WO2011157873A1

WO2011157873A1 - System and method for obtaining data on the breathing cycle of a patient

Info

Publication number: WO2011157873A1
Application number: PCT/ES2011/070242
Authority: WO
Inventors: Luís GOLDMAN TARLOVKY
Original assignee: Fundación Para La Investigación Biomédica Del Hospital Universitario De La Paz
Priority date: 2010-06-16
Filing date: 2011-04-11
Publication date: 2011-12-22
Also published as: ES2387153A1; ES2387153B1

Abstract

The invention relates to a system and non-invasive method for obtaining data on the breathing cycle of a patient from a video sequence of the patient recorded by a thermographic camera, that includes the following operations: identifying the thoracic region (TX), abdominal region (AB) and at least one nasal vestibule (NS) of the patient in the thermographic video sequence; and processing the thermographic video sequence in order to infer, as a function of periodic changes in said regions, data on the breathing cycle of the patient.

Description

SYSTEM AND PROCEDURE FOR OBTAINING DATA ABOUT A PATIENT'S RESPIRATORY CYCLE

OBJECTIVE OF THE INVENTION

The object of the present invention is a non-invasive system and method for obtaining data about the respiratory cycle of a patient. Specifically, the procedure evaluates: 1) periodic changes related to breathing that occur in certain areas of a patient's body by processing a video sequence obtained by means of a thermal imager and 2) from the curves of the chest, abdomen and Nasal airflow resulting from the previous procedure, accurately measures respiratory synchrony without virtually no instrumental lag. BACKGROUND OF THE INVENTION

It has been shown that a multitude of respiratory pathologies produce alterations in the inspired and exhaled volumes of air and in the respiratory rate. Thus, for example, the Obstructive Sleep Apnea Syndrome is characterized by more or less prolonged respiratory pauses and an obstruction in the upper airways that decreases the air flow. In children, respiratory pauses can also occur due to neurological and muscular conditions or prolonged apnea of etiology not yet fully elucidated as Sudden Infant Death Syndrome. Therefore, it is necessary to know data about the patient's respiratory cycle to be able to make an accurate differential diagnosis and take the necessary life support measures if the event is prolonged putting the patient's life at risk.

The respiratory cycle of a patient originates primarily in the cyclic contraction of the diaphragm muscle and results in displacements of the chest and abdomen causing a bidirectional air flow detectable at the level of the nostrils and / or in the mouth. The lack of relative synchrony between Displacements of the thorax and abdomen and air flow may be indicative of different respiratory pathologies. Nasal air flow refers to the air entering and exiting through the nose during inspiration and expiration, while thoracic and abdominal displacements refer respectively to the way in which the chest and abdomen swell and deflate during the breathing. Under physiological conditions these processes are synchronous, accepted as a normal angle thoraco abdominal desfazamiento between 0 and 54 ° ^or. There are currently some methods that allow obtaining data about the respiratory cycle of a patient in order to diagnose respiratory diseases. For example, some very skilled doctors can identify small respiratory pauses and apnea episodes through the simple observation of thoracic and abdominal displacements. They can even detect episodes of airway obstruction through the recognition of inspiratory or forced expiratory maneuvers. However, these methods are not useful for the objective and long-term monitoring of a patient's respiratory cycle. In patients who require monitoring of the respiratory cycle, it is common to use elastic bands provided with transducers around the thorax and abdomen of the patient that measure thoracic and abdominal displacements, as well as the use of a mask equipped with a pneumotachograph to detect air flow nasal. However, this system is very invasive and can influence the patient's respiratory pattern. In addition, these systems are especially inadequate for taking measurements in young children, who may feel uncomfortable and accidentally get hooked on the cables and sensors used. Consequently, there is still a need for a system that allows a patient's respiratory cycle to be obtained reliably and repetitively. DESCRIPTION OF THE INVENTION

The inventors of the present application have developed a procedure that allows obtaining data about the respiratory cycle of a patient from a thermographic video sequence of the patient, thus avoiding the need to connect measuring devices such as elastic bands provided with transducers or the like. and air flow sensors. To do this, a thermographic video sequence of a region of interest of a patient and of a duration that can range from a few seconds to a few minutes is taken, and then the thermal map variations (of size, of shape, position and / or magnitude) in said region of interest. That is, considering that the map of surface temperatures of, for example, a patient's chest is relatively constant, the periodic changes that occur in a thermographic image of the region of interest should be the consequence of movements related to breathing.

One aspect of the invention, therefore, describes a method for obtaining data about a patient's respiratory cycle from a video sequence of the patient recorded with a thermal imager. In this document, it is understood that the thermal imager is capable of detecting thermal emissions whose wavelength corresponds to the range of surface temperatures of a person and their environment in a hospital ward, for example between 25 ° C and 40 ° C. On the other hand, the data obtained about the respiratory cycle are represented as graphs that allow a user to identify the moments of beginning and end of the inspiration and expiration phases, as well as possible respiratory pauses, inspiration or forced expiration, or other respiratory abnormalities and provide the angle of lag between the respiratory movements of the thorax and abdomen. The process of the invention comprises the following steps:

1) Identify in the thermographic video sequence the regions of the thoracic, abdominal patient and at least one nasal vestibule.

2) Deducing from the periodic changes in the thermographic video sequence in said regions data about the patient's respiratory cycle, more preferably the degree of synchronization between the movements of the chest and abdomen and the air flow through the patient's airways .

According to a preferred embodiment of the invention, this step in turn comprises the steps of:

- Subtract pixel by pixel consecutive frames of the video sequence in the region of interest. In each pixel, this operation can result in zero or positive or negative values depending on the thermal variation between the corresponding pixels.

- Eliminate pixels whose values fluctuate around 0 and that represent insignificant thermal variations, probably undetermined noise.

- Binarize the resulting differential image obtaining a series of matrices whose elements take the value of 1 or 0 depending on whether the pixel-to-pixel differences between consecutive frames have respectively increased or decreased.

- Estimate the area of pixels 1 and 0 in the binary matrices obtained taking into account the relative weight of the 2x2 neighboring pixels. For example, if none of the 4 neighboring pixels is significant the area is 0, if two diagonal pixels are positive the area is worth ¾ and if all pixels are positive the area is worth the unit.

- The difference between areas 1 and 0 represents the value of the thermal variation in each region of interest. This value is updated 50 times per second (which is the number of frames recorded per second) giving rise to a series of time whose magnitude varies with the respiratory displacements of the thorax and abdomen and the change in temperature in the nostrils. Finally, a cross-correlation analysis allows to obtain the offset angle or delay time between the signals.

In normal individuals, the respiratory rate may range between approximately 12 and 40 breaths per minute, while the heart rate may be between approximately 60 and 100 pulses per minute, both values being dependent on the subject's age. Accordingly, it is possible to filter frequencies greater than 0.8 Hz to prevent changes due to heart rate interfering with the identification of data about the respiratory cycle. That is, the filter will eliminate artifacts at more than 48 cycles per second. More preferably, a low-pass filter of Chebyshev type II and order 6 is used.

Furthermore, although the described invention is directed to a procedure, it is understood that said procedure is capable of being encoded as a computer program. The program may have the form of source code, object code, an intermediate source of code and object code (for example, in partially compiled form), or in any other form suitable for the implementation of the method according to the invention.

Said computer program may be located on or within a carrier, any entity or device capable of supporting the program being a carrier. The carrier can be a storage medium such as a ROM, a CD ROM or a semiconductor ROM, or a magnetic recording medium, for example, a flexible disk or a hard disk. The carrier can also be a transmissible carrier, for example, an electrical or optical signal that could be transported through electrical or optical cable, by radio or by any other means. When the program is incorporated into a signal that can be directly transported by a cable or other device or medium, the carrier may be constituted by said cable or other device or means. The carrier could also be an integrated circuit in which the program is included, the integrated circuit adapted to execute, or to be used in the execution of, the corresponding processes.

A second aspect of the present invention describes a system capable of carrying out the process described above comprising

- a thermographic video camera capable of obtaining a video sequence of at least one region of interest of a patient; and - a processing medium, preferably a PC, connected to the thermographic video camera and configured to carry out the procedure described above.

BRIEF DESCRIPTION OF THE FIGURES

Fig. 1 shows an example in grayscale of a frame corresponding to a thermographic video sequence of the torso and face of a patient where thermal zones are displayed that move with the respiratory movements of the thorax and abdomen.

Fig. 2 shows a pair of thermographic frames corresponding to the nasal region highlighting the predominance of the gray areas (lower temperature) at the end of inspiration (Fig. 2A) and a decrease of these areas (higher temperature) at the end of the expiration (Fig. 2B). The inhalation of air at room temperature is what causes the thermal decrease in the nostrils.

Fig. 3 shows the result of subtracting pixel by pixel consecutive frames. The non-significant results (around 0) were eliminated since they represent undetermined noise while the positive and negative results were respectively assigned to 1 and 0. The complementary pattern of areas 1 and 0 is observed during inspiration and expiration. Figs. 4A-C show respectively the fluctuations corresponding to the respiratory movements of the regions of interest TX, AB and NS, the nasal pressures obtained by means of a nasal probe used to check results and the correlation between thoracic and abdominal movements.

Figs. 4A-C show respectively the fluctuations corresponding to the respiratory movements of the regions of interest TX, AB and NS, the nasal pressures obtained by means of a nasal probe used to check results and the correlation between thoracic and abdominal movements.

PREFERRED EMBODIMENT OF THE INVENTION An exemplary embodiment of the present invention is described below with reference to the attached figures.

The thermal imager used in this example is a Flir Thermovision A40 M (Flir Systems, AB, Sweden) controlled by software from the same company (Therma Cam Researcher software, Flir Systems, AB, Sweden) that measures thermal emissivity in a range spectral between 7.5-13 μηπ. The camera was connected to a PC through a firewire connection. The camera has a standard lens of 24 ° and F = 1 directed towards the face, thorax and abdomen of patients located between 70-120 cm. The thermographic video sequences acquired in AVI format (50 frames per second, 76000 pixels per frame), lasting at least 10 minutes, were transformed to Matlab format (Matlab 7.0, Mathworks, Inc., Natick, MA) for post processing The room temperature was maintained between 22 ⁰ and 24 °. Fig. 1 shows a frame corresponding to a patient during the inspiration process. The lighter areas correspond to warmer temperatures, while the dark ones correspond to colder temperatures, such as It is seen on the scale on the right. The three regions of interest usually used for the diagnosis of respiratory problems are also represented: abdomen (AB), thorax (TX) and nose (NS). Specifically, Fig. 2 shows the frames of a nostril corresponding to the end of inspiration (Fig. 2A) and expiration (Fig. 2B) in a normal patient. It can be seen how, during inspiration (Fig. 2A), the distribution of temperatures in the central zone has cold areas that do not exist during expiration (Fig. 2B). This is because the air inhaled during inspiration is at a lower temperature compared to the temperature of the exhaled air. This change in the distribution of temperatures in the nasal area will therefore occur periodically with breathing. Similarly, variations in abdominal and thoracic volume that occur during breathing cause periodic changes in the distribution of temperatures in these areas. As previously mentioned herein, the process of the invention takes advantage of these periodic changes in temperature distributions to obtain data about the respiratory cycle. First, consecutive pixel-to-pixel frames are subtracted: the resulting pixels whose value has practically not changed are eliminated as they represent noise, while the pixels whose value has increased or decreased are assigned respectively the values 1 and 0. Fig. 3 shows the result of this operation in the abdominal and nasal regions.

The areas formed by pixels 1 and 0 were calculated taking into account the value of each pixel relative to the 2x2 of its environment. The difference between these areas calculated 50 times per second generated the time series of Figs. 4A-C and 5A-C containing respiratory fluctuations in the chest, abdomen and nostrils corresponding to two patients. Specifically, Figs. 4A-C show the signals obtained non-invasively and without any instrumental delay corresponding to the respiratory pattern in the abdominal, thoracic and nasal regions of a healthy patient. In Fig. 4A, the thermographic curves corresponding to the regions mentioned and represented in arbitrary units have a totally physiological synchrony. It is predictable in a normal subject from the respiratory point of view where the expansion of the chest and abdomen and the consequent inspiration of air through the nostrils is practically simultaneous. In Fig. 4B the control pressure is shown in the nostrils where a decrease in amplitude is observed that corresponds totally with the thermographic curves (compare with Fig. 4A) Finally in Fig. 4C, the cross correlation between the movements of the thorax and abdomen depending on the number of samples obtained by mutually displacing both curves for a certain time or "lag" and calculating the correlation again. If the peak of the cross-correlation function corresponds to a zero displacement of the curves analyzed, these are absolutely synchronous. If the peak occurs with a certain delay, this is the measure of the lack of synchrony between the thorax and abdomen. Fig. 5, on the other hand, shows the same curves but obtained in a patient with respiratory involvement, asynchrony is evident between the thermographic signals obtained (Fig. 5A). In turn, the curves of the thorax and abdomen do not coincide with the nasal pressure (Fig. 5B) considered as a reference and confirming its lack of synchrony. Cross correlation (Fig. 5C) allows quantification of thoraco-abdominal displacement as the difference between the times corresponding to the maximum correlation function and the "lag" = 0. Its value in this case is clearly abnormal.

Claims

1. Procedure to obtain data about the respiratory cycle of a patient from a video sequence of the patient recorded by means of a thermal imager, characterized in that it comprises the following operations:

- identify in the thermographic video sequence the regions of the thoracic (TX), abdominal (AB) patient and at least one nasal vestibule (NS); Y

- process the thermographic video sequence to deduce, based on periodic changes in said regions, data about the patient's respiratory cycle.

2. Method according to claim 1, wherein the processing operation comprises:

- Subtract pixel by pixel consecutive frames from video images of the area of interest, filtering those virtually unchanged pixels and assigning 1 or 0 to pixels whose value has respectively increased or decreased

- apply a morphological transformation to the obtained frames that allow calculating the differences between positive and negative areas by relativizing the weight of each pixel according to the adjacent 2x2, obtaining time series representing respiratory fluctuations in the chest, abdomen and nose regions .

3. Method according to claim 2, further comprising determining the degree of synchrony between the movements of the thorax and the abdomen.

4. Method according to any of claims 2-3, further comprising determining the air flow through the patient's nasal vestibule.

5. Method according to any of claims 2-4, further comprising a step of filtering frequencies greater than about 0.8 Hz.

6. The method according to claim 5, wherein the filtering step is carried out by means of a Chebyshev low-pass filter of type I I and order 6.

7. Computer program comprising program instructions for making a computer carry out the method according to any one of claims 1 to 6.

8. Computer program according to claim 7, incorporated into storage media.

9. Computer program according to claim 7, supported on a carrier signal.

10. System for obtaining data about the respiratory pattern of a patient, characterized in that it comprises a thermal imager connected to a processing means configured to carry out the procedure of any one of claims 1-7.