lungs of a person, the essence of which is to record and analyze the amplitude-frequency characteristics of the main respiratory sounds at target speeds of the respiratory flow separately in the inspiration phase.

The technical result of the proposed method is to increase the efficiency of diagnosis of focal lung diseases in a clinical setting, X-ray monitoring of focal formations in the lungs on an outpatient basis.

A comparative analysis of the claimed method and the prototype shows that in the inventive method, when registering, the main respiratory noise of the breath is filtered in the “A” mode, for the acoustic characteristic of the respiratory noise, the upper cut-off frequency of the spectrum is determined from the level of -3 dB from the level of the lower frequency of 10 dB above the studied spectrum over the background measured at breath holding and the upper cutoff frequency at a level of -20 dB from the average value of the noise amplitude in the range between the lower frequency of 10 dB above the background of the studied spectrum and often the cut-off of the spectrum at a level of -3 dB, and local pathological areas are diagnosed: RF ₁₈ when the cutoff frequency “-3” dB exceeds the threshold 443.6 and / or the cutoff frequency “-20” dB threshold 631.1, the RF ₁₇ when the frequency is exceeded cut-off “-3” dB threshold 463.1 and / or cut-off frequency “-20” dB threshold 713.1, RF ₁₆ when exceeding cut-off frequency “-3” dB threshold 451.4 and / or cut-off frequency “-20” dB threshold 631.1, RF ₁₅ when exceeding the cutoff frequency “-3” dB threshold 459.2 and / or frequency off the “-20” dB threshold 650.6, RF ₁₄ exceeding the cutoff frequency “-3” dB threshold 428.0 and / or cut-off frequency and "-20" dB threshold 658.4, CP ₁₃ etc. exceeding a cutoff frequency "-3" 392.8 dB threshold and / or cut-off frequency "-20" dB threshold 664.3, CP ₁₂ when exceeding a cutoff frequency "-3" 396.7 dB threshold and / or cut-off frequency "-20 ”DB threshold 619.4, RF ₁₁ when the cutoff frequency is“ -3 ”dB threshold 482.7 and / or the cutoff frequency“ -20 ”dB threshold 662.3 is exceeded.

The method is as follows.

The examined patient is in a sitting position, the nose is closed with a nasal clip, 4-6 breathing cycles of calm breathing are performed, at the same time the doctor installs an acoustic sensor in the form of a microphone with a stethoscopic nozzle at the classical auscultation points, additionally fixing it with a movable rubber band (Martens bandage), covering the chest of the subject (Korenbaum V.I., Tagiltsev A.A., Kostiv A.E., Gorovoi S.V., Pochekutova I.A., Bondar G.N. Acoustic equipment for the study of respiratory sounds oveka // Instruments and Experimental Techniques, 2008. T.51, №2, s.147-154; Geltser BI, seven hundreds EF Propaedeutics Internal Medicine Vladivostok Dal'nauka, 2001)..

The acoustic sensor uses a highly sensitive 1-inch condenser microphone type MK102 (RFT). The stethoscopic nozzle of the acoustic sensor has the shape of a truncated cone with a base diameter of 40 mm and a depth of 6 mm. The signal from the acoustic sensor is fed to the sound level meter 00023 (RFT), on which the frequency response of type “A” is set (Handbook of Technical Acoustics / Edited by M. Hekl, H.A. Mueller. L .: Shipbuilding. 1980. p.64 ), sharply lowering the level of recorded signals in the low-frequency region. This frequency response corresponds to the perception of low-volume sounds by a human hearing, i.e. corresponds to the auditory perception of the doctor during auscultation. Its use for the proposed method is important, because allows you to suppress low-frequency pseudo-sonic interference associated with the movement of the respiratory muscles. The microphone signals passed through the sound level meter were fed to one of the channels of the PowerLab electronic recorder (ADInstruments). A spirometer (ADInstruments) is connected to the other input of the electronic recorder, to the input of which a Lily tube is connected. A spirometer with a Lily tube allows the volumetric velocity of a passing air stream to be recorded in a pneumotachograph mode. The Lily tube gives a signal proportional to the pressure gradient on its lattice, then the spirometer device is calibrated using the 3 liter calibration syringe in units of l / s.

The signals were processed in the Chart program (ADInstruments). When analyzing noise from a general flow of acoustic data using a pneumotachometer, inspiration and background (breath holding) were isolated. To do this, by visual adjustment of the cursor, horizontal sections were selected with a space velocity of about 1 l / s for inspiration and about 0 l / s for the background. As a result, we obtained a set of fragments of recording noise with a constant volumetric flow rate for inspiration and background. The resulting cut out sets of fragments of the acoustic signal (breath, background) are then saved in the wave format. For each signal, the corresponding average value and standard deviation of the volumetric flow rate, which is calculated in the DataPad extension of the Chart package, are determined.

Then, the wave file was processed in the SpectraLab software package (SoundTech). To smooth out gaps formed when cutting fragments with a constant flow rate, the signal is passed through a high-pass filter with a cut-off frequency of 10 Hz. Then, the amplitude spectrum of the signal was calculated (the amplitude scale is logarithmic, the number of samples is 1024, the overlap is 50%, the Haning window). The resulting spectrum (Figure 1) was saved as a text file as text and entered into the MS Excel table.

Then, the spectral values of the background were subtracted from the spectral values of the noise of the inspiration. For the difference spectrum, the lower and upper values of the frequencies were determined at which an excess of the studied spectrum over the background of 10 dB is observed. Next, we analyzed the spectrum of the main respiratory sounds in these frequency ranges. The level of “-3” dB was postponed from the spectrum level at the lower frequency of the excess of the investigated spectrum over the background of 10 dB and the cutoff frequency f _{-3dB was found} . In the range from the lower frequency of the excess of the investigated spectrum over the background of 10 dB to the cutoff frequency f _-3dB , the average value of the noise amplitude was determined. The level of “-20” dB was set aside from the average value of the amplitude and the cutoff frequency f _{-20dB was found} . Data analysis was performed in the laboratory and took 12-16 minutes for one patient.

Recording of respiratory sounds and respiratory air flow rate was carried out synchronously at 8 points corresponding to the topographic points of classical auscultation (ZP ₁₁ - the right suprascapular line, ZP ₁₂ - the left suprascapular region, ZP ₁₃ - the right interscapular region at the level of V-VI of the thoracic vertebrae, paravertebral line , RFP ₁₄ - the left interscapular region at the level of the V-VI thoracic vertebrae, the paravertebral line, RFP ₁₅ - the right interscapular region at the level of the VI thoracic vertebra, the paravertebral line, RFP ₁₆ - the left interscapular region level VI of the thoracic vertebra, paravertebral line, RFP ₁₇ - the right subscapular region, RFP ₁₈ - the left subscapular region), along all topographic lines of the chest.

When recording, the subject breathes through the Lily tube and independently monitors (adjusts) the flow rate in real time on the computer screen, trying to breathe so that the velocity curve does not go beyond the target flow value.

Acoustic values f _-3dB and f _-20dB obtained from the survey points were compared with threshold values (Table 1).

Table 1 Point Thresholds Acoustic parameters f _-3dB f _-20dB PO 11 482.7 662.3 PO 12 396.7 619.4 PO 13 392.8 664.3 PO 14 428.0 658.4 ZP 15 459.2 650.6 PO 16 451.4 631.1 ZP 17 463.1 713.1 PO 18 443.6 631.1

The threshold values of the proposed parameters for various types of focal pulmonary pathology can be determined in various ways, for example, by comparing the presented samples of healthy and sick people, representative by sex and age (O.Yu. Rebrova. Statistical analysis of medical data. M: Publishing House Media Sphere , 2006, p.305).

The obtained threshold values are determined by comparing the indicators for groups of patients with pneumonia and healthy individuals by the method of ROC analysis (optimization of sensitivity and specificity).

In the proposed method, the threshold values were determined by the training sample, which consisted of 36 healthy men aged 18 to 74 years and 36 patients of the same age who were diagnosed with community-acquired pneumonia: in 11 patients, the process was left-sided, 16 - right-sided, and at 9 - bilateral. The diagnosis was established on the basis of clinical and laboratory data and confirmed by the radiological method (radiography, computed tomography). The subjects were determined by the above parameters at each point of the survey. Threshold values were determined by maximizing the sensitivity and specificity indicators for the sample under study (ROC analysis method) (Vlasov V.V. Efficiency of diagnostic studies. M: Medicine, 1988, pp. 104-127). The maximum specificity is achieved in the healthy group of 80.5% and the maximum sensitivity of detection in a patient of at least one focus of pneumonia in the patient group of 83.3% at selected thresholds (Table 1).

For the case of detecting pneumonia experimentally determined that an acoustic indication of disorders associated with the presence of hearth pneumonia is a condition exceeding the parameters f _-3dB and / or f _{20 dB} above threshold.

According to the obtained acoustic parameters in the studied training sample of patients with pneumonia and healthy individuals, the use of the proposed method can significantly increase the efficiency of acoustic detection of focal lesions in the lungs of a person by detecting new, reliable, objectively and automatically evaluated acoustic characteristics of respiratory noise f _-3dB and / or f _-20dB and their threshold values.

The estimated sensitivity of the proposed method 83% is much higher than the sensitivity of subjective auscultation of 45% (Blauert I. Spatial hearing. Translated from German by I.D. Goodwitz. M.: Energy, 1978. - 222 p .; Votchal B.E. et al. Acoustic characteristics of stethophonendoscopes and their measurement // Medical Technology. - 1972 - No. 2. - p.16-20).

The achieved values of sensitivity and specificity of the proposed method are quite high and allow you to use it not only for inter-x-ray monitoring of the pneumonia, which is valuable, but also for their primary dorentgen detection.

The proposed method improves the diagnosis of focal lung diseases in clinical conditions, makes it possible to use remotely in modern digital technologies (telemedicine), to be an additional method in the diagnosis of lobar and focal pulmonary tissue densification, and in some cases to be an alternative diagnostic method for repeated studies due to its safety and the absence of any exposure, sufficiently affordable economically for practical health care.

An example implementation of the method.

Patient S., 32 years old, was admitted to the hospital with complaints of frequent cough with sputum that was difficult to separate, fever to febrile numbers, lethargy, general weakness. From the anamnesis of the disease: sick on the 4th day, the disease began with fever up to 39 °, dry cough. The patient was examined by a therapist and sent to a hospital with a diagnosis of Community-acquired pneumonia, of unspecified etiology, in the lower lobe on the right (segment 9, 8). Respiratory failure of 1 degree. At admission: a state of moderate severity. Respiratory rate 22 per minute. Percussion: shortening of percussion sound in the lower sections along the rear axillary and scapular line. According to computed tomography: In the study of the chest cavity in the lungs, infiltrative changes in 8, 9. Conclusion: right-sided lower lobe pneumonia (S8, S9).

At each point of the examination, the patient in accordance with the claimed method defined the parameters f _-3dB , f _-20dB (Table 1). The parameter values f _-3dB , f _{-20dB are} compared with the corresponding thresholds according to the procedure described above. As a result, in the table of calculated parameters, italicized values exceeding the threshold (Table 2).

table 2 Acoustic parameters at the points of examination of patient S. 32 years old with a diagnosis of right-sided lower lobe pneumonia (S8, S9) PO point f _-3dB f _-20dB eleven 234.4 464.8 12 296.9 539.1 13 359.4 550.8 fourteen 289.1 519.5 fifteen 500,0 707.0 16 277.3 582.0 17 472.7 715.1 eighteen 343.8 589.8 Note: ZP - back surface; f _-3dB - cutoff frequency "-3"dB; f _-20dB - cutoff frequency "-20" dB.

At points 15 and 17 (Table 2), acoustic deviations are detected. Given the projection of these points of examination of the patient, it is obvious that on the right in the projection of the 8th, 9th segments there is a zone of acoustic disturbances. This is consistent with the focus of pneumonia detected in 8, 9 segments of the right lung according to x-ray data.

The inventive method with a fairly high sensitivity (83%) and specificity (80.5%) is completely safe for the subjects, is not associated with harmful radiation and is very simple to implement. Its main purpose is inter-x-ray monitoring of focal formations in the lungs on an outpatient basis. When specifying threshold values, the inventive method can be used at different inspiratory flow rates and using other types of acoustic sensors, including standard electronic stethoscopes.

Claims (1)

A method for the acoustic diagnosis of focal lesions in a human lung, which consists in recording the main respiratory noise on the surface of the chest, calculating their spectra, measuring the acoustic parameters characterizing respiratory noise, mapping the acoustic parameters of respiratory noise on the surface of the chest, comparing the acoustic parameters of respiratory noise with a threshold, separating the norm from pathology, identifying local pathological areas, characterized in that during registration the main respiratory inspiratory noise is filtered in the “A” mode; for the acoustic characteristic of respiratory noise, the upper cut-off frequency of the spectrum is determined from the level of -3 dB from the level of the lower frequency of 10 dB above the background of the spectrum under study, measured by breath holding, and the upper cut-off frequency from -20 dB from the average value of the noise amplitude in the range between the lower frequency of 10 dB above the studied spectrum above the background and the cutoff frequency of the spectrum at the level of -3 dB, and local pathological areas are diagnosed: RF ₁₈ when the cutoff frequency exceeds -3 dB of the threshold 443 , 6 and / or a cutoff frequency of -20 dB threshold 631.1, RFP ₁₇ when the cutoff frequency exceeds -3 dB threshold 463.1 and / or cutoff frequency -20 dB threshold 713.1, RF ₁₆ when the cutoff frequency exceeds -3 dB threshold 451.4 and / or cut-off frequency -20 dB threshold 631.1, RF ₁₅ when exceeding a cutoff frequency of -3 dB threshold 459.2 and / or frequency cut-off -20 dB threshold 650.6, RF ₁₄ when exceeding a cut frequency - 3 dB threshold 428.0 and / or cut-off frequency -20 dB threshold 658.4, RF ₁₃ when the cut-off frequency is -3 dB threshold 392.8 and / or cut-off frequency -20 dB threshold 664.3, RF ₁₂ when the frequency is exceeded a cut-off of -3 dB threshold 396.7 and / or a cut-off frequency of -20 dB p Orog 619.4, ZP ₁₁ when the cutoff frequency exceeds -3 dB threshold 482.7 and / or the cutoff frequency -20 dB threshold 662.3.

Images