Helium inhalation and infrared imaging combined method for three-dimensional reconstruction of airway

Abstract

The invention provides a helium inhalation and infrared imaging combined method for three-dimensional reconstruction of an airway, which comprises the following steps of: S1, aiming at different patients, setting an accurately quantified helium optimal action parameter on a friendly helium inhalation device; s2, introducing helium into the airway of the patient by using a friendly helium inhalation device; s3, using infrared imaging equipment, adjusting parameters of the infrared imaging equipment, and scanning the airway of the patient to obtain an airway image; s4, performing image processing on the airway image; s5, using a three-dimensional reconstruction visualization engine to construct a complete airway three-dimensional model for the obtained airway image; according to the method, the three-dimensional airway image is processed through a helium and infrared scanning method, a multi-mode denoising algorithm, an image registration method based on child respiratory cycle feature recognition and other methods, the accuracy of the three-dimensional airway image is improved, meanwhile, a two-dimensional image is converted into a three-dimensional model through a three-dimensional construction technology, and the accuracy of the three-dimensional model is improved. And the airway three-dimensional model can be reflected more accurately.

Description

Technical Field

[0001] The invention relates to the field of ear, nose and throat examination equipment, and in particular to a helium inhalation combined with infrared imaging method for three-dimensional reconstruction of airways. Background Art

[0002] Airway three-dimensional reconstruction is a technology that uses imaging data to construct a three-dimensional model of the airway. It is currently widely used in respiratory disease diagnosis, preoperative planning, and disease research.

[0003] However, there are still some problems with airway 3D reconstruction technology: 1. Invasive procedures such as bronchoscopy can cause significant pain and discomfort to patients, especially children, resulting in extremely low levels of cooperation and difficulty in completing the examination process, which affects the quality and integrity of image acquisition.

[0004] 2. CT radiation concerns: Patients who undergo repeated CT scans to obtain three-dimensional reconstructed images will significantly increase their long-term risk of developing serious diseases such as cancer in the future, posing health risks.

[0005] 3. Deviations in image quality and accuracy: During the reconstruction process, uncontrollable physiological activities such as human respiratory movement and heartbeat continue to interfere with scanning imaging, generating artifacts, resulting in missing details and blurred edges of airway images, which seriously hinders the precise observation and analysis of the fine structure of the airway and affects the accuracy of lesion identification and positioning. Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a helium inhalation combined with infrared imaging method for airway three-dimensional reconstruction.

[0007] To achieve the above object, the present invention adopts the following technical solution: a helium inhalation combined with infrared imaging method for airway three-dimensional reconstruction, comprising the following steps: S1: Set the optimal helium action parameters on the friendly helium inhalation device for different patients; The friendly helium inhalation device includes an intelligent flow regulation module, a real-time concentration matching subsystem, and an all-round safety monitoring kit; The intelligent flow regulation module is used to regulate the flow of helium, with a flow rate precision control range of 0-2 L / min and an accuracy of ±0.05 L / min; The concentration real-time proportioning subsystem is used to adjust the mixing ratio of helium and air between 10% and 80%, with an error of <±2%; Comprehensive safety monitoring suite includes oxygen concentration and pressure detection, abnormal flow alarm, and respiratory rate monitoring functions; The optimal parameters for helium action are to adjust the helium inhalation concentration to 10% - 80%, the flow rate to 0-2 L / min, and the duration to 3 - 8 minutes, so as to maximize the infrared imaging effect under the premise of ensuring the safety of the patient.

[0008] S2: Use a friendly helium inhalation device to insulate helium into the patient's airway; The helium inhalation concentration is controlled within a safe and effective range of 30% -50%, the helium flow rate is finely adjusted to 0.5 - 2L / min, with a step length of 0.1 L / min, and the duration is flexibly set within the range of 3 - 8 minutes according to age and condition; S3: using an infrared imaging device, adjusting parameters of the infrared imaging device, scanning the patient's airway, and obtaining an airway image; The infrared imaging device has high thermal sensitivity, high resolution, high frame rate and strong anti-interference ability; The high thermal sensitivity is 0.03°C - 0.05°C, the high resolution is 800×600 pixels, and the high speed frame rate is 60 - 100 frames / second; The airway image includes a multi-frame image sequence collected at different stages of the patient's respiratory cycle, and each frame of the image records the two-dimensional thermal distribution and structural information of the helium-containing airway at that moment; The structural information includes the airway bifurcation point and the location of the landmark texture of the tube wall; S4: Perform image processing on the airway image; The following steps are involved: S41: Preprocess airway images using a multimodal denoising algorithm to deeply remove thermal noise; The following steps are involved: S411: aligning a multi-frame image sequence of airway images with different phases of the respiratory cycle; The multi-frame image sequence of the airway image is marked as A, and the multi-frame images of the airway image are marked as A1, A2, A3, ..., An in time sequence; The respiratory cycle is marked as T, and the different stages of the respiratory cycle are marked as T1, T2, T3, ..., Tn; The A1 frame image corresponds to the T1 stage, the A2 frame image corresponds to the T2 stage, the A3 frame image corresponds to the T3 stage, ..., the An frame image corresponds to the Tn stage; S412: Normalizing the pixel value range of the airway image; First, confirm the pixel value range of each frame of the airway image and mark the pixel value as X, where the maximum pixel value is marked as and the lowest pixel value is marked as X. min ; The pixel value is normalized to the interval [0-1] using the normalization formula F, the normalization formula F=(XX min ) / (X Max -X min ); S413: Selecting a Daubechies wavelet basis function to perform wavelet decomposition on the airway image, decomposing the airway image into a low-frequency sub-band and multiple high-frequency sub-bands; S414: using hard threshold and soft threshold strategies for the obtained multiple high frequency sub-bands; The hard threshold is to set the wavelet coefficients whose absolute values ​​are less than the threshold directly to 0, and the soft threshold is to shrink the wavelet coefficients less than the threshold in a ratio of 10:3; S415: performing inverse wavelet transform on the wavelet coefficients after threshold processing to obtain modal data after wavelet transform denoising; The wavelet coefficients include a low-frequency sub-band and a processed high-frequency sub-band; S416: extracting texture features from the modal data; S417: Fusing texture features of data of different modalities, and reconstructing multi-modal denoised data using an inverse transform algorithm based on the fused texture features.

[0009] S42: Use the image registration method based on respiratory cycle feature recognition to accurately correct the image deviation caused by patient movement and position changes; The following steps are involved: S421: Detecting the respiratory cycle; Wrap the breathing belt around the patient's abdomen, and obtain the breathing cycle data by detecting the expansion and contraction of the chest and abdomen. Mark the starting time of the patient's inhalation as M0, the peak of inhalation as M1, and the end of exhalation as M2; the expansion degree of the chest during breathing is marked as y1, and the contraction degree of the chest during breathing is marked as y2; S422: extracting respiratory cycle features and image features; Extract key features of the respiratory cycle from respiratory monitoring data; The key features of the respiratory cycle include respiratory frequency V, inhalation time B1, exhalation time B2, and respiratory amplitude Y; The respiratory frequency V=60 / (M2-M0), where the respiratory frequency V is the number of breaths in 1 minute; The inspiratory time B1=M1-M0, the inspiratory time is the time from the start of inhalation to the peak of inhalation; Exhalation time B2 = M2-M1, exhalation time is the time from the peak of inspiration to the end of exhalation; The respiratory amplitude Y reflects the degree of expansion and contraction of the chest or abdomen during breathing; The image features are airway contour, tracheal structure, and inner wall structure; S423: Select an image at a certain stage of the respiratory cycle as a reference; A certain stage in the respiratory cycle can be the end of exhalation or the end of inspiration; Selecting image features in the image as reference features; S424: Optimizing the initial registration result according to the extracted respiratory cycle features; Matching the extracted respiratory cycle feature with the reference feature, calculating the difference between the respiratory cycle feature and the reference feature, and when the difference reaches a set threshold, the optimization is considered successful; The threshold is the difference between the respiratory cycle feature and the reference feature, which is equal to 0.1 mm; S5: The processed airway image is used to construct a complete airway three-dimensional model using a three-dimensional reconstruction visualization engine; The three-dimensional reconstruction visualization engine converts the processed two-dimensional airway image sequence into a three-dimensional model; It includes the following sub-steps: S51: Constructing airway surface model using Marching Cubes algorithm; S52: Use Laplace smoothing algorithm to smooth the model surface; S53: According to the physiological characteristics of the airway mucosa, add corresponding texture mapping to the model; S54: Use the functions provided by the 3D reconstruction visualization engine to set appropriate lighting models, viewing angles, and colors.

[0010] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention uses helium and infrared imaging technology to perform three-dimensional reconstruction of the patient's airway, thereby avoiding the harm and discomfort to the patient caused by traditional methods; the amount of helium introduced is accurately controlled by the designed algorithm, and a multimodal denoising algorithm and deep learning semantic segmentation model are developed to construct an airway three-dimensional reconstruction visualization engine, thereby improving the accuracy and image quality of airway three-dimensional images. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 This is a diagram of the implementation steps of Example 1 of the present invention. DETAILED DESCRIPTION

[0012] In order to provide a further understanding of the purpose, structure, features, and functions of the present invention, the following detailed description is given in conjunction with the embodiments.

[0013] like Figure 1 As shown, a helium inhalation combined with infrared imaging method for airway three-dimensional reconstruction includes the following steps: S1: Set the optimal helium action parameters on the friendly helium inhalation device for different patients; The friendly helium inhalation device includes an intelligent flow regulation module, a real-time concentration matching subsystem, and an all-round safety monitoring kit; The intelligent flow regulation module is used to regulate the flow of helium, with a flow rate precision control range of 0-2 L / min and an accuracy of ±0.05 L / min; The concentration real-time proportioning subsystem is used to adjust the mixing ratio of helium and air between 10% and 80%, with an error of <±2%; Comprehensive safety monitoring suite includes oxygen concentration and pressure detection, abnormal flow alarm, and respiratory rate monitoring functions; The optimal parameters for helium are to adjust the concentration, flow rate, and duration of helium inhalation to maximize the infrared imaging effect while ensuring patient safety. Helium has high thermal conductivity (about 6 times that of air) and low density. When children inhale an appropriate amount of helium, the gas composition in the airway changes, and the significant increase in thermal conductivity makes the heat transfer in the airway more efficient. During infrared imaging, a more obvious temperature gradient is formed with the surrounding tissue, making the airway boundary present a clearer and sharper outline on the thermal image, which is different from the relatively blurred imaging performance when the airway is filled with air alone. This greatly enhances the imaging contrast and lays the foundation for the subsequent accurate extraction of airway structure information; Such equipment ensures that when helium is introduced into the patient's airway, the helium can be evenly diffused in the patient's airway while ensuring the patient's safety.

[0014] S2: Use a friendly helium inhalation device to insulate helium into the patient's airway; The helium inhalation concentration should be controlled within a safe and effective range of 30% -50%, and the helium flow rate should be finely adjusted to 0.5 -2 L / min, with a step length of 0.1 L / min. The duration should be flexibly set within the range of 3 - 8 minutes according to age and condition; This design adjusts the inhaled helium concentration so that the infrared imaging device can maximize the infrared imaging effect when scanning the patient's airway.

[0015] S3: using an infrared imaging device, adjusting parameters of the infrared imaging device, scanning the patient's airway, and obtaining an airway image; The infrared imaging device has high thermal sensitivity, high resolution, high frame rate and strong anti-interference ability; The high thermal sensitivity is 0.03℃ - 0.05℃, the high resolution is improved to no less than 800×600 pixels, the high-speed frame rate can reach 60 - 100 frames / second, and the strong anti-interference ability is mainly aimed at thermal noise, ambient light interference and slight body movement of children; This design ensures that the impact of artifacts is reduced and that the rapidly changing and subtle temperature, temperature changes, and structural dynamics of the child's airway are captured; Adjusting the parameters of infrared imaging equipment mainly adjusts the infrared wavelength; The infrared wavelength is 8-12 μm; Helium has different absorption and reflection characteristics for infrared light than air. At specific infrared wavelengths (such as 8-12μm), the presence of helium changes the propagation path and reflection pattern of light in the airway, reducing image blur and information loss caused by gas scattering and random reflection; The airway image includes a multi-frame image sequence collected at different stages of the patient's respiratory cycle, and each frame of the image records the two-dimensional thermal distribution and structural information of the helium-containing airway at that moment; The structural information includes the airway bifurcation point and the location of the landmark texture of the tube wall; S4: Perform image processing on the airway image; The following steps are involved: S41: Preprocess airway images using a multimodal denoising algorithm to deeply remove thermal noise; The following sub-steps are included: S411: aligning a multi-frame image sequence of airway images with different phases of the respiratory cycle; The multi-frame image sequence of the airway image is marked as A, and the multi-frame images of the airway image are marked as A1, A2, A3, ..., An in time sequence; The respiratory cycle is marked as T, and the different stages of the respiratory cycle are marked as T1, T2, T3, ..., Tn; The A1 frame image corresponds to the T1 stage, the A2 frame image corresponds to the T2 stage, the A3 frame image corresponds to the T3 stage, ..., the An frame image corresponds to the Tn stage; S412: Normalizing the pixel value range of the airway image; First, confirm the pixel value range of each frame of the airway image and mark the pixel value as X, where the maximum pixel value is marked as and the lowest pixel value is marked as X. min ; The pixel value is normalized to the interval [0-1] using the normalization formula F, the normalization formula F=(XX min ) / (X Max -X min ); S413: Selecting a Daubechies wavelet basis function to perform wavelet decomposition on the airway image, decomposing the airway image into a low-frequency sub-band and multiple high-frequency sub-bands; S414: using hard threshold and soft threshold strategies for the obtained multiple high frequency sub-bands; The hard threshold is to set the wavelet coefficients whose absolute values ​​are less than the threshold directly to 0, and the soft threshold is to shrink the wavelet coefficients less than the threshold in a ratio of 10:3; By threshold processing, the noise components in the high-frequency sub-band can be removed.

[0016] S415: performing inverse wavelet transform on the wavelet coefficients after threshold processing to obtain modal data after wavelet transform denoising; The wavelet coefficients include a low-frequency sub-band and a processed high-frequency sub-band; S416: extracting texture features from the modal data; S417: Fusing texture features of data of different modalities, and reconstructing multi-modal denoised data using an inverse transform algorithm based on the fused texture features.

[0017] This design can address image noise and artifact problems caused by children's activeness and unstable breathing rhythm, deeply eliminate thermal noise and motion artifacts, and robustly enhance image contrast and clarity.

[0018] S42: Use the image registration method based on respiratory cycle feature recognition to accurately correct the image deviation caused by patient movement and position changes; The following sub-steps are included: S421: Detecting the respiratory cycle; Wrap the breathing belt around the patient's abdomen, and obtain the respiratory cycle data by detecting the expansion and contraction of the chest and abdomen. Mark the starting time of the patient's inhalation as M0, the peak of inhalation as M1, and the end of exhalation as M2. The expansion degree of the chest during breathing is marked as y1, and the contraction degree of the chest during breathing is marked as y2; S422: extracting respiratory cycle features and image features; Extract key features of the respiratory cycle from respiratory monitoring data; The key features of the respiratory cycle include respiratory frequency V, inhalation time B1, exhalation time B2, and respiratory amplitude Y; The respiratory frequency V=60 / (M2-M0), where the respiratory frequency V is the number of breaths in 1 minute; The inspiratory time B1=M1-M0, the inspiratory time is the time from the start of inhalation to the peak of inhalation; Exhalation time B2 = M2-M1, exhalation time is the time from the peak of inspiration to the end of exhalation; The respiratory amplitude Y reflects the degree of expansion and contraction of the chest or abdomen during breathing; The image features are airway contour, tracheal structure, and inner wall structure; S423: Select an image at a certain stage of the respiratory cycle as a reference; A certain stage in the respiratory cycle can be the end of exhalation or the end of inspiration; Selecting image features in the image as reference features; S424: Optimizing the initial registration result according to the extracted respiratory cycle features; Matching the extracted respiratory cycle feature with the reference feature, calculating the difference between the respiratory cycle feature and the reference feature, and when the difference reaches a set threshold, the optimization is considered successful; The threshold is the difference between the respiratory cycle feature and the reference feature, which is equal to 0.1 mm; Such a design eliminates image deviation and achieves high-precision alignment of multi-frame image sequences in time and space dimensions.

[0019] S5: The processed airway image is used to construct a complete airway three-dimensional model using a three-dimensional reconstruction visualization engine; The three-dimensional reconstruction visualization engine converts the processed two-dimensional airway image sequence into a three-dimensional model; The following sub-steps are included: S51: Constructing airway surface model using Marching Cubes algorithm; Such an algorithm divides the volume data into small cubic units, determines whether it intersects with the isosurface based on the grayscale value of each unit vertex, then connects the intersecting vertices to form triangular patches, and finally constructs the surface model of the airway.

[0020] S52: Use Laplace smoothing algorithm to smooth the model surface; By using such an algorithm, the uneven or jagged edges on the surface of the reconstructed airway model can be avoided, and the visual effect of the model can be improved without affecting the overall shape of the model.

[0021] S53: According to the physiological characteristics of the airway mucosa, add corresponding texture mapping to the model; This design adds corresponding texture mapping to the model based on the physiological characteristics of the airway mucosa, making the model look closer to the real airway structure.

[0022] S54: Use the functions provided by the 3D reconstruction visualization engine to set a suitable illumination model, viewing angle, and color. Furthermore, the interior of the airway can be set to be transparent so as to observe the internal structure of the airway, and the airway wall can be set to different colors to distinguish different anatomical regions.

[0023] The present invention is a non-invasive operation, which is easier for patients to accept. It uses infrared imaging to avoid the harm of CT scanning to patients. At the same time, helium is non-toxic and can be used clinically for respiratory support, improving coronary artery disease, and treating upper respiratory tract obstruction. Therefore, it can reduce respiratory irritation to children and reduce the occurrence of laryngeal spasm, tracheospasm, etc. The three-dimensional model constructed by the present invention is more accurate and more complete, and it presents the overall structure of the airway, the direction of the branches, and the precise spatial position and volume of the internal lesions in three dimensions.

[0024] The present invention has been described by the above-mentioned relevant embodiments, however, the above-mentioned embodiments are only examples for implementing the present invention. It must be pointed out that the disclosed embodiments do not limit the scope of the present invention. On the contrary, changes and modifications made without departing from the spirit and scope of the present invention are all within the scope of patent protection of the present invention.

Claims

1. A helium inhalation combined with infrared imaging method for airway three-dimensional reconstruction, characterized in that: The following steps are involved: S1: Set the optimal helium action parameters on the friendly helium inhalation device for different patients; The friendly helium inhalation device includes an intelligent flow regulation module, a real-time concentration matching subsystem, and an all-round safety monitoring kit; The intelligent flow regulation module is used to regulate the flow of helium, with a flow rate precision control range of 0-2 L / min and an accuracy of ±0.05 L / min; The concentration real-time proportioning subsystem is used to adjust the mixing ratio of helium and air between 10% and 80%, with an error of <±2%; Comprehensive safety monitoring suite includes oxygen concentration and pressure detection, abnormal flow alarm, and respiratory rate monitoring functions; The optimal helium action parameters are to adjust the helium inhalation concentration to 10% - 80%, the flow rate to 0-2 L / min, and the duration to 3-8 minutes, so as to maximize the infrared imaging effect under the premise of ensuring the safety of the patient; S2: Use a friendly helium inhalation device to insulate helium into the patient's airway; The helium inhalation concentration should be controlled within a safe and effective range of 30% -50%, and the helium flow rate should be finely adjusted to 0.5 - 2L / min, with a step length of 0.1 L / min. The duration should be within the range of 3 - 8 minutes depending on age and condition. S3: using an infrared imaging device, adjusting parameters of the infrared imaging device, scanning the patient's airway, and obtaining an airway image; The infrared imaging device has high thermal sensitivity, high resolution, high frame rate and strong anti-interference ability; The high thermal sensitivity is 0.03℃ - 0.05℃, the high resolution is improved to no less than 800×600 pixels, the high-speed frame rate can reach 60 - 100 frames / second, and the strong anti-interference ability is mainly aimed at thermal noise, ambient light interference and slight body movement of children; Adjusting the parameters of infrared imaging equipment mainly adjusts the infrared wavelength; The infrared wavelength is 8-12 μm; The airway image includes a multi-frame image sequence collected at different stages of the patient's respiratory cycle, and each frame of the image records the two-dimensional thermal distribution and structural information of the helium-containing airway at that moment; The structural information includes the airway bifurcation point and the location of the landmark texture of the tube wall; S4: Perform image processing on the airway image; The following steps are involved: S41: Preprocess airway images using a multimodal denoising algorithm to deeply remove thermal noise; S42: Use the image registration method based on respiratory cycle feature recognition to accurately correct the image deviation caused by patient movement and position changes; S5: The processed airway image is used to construct a complete airway three-dimensional model using a three-dimensional reconstruction visualization engine; The three-dimensional reconstruction visualization engine converts the processed two-dimensional airway image sequence into a three-dimensional model; S51: Constructing airway surface model using Marching Cubes algorithm; S52: Use Laplace smoothing algorithm to smooth the model surface; S53: According to the physiological characteristics of the airway mucosa, add corresponding texture mapping to the model; S54: Use the functions provided by the 3D reconstruction visualization engine to set appropriate lighting models, viewing angles, and colors.

2. The helium inhalation combined with infrared imaging method for airway three-dimensional reconstruction according to claim 1, characterized in that: The step S41 comprises the following steps: S411: aligning a multi-frame image sequence of airway images with different phases of the respiratory cycle; The multi-frame image sequence of the airway image is marked as A, and the multi-frame images of the airway image are marked as A1, A2, A3, ..., An in time sequence; The respiratory cycle is marked as T, and the different stages of the respiratory cycle are marked as T1, T2, T3, ..., Tn; The A1 frame image corresponds to the T1 stage, the A2 frame image corresponds to the T2 stage, the A3 frame image corresponds to the T3 stage, ..., the An frame image corresponds to the Tn stage; S412: Normalizing the pixel value range of the airway image; First, confirm the pixel value range of each frame of the airway image and mark the pixel value as X, where the maximum pixel value is marked as and the lowest pixel value is marked as X. min ; The pixel value is normalized to the interval [0-1] using the normalization formula F, the normalization formula F=(XX min ) / (X Max -X min ); S413: Selecting a Daubechies wavelet basis function to perform wavelet decomposition on the airway image, decomposing the airway image into a low-frequency sub-band and multiple high-frequency sub-bands; S414: using hard threshold and soft threshold strategies for the obtained multiple high frequency sub-bands; The hard threshold is to set the wavelet coefficients whose absolute values ​​are less than the threshold directly to 0, and the soft threshold is to shrink the wavelet coefficients less than the threshold in a ratio of 10:3; S415: performing inverse wavelet transform on the wavelet coefficients after threshold processing to obtain modal data after wavelet transform denoising; The wavelet coefficients include a low-frequency sub-band and a processed high-frequency sub-band; S416: extracting texture features from the modal data; S417: Fusing texture features of data of different modalities, and reconstructing multi-modal denoised data using an inverse transform algorithm based on the fused texture features.

3. The helium inhalation combined with infrared imaging method for airway three-dimensional reconstruction according to claim 1, characterized in that: The step S42 comprises the following steps: S421: Detecting the respiratory cycle; Wrap the breathing belt around the patient's abdomen, and obtain the respiratory cycle data by detecting the expansion and contraction of the chest and abdomen. Mark the starting time of the patient's inhalation as M0, the peak of inhalation as M1, and the end of exhalation as M2. The expansion degree of the chest during breathing is marked as y1, and the contraction degree of the chest during breathing is marked as y2; S422: extracting respiratory cycle features and image features; Extract key features of the respiratory cycle from respiratory monitoring data; The key features of the respiratory cycle include respiratory frequency V, inhalation time B1, exhalation time B2, and respiratory amplitude Y; The respiratory frequency V=60 / (M2-M0), where the respiratory frequency V is the number of breaths in 1 minute; The inspiratory time B1=M1-M0, the inspiratory time is the time from the start of inhalation to the peak of inhalation; Exhalation time B2 = M2-M1, exhalation time is the time from the peak of inspiration to the end of exhalation; The respiratory amplitude Y reflects the degree of expansion and contraction of the chest or abdomen during breathing; The image features are airway contour, tracheal structure, and inner wall structure; S423: Select an image at a certain stage of the respiratory cycle as a reference; A certain stage in the respiratory cycle can be the end of exhalation or the end of inspiration; Selecting image features in the image as reference features; S424: Optimizing the initial registration result according to the extracted respiratory cycle features; Matching the extracted respiratory cycle feature with the reference feature, calculating the difference between the respiratory cycle feature and the reference feature, and when the difference reaches a set threshold, the optimization is considered successful; The threshold is a difference between the respiratory cycle feature and the reference feature equal to 0.1 mm.

Images