Image analysis device and program

Abstract

To support a user to appropriately determine presence or absence of a complication due to mounting of an artificial respirator.SOLUTION: A control part of a console acquires an image obtained by executing still image capturing or kymography for a subject at least during mounting of an artificial respirator or in a predetermined time after the extubation, calculates a feature amount related to a predetermined structure on the basis of the acquired image, and creates information on presence or absence of a complication related on the artificial respirator of the subject on the basis of the calculated feature amount.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to an image analysis apparatus and a program.

Background Art

[0002] Conventionally, based on a plurality of frame images showing the dynamics of the chest obtained by radiographing the chest of a subject wearing a ventilator, the amount of morphological change of a predetermined structure in the chest is calculated, and based on the calculated amount of morphological change, a technique for evaluating the respiratory state of the subject when wearing or extubating the ventilator has been proposed (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, there is a risk of complications when wearing a ventilator, but Patent Document 1 does not mention complications.

[0005] An object of the present invention is to assist a user in appropriately determining the presence or absence of complications due to wearing a ventilator.

Means for Solving the Problems

[0006] To solve the above problems, an image analysis apparatus of the present invention includes: an acquisition unit that acquires a radiation image including a plurality of frame images obtained by performing dynamic imaging on a subject during wearing of a ventilator; a generation unit that generates information regarding the presence or absence of complications related to the ventilator of the subject based on a predetermined frame image among the acquired radiation images including the plurality of frame images; Equipped with 、 The generation unit calculates information regarding the size of the left and right lung fields based on images obtained by performing dynamic imaging of the subject's chest while the ventilator is attached, and generates the calculated information as information regarding the presence or absence of complications related to the ventilator in the subject.

[0007] The program of the present invention, Computers An acquisition unit that acquires radiographic images including multiple frame images obtained by performing dynamic imaging on a subject while on a ventilator. A generation unit generates information regarding the presence or absence of complications related to the ventilator of the subject based on a predetermined frame image from among the acquired multiple frame images of the radiographic image. To function as 、 The generation unit calculates information regarding the size of the left and right lung fields based on images obtained by performing dynamic imaging of the subject's chest while the ventilator is attached, and generates the calculated information as information regarding the presence or absence of complications related to the ventilator in the subject. [Effects of the Invention]

[0008] According to the present invention, it becomes possible to support users in appropriately determining whether or not complications have occurred due to the use of a ventilator. [Brief explanation of the drawing]

[0009] [Figure 1] This figure shows the overall configuration of the dynamic analysis system according to this embodiment. [Figure 2] This is a block diagram showing the functional configuration of the console in Figure 1. [Figure 3] This diagram schematically shows the movement of the vocal cords in a frontal radiographic image of the airway. [Figure 4] This diagram schematically shows a normal larynx and a larynx with laryngeal edema in a radiographic image of the airway (lateral view). [Figure 5] This diagram schematically shows the trachea during exhalation and inspiration in a frontal radiographic image of the airway of a patient with tracheal stenosis. [Figure 6] Figure 2 is a flowchart showing the airway complication evaluation process performed by the control unit. [Figure 7] This figure shows an example of how to calculate the length of the airway diameter in the larynx. [Figure 8]It is a diagram showing an example of an evaluation screen displayed on a display unit in step S6 of FIG. 6. [Figure 9] It is a diagram showing an example of an evaluation screen displayed on a display unit in step S6 of FIG. 6. [Figure 10] It is a flowchart showing the pulmonary complication evaluation process executed by the control unit of FIG. 2. [Figure 11] It is a diagram showing an example of an evaluation screen displayed on a display unit in step S26 of FIG. 10. [Figure 12] It is a diagram showing an example of an evaluation screen displayed on a display unit in step S26 of FIG. 10. [Figure 13] It is a flowchart showing the cardiac complication evaluation process executed by the control unit of FIG. 2. [Figure 14] It is a diagram showing an example of an evaluation screen displayed on a display unit in step S36 of FIG. 13.

Mode for Carrying Out the Invention

[0010] Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. Note that the present invention is not limited to the illustrated examples.

[0011] (Configuration of Dynamic Analysis System 100) First, the configuration of the embodiment according to the present invention will be described. FIG. 1 shows an example of the overall configuration of the dynamic analysis system 100 in the present embodiment. The dynamic analysis system 100 is, for example, a system for making rounds for taking pictures of patients who are difficult to move and are staying in an intensive care unit or an operating room as subjects. The dynamic analysis system 100 comprises a radiation generator 1, a console 2, an access point 3, and an FPD (Flat Panel Detector) cassette 4. The radiation generator 1 is configured as a mobile rounds vehicle having wheels and installed with the console 2 and the access point 3. In the dynamic analysis system 100, the console 2 can be communicatively connected to the radiation generator 1 and the FPD cassette 4 via the access point 3.

[0012] As shown in Figure 1, the dynamic analysis system 100 is brought into an operating room (intensive care unit) Rc, etc., and with the FPD cassette 4 inserted, for example, between the subject H lying on bed B and bed B, or into an insertion opening provided on the opposite side of bed B from the subject H (not shown), the system irradiates the subject H with radiation from the portable radiation source 11 of the radiation generator 1 to take dynamic images or still images of the subject H. Dynamic imaging refers to acquiring multiple images by repeatedly irradiating a subject H with pulsed radiation such as X-rays at predetermined time intervals (pulsed irradiation), or by continuously irradiating them at a low dose rate without interruption (continuous irradiation). A series of images obtained through dynamic imaging is called a dynamic image. Each of the multiple images that make up a dynamic image is called a frame image. Dynamic photography includes video recording, but does not include taking still images while displaying video. Similarly, dynamic images include video, but do not include images obtained by taking still images while displaying video.

[0013] The following describes each component of the dynamic analysis system 100. The radiation generating device 1 is comprised of a radiation source 11 that emits radiation, a radiation irradiation control unit 12, an exposure switch 13, and the like.

[0014] The radiation source 11 irradiates the subject H with radiation (X-rays) according to the control of the radiation irradiation control unit 12. The radiation irradiation control unit 12 controls the radiation source 11 based on the radiation irradiation conditions transmitted from the console 2 to perform radiography (dynamic or still image capture). The radiation irradiation conditions input from the console 2 include, for example, tube current, tube voltage, radiation irradiation time, frame rate (number of frames captured per unit time (1 second)), total exposure time or total number of frames captured per exposure, and type of additional filter. When the exposure switch 13 is pressed, it inputs a radiation irradiation instruction signal to the console 2.

[0015] Console 2 outputs radiation irradiation conditions to the radiation generator 1 according to the input examination information, and outputs image reading conditions to the FPD cassette 4 to control radiation irradiation and radiation image reading operations. It also acts as an image analysis device, analyzing radiation images (dynamic or still images) transmitted from the FPD cassette 4, and generating and displaying information regarding the presence or absence of complications related to the ventilator in subject H.

[0016] Figure 2 shows an example of the functional configuration of console 2. As shown in Figure 2, console 2 is composed of a control unit 21, a storage unit 22, an operation unit 23, a display unit 24, a communication unit 25, a connector 26, etc., and each unit is connected by a bus 27.

[0017] The control unit 21 includes a CPU (Central Processing Unit) and RAM (Random Access Memory). The control unit 21 is composed of the following components. The CPU of the control unit 21 reads system programs and various processing programs stored in the memory unit 22 in response to the operation of the operation unit 23, expands them into RAM, and centrally controls the operation of each part of the console 2, as well as the operation of the radiation generator 1 and the FPD cassette 4, according to the expanded programs. The control unit 21 also functions as an acquisition unit and a generation unit, executing various processes, including the airway complication evaluation process, pulmonary complication evaluation process, and cardiac complication evaluation process, which will be described later, according to the expanded programs.

[0018] The memory unit 22 is composed of non-volatile semiconductor memory, a hard disk, or the like. The memory unit 22 stores data such as various programs executed by the control unit 21, parameters necessary for processing by those programs, or processing results. For example, the memory unit 22 stores programs for executing the airway complication evaluation process, pulmonary complication evaluation process, and cardiac complication evaluation process, which will be described later. The various programs are stored in the form of readable program code, and the control unit 21 sequentially executes operations according to the program code. Furthermore, the memory unit 22 stores the radiation irradiation conditions and image reading conditions during dynamic imaging. The radiation irradiation conditions and image reading conditions can be set by the user using the operation unit 23.

[0019] Furthermore, the memory unit 22 stores the radiographic images transmitted from the FPD cassette 4 in association with patient information (attribute information) of the subject H at the time of imaging, examination information (examination date, examination target area (e.g., chest, airway, etc.), imaging type (dynamic imaging / still imaging), imaging period (e.g., before / during / after extubation of the ventilator)), evaluation items (e.g., airway complications, pulmonary complications, etc.)), calculated feature quantities, and information regarding the presence or absence of generated complications. Furthermore, the memory unit 22 may store statistical data of the results of calculations obtained from radiographic images of healthy individuals, which are used to calculate the characteristic quantities of predetermined structures calculated in the airway complication evaluation process described later.

[0020] The operation unit 23 is configured with a keyboard equipped with cursor keys, number input keys, and various function keys, and a pointing device such as a mouse, and outputs instruction signals input by key operations on the keyboard or mouse to the control unit 21. The operation unit 23 may also be equipped with a touch panel on the display screen of the display unit 24, in which case it outputs instruction signals input via the touch panel to the control unit 21.

[0021] The display unit 24 is composed of monitors such as LCDs (Liquid Crystal Displays) and CRTs (Cathode Ray Tubes), and displays input instructions and data from the operation unit 23 according to the instructions of the display signals input from the control unit 21.

[0022] The communication unit 25 is equipped with a wireless LAN adapter and controls the transmission and reception of data between the radiation generator 1, FPD cassette 4, and other external devices connected to a communication network such as a wireless LAN via the access point 3.

[0023] Connector 26 is a connector for communicating with the FPD cassette 4 via a cable (not shown).

[0024] Returning to Figure 1, access point 3 relays communication between the radiation generator 1 and console 2, as well as communication between console 2 and FPD cassette 4, etc.

[0025] The FPD cassette 4 is a portable radiation detector compatible with dynamic imaging. The FPD cassette 4 is constructed by arranging multiple radiation detection elements in a matrix (two-dimensional) at predetermined positions on a substrate such as a glass substrate. These elements detect radiation irradiated from a radiation source 11 that has passed through at least the subject H, according to its intensity, and convert the detected radiation into an electrical signal for storage. A switching unit, such as a TFT (Thin Film Transistor), is connected to each radiation detection element, and the storage and reading of electrical signals to each radiation detection element is controlled by the switching unit, thereby acquiring image data (frame images). FPDs can be of the indirect conversion type, which converts radiation into an electrical signal via a scintillator using a photoelectric conversion element, or the direct conversion type, which directly converts radiation into an electrical signal; either type may be used.

[0026] The FPD cassette 4 includes a read control unit that controls the accumulation and reading of electrical signals by the switching unit, and a communication unit for communicating with the console 2 via the access point 3 (neither of which are shown). Image reading conditions such as frame rate, number of captured frames per scan, and image size (matrix size) are set by the console 2 via the communication unit. Based on the set image reading conditions, the read control unit controls the accumulation and reading of electrical signals to each radiation detection element by the switching unit. The FPD cassette 4 also has a connector and can communicate with the console 2 via a cable (not shown).

[0027] While the FPD cassette 4 may be carried by the radiologist or other person performing the imaging, it is relatively heavy and could break or malfunction if dropped. Therefore, it is designed to be transported in a cassette pocket 61a provided on the mobile medical cart.

[0028] (Operation of the dynamic analysis system 100) Next, we will describe the operation of the dynamic analysis system 100. Complications can occur when a patient is placed on a ventilator. Potential complications associated with ventilator use include airway complications, lung complications, and cardiac complications.

[0029] In the console 2 of this embodiment, it is possible to generate information regarding the presence or absence of complications related to the airway, the lungs, and the heart, based on radiographic images (dynamic or still images) obtained by taking dynamic or still images while the ventilator is attached or after extubation. The following describes the process for generating information regarding the presence or absence of each complication.

[0030] (Airway complication assessment and processing) Airway complication assessment processing is a process for generating information regarding the presence or absence of airway complications. Airway complications include vocal cord paralysis, laryngeal edema, and tracheal stenosis.

[0031] Vocal cord paralysis is a condition in which the vocal cords remain open and do not close even when speaking. Figure 3 schematically shows the movement of the vocal cords in a radiographic image of the airway (front view). In Figure 3, the label 51 indicates the vocal cords. As shown in Figure 3, the vocal cords 51 close when speaking and spread when not speaking, but when vocal cord paralysis occurs, the vocal cords remain open.

[0032] Laryngeal edema is a condition in which the mucous membrane inside the larynx swells, impairing breathing. Figure 4 schematically shows a normal larynx and a larynx with laryngeal edema (abnormality) in a radiographic image of the airway (lateral view). In Figure 4, reference numeral 52 indicates the larynx, and 521 indicates laryngeal edema. When laryngeal edema occurs, the larynx is compressed and narrowed compared to a normal larynx, as shown in Figure 4.

[0033] Tracheal stenosis is a condition in which the trachea narrows during inspiration, resulting in a difference in tracheal length between inspiration and exhalation. Figure 5 schematically shows the trachea during inspiration and exhalation in a radiographic image (front view) of the airway of a patient with tracheal stenosis. In Figure 5, the label 53 refers to the trachea.

[0034] In this case, because the airway is widened by a tube while the patient is on a ventilator, it is difficult to identify airway complications without extubating the ventilator. Therefore, in this embodiment, an airway complication evaluation process is performed within a predetermined time after extubation from a ventilator, and information regarding airway complications is generated.

[0035] Figure 6 is a flowchart showing the flow of the airway complication assessment process performed in console 2. The airway complication assessment process is performed in cooperation with the control unit 21 and the program stored in the memory unit 22. The airway complication assessment process will be described below. In the explanation of Figure 6, we describe a case where dynamic imaging is performed to acquire dynamic images, and information regarding the presence or absence of airway complications is generated based on the acquired dynamic images. Furthermore, in the dynamic analysis system 100, it is assumed that dynamic images acquired by dynamic imaging of the subject H's airway before the ventilator is attached are stored in the storage unit 22.

[0036] First, the control unit 21 receives input from the operation unit 23 of patient H's information (name, age, sex, disease, etc.) and examination information (area to be examined (here, for example, airway), type of imaging (here, dynamic imaging), time of imaging (here, after extubation), evaluation items (here, airway complications)) (step S1).

[0037] Next, the control unit 21 controls the radiation irradiation control unit 12 and the FPD cassette 4 based on the input examination information to perform dynamic imaging including the airway of the subject H in response to the pressing of the exposure switch 13, and acquires a dynamic image of the airway of the subject H after extubation from the ventilator (step S2).

[0038] Next, the control unit 21 calculates characteristic features of predetermined structures related to airway complications from the acquired dynamic images (step S3).

[0039] For example, the control unit 21 calculates the amount of vocal cord movement or the amount of change in glottal width as a feature of the vocal cords. For example, the control unit 21 recognizes the vocal cords from each frame image of the acquired dynamic image using image processing such as edge detection or machine learning, tracks the recognized point P on the vocal cord (for example, the part indicated by point P on the vocal cord 51 in Figure 3) from each frame image, and calculates the amount of movement (maximum movement) of point P. Alternatively, the width of the glottis may be calculated from each frame image, and the difference between the maximum and minimum values ​​may be calculated as the amount of change in glottal width.

[0040] Furthermore, for example, the control unit 21 calculates the length of the airway diameter of the larynx (indicated by reference numeral 522 in Figure 7) as a characteristic feature of the larynx, as shown in Figure 7. For example, the control unit 21 recognizes the larynx from predetermined frame images of the acquired dynamic image using image processing such as edge detection or machine learning, and calculates the airway diameter of the recognized larynx. Since the larynx extends in the vertical direction, the maximum and minimum diameters of the larynx may also be calculated. Furthermore, when laryngeal edema occurs, the diameter of the laryngeal airway may narrow in the depth direction, as shown in Figure 4. In this case, in a radiographic image of the airway taken from the front, the amount of radiation transmitted through the laryngeal edema decreases, and therefore the signal value (density) in this area decreases. Thus, the control unit 21 may calculate the signal value of the larynx (representative value, for example, mean or median) as a characteristic value of the larynx.

[0041] Furthermore, for example, the control unit 21 calculates the amount of tracheal wall movement or change in tracheal diameter at predetermined positions within the trachea from the larynx to the bronchi as a characteristic feature of the trachea. For example, the control unit 21 recognizes the airway from below the larynx to the bronchi as the trachea using image processing such as edge detection or machine learning from each frame image of the acquired dynamic image, and calculates the amount of tracheal wall movement (maximum movement) at a predetermined position (a predetermined position in the vertical direction) of the recognized trachea. Alternatively, the tracheal diameter at a predetermined position (a predetermined position in the vertical direction) of the recognized trachea may be calculated from each frame image, and the difference between the maximum and minimum values ​​may be calculated as the change in tracheal diameter. Furthermore, as mentioned above, when tracheal stenosis occurs, the tracheal diameter changes during exhalation and inhalation. In particular, when the diameter changes in the depth direction, the amount of radiation transmitted in the radiographic image taken from the front of the airway changes, and therefore the signal value (density) changes. For this reason, the control unit 21 may calculate the amount of change in the tracheal signal value (representative value, for example, mean or median) (signal value change) as a characteristic quantity of the trachea. Alternatively, the ratio of the upper airway to the lower airway may be determined as a characteristic feature of the trachea.

[0042] Next, the control unit 21 acquires characteristic features of predetermined structures related to airway complications in the dynamic image of the subject H's airway before the ventilator is attached (step S4). The control unit 21 reads dynamic images of the subject H's airway before the ventilator was attached from the memory unit 22, and performs the same processing on the read dynamic images as described in step S3 to obtain characteristic quantities of predetermined structures related to airway complications in the dynamic images of the subject H's airway before the ventilator was attached. Alternatively, the characteristic quantities of predetermined structures related to airway complications may be calculated in advance from dynamic images of the airway of subject H before the ventilator is attached and stored in the memory unit 22, and in step S4, the characteristic quantities may be retrieved from the memory unit 22.

[0043] Next, the control unit 21 generates comparative information of the characteristic quantities of a predetermined structure before and after the attachment of the ventilator as information regarding the presence or absence of airway complications (step S5). Then, the control unit 21 displays an evaluation screen 241 on the display unit 24 that shows information regarding the presence or absence of complications related to the generated airway (step S6), and terminates the airway complication evaluation process.

[0044] Figure 8 shows an example of the evaluation screen 241 displayed on the display unit 24 in step S6. As shown in Figure 8, the evaluation screen 241 displays, for example, patient information 241a of the patient being evaluated (subject H), dynamic images 241b acquired in step S2, vocal cord characteristics (here, vocal cord displacement) 241c before and after ventilator attachment and extubation, and larynx characteristics (here, laryngeal airway diameter) 241d before and after ventilator attachment and extubation.

[0045] Figure 9 shows an example of the evaluation screen 242 displayed on the display unit 24 in step S6. As shown in Figure 9, the evaluation screen 242 displays, for example, patient information 242a of the patient being evaluated, a frame image 242b of the maximum expiratory position of the dynamic image acquired in step S2, a frame image 242c of the maximum inspiratory position, and tracheal characteristics (in this case, change in tracheal diameter) 242d before and after the attachment of the ventilator.

[0046] Thus, evaluation screens 241 and 242 display comparative information regarding the presence or absence of airway complications, including characteristic quantities of predetermined structures related to airway complications (in this case, vocal cord displacement, laryngeal airway diameter, and tracheal diameter change) before and after the use of a ventilator. This allows users to easily identify the presence or absence of airway complications, such as vocal cord paralysis, laryngeal edema, and tracheal stenosis.

[0047] Furthermore, on the evaluation screen 241, the dynamic images acquired in step S2 may be displayed as a video, or only representative frame images may be displayed as dynamic image 241b. Additionally, dynamic images (or representative frame images) before and after extubation may be displayed side by side, and annotations may be added to the parts of the displayed dynamic images (or representative frame images) where vocal cord movement and airway diameter were measured.

[0048] In the above-described airway complication evaluation process, we explained that dynamic images of the airway are taken and the resulting dynamic images are analyzed to generate information regarding the presence or absence of ventilator-related complications. However, it is also possible to take still images of the airway and analyze the resulting still images to generate information regarding the presence or absence of ventilator-related complications. For example, since laryngeal edema does not involve movement, the laryngeal airway diameter or signal value of the larynx can be calculated from still images obtained by taking still images of the airway of subject H after extubation from the ventilator. Based on a comparison with the laryngeal airway diameter or signal value of the larynx before subject H was placed on the ventilator, information regarding the presence or absence of ventilator-related complications can be generated. Furthermore, for example, after extubation from a ventilator, still images of the subject H's airway can be taken during both exhalation and inhalation. From these still images, the change in tracheal diameter, tracheal wall displacement, or tracheal signal value can be calculated. Based on a comparison with the changes in tracheal diameter, tracheal wall displacement, or tracheal signal value during exhalation and inhalation of subject H before ventilator use, information regarding the presence or absence of ventilator-related complications (tracheal stenosis) can be generated.

[0049] Furthermore, laryngeal edema can be detected in both images of the airway taken from the front (see Figure 7) and images taken from the side (see Figure 5). Therefore, information regarding the presence or absence of ventilator-related complications may be generated using either one of these images. Alternatively, if laryngeal edema is suspected based on a comparison between one feature and a predetermined threshold, information regarding the presence or absence of ventilator-related complications (tracheal stenosis) may also be generated using the other image.

[0050] Furthermore, in the above-described airway complication evaluation process, comparative information is generated by comparing the characteristic quantities after extubation from a ventilator with the characteristic quantities of the same patient before ventilator use, as information regarding the presence or absence of ventilator-related complications. However, the process is not limited to this, and comparative information is also to be generated by comparing the characteristic quantities after extubation from a ventilator with statistical data from healthy individuals (for example, statistical data obtained by calculating the characteristic quantities calculated in the airway complication evaluation process from radiographic images of healthy individuals) as information regarding the presence or absence of ventilator-related complications.

[0051] Furthermore, in the above-described airway complication evaluation process, information regarding the presence or absence of complications is generated for all items related to airway complications, including vocal cord paralysis, laryngeal edema, and tracheal stenosis. However, the system may be configured so that the user can select which items to generate information for from the control unit 23.

[0052] Additionally, an alert may be issued (displayed, audio output, etc.) if the difference between the feature quantities after extubation from a ventilator and the feature quantities (or statistical data) before ventilator placement exceeds a predetermined threshold.

[0053] (Pulmonary complication assessment) The pulmonary complication assessment process is a process for generating information regarding the presence or absence of lung complications. Lung-related complications include pneumothorax, pneumonia, atelectasis, pulmonary edema, and pleural effusion. These complications primarily occur during mechanical ventilation. Therefore, in this embodiment, a pulmonary complication evaluation process is performed while the ventilator is attached, and information regarding lung complications is generated.

[0054] Figure 10 is a flowchart showing the flow of the pulmonary complication assessment process performed in console 2. The pulmonary complication assessment process is performed through the cooperation of the control unit 21 and the program stored in the memory unit 22. The pulmonary complication assessment process will be described below. In the explanation of Figure 10, we describe a case where dynamic imaging is performed to acquire dynamic images, and information regarding the presence or absence of lung complications is generated based on the acquired dynamic images. Furthermore, in the dynamic analysis system 100, it is assumed that dynamic images acquired by dynamic imaging of the subject H's chest before the ventilator is attached, and dynamic images and feature quantities acquired in a pulmonary complication evaluation process previously performed while the subject was on the ventilator, are stored in the memory unit 22.

[0055] First, the control unit 21 receives input from the operation unit 23 of patient information (name, age, sex, disease, etc.) and examination information (area to be examined (here, for example, chest), type of imaging (here, dynamic imaging), time of imaging (here, while wearing the device), evaluation items (here, for example, pulmonary complications)) (step S21).

[0056] Next, the control unit 21 controls the radiation irradiation control unit 12 and the FPD cassette 4 based on the input examination information to perform dynamic imaging of the chest of the subject H, who is on a ventilator, in response to the pressing of the exposure switch 13, and acquires a dynamic image (step S22).

[0057] Next, the control unit 21 calculates characteristic features of predetermined structures related to lung complications from the acquired dynamic images (step S23).

[0058] Here, among the complications that can occur while on a ventilator, pneumonia is a condition in which pathogens that have entered through the airway multiply in the lungs, causing inflammation. Atelectasis is a condition in which air does not reach the lung tissue for some reason, resulting in a lack of air in part or all of the lung, causing the lung to collapse. Pulmonary edema is a condition in which fluid components from the blood leak out and accumulate in the alveoli. Pleural effusion is a condition in which fluid accumulates abnormally in the pleural cavity. In radiographic images of lungs affected by pneumonia, atelectasis, pulmonary edema, or pleural effusion, the signal intensity (density) within the lung field is lower than normal. Furthermore, when atelectasis or pleural effusion is present, the lung area is smaller than normal. Pulmonary edema is often accompanied by cardiomegaly, in which case the cardiothoracic ratio (the ratio of heart width to chest width) is larger than normal.

[0059] Therefore, in step S23, for example, the control unit 21 calculates at least one of the following as characteristic quantities of a predetermined structure related to lung complications: signal values ​​(representative values, e.g., mean or median) of the left and right lung field regions, lung field area, and cardiothoracic ratio. For example, the control unit 21 recognizes the lung region from each frame image of the acquired dynamic image using known image processing such as edge detection or machine learning, and calculates the signal value (representative value, e.g., mean or median) and lung area for each of the left and right lung regions. Then, for example, the representative values ​​of the signal value and lung area calculated from each frame image (e.g., maximum value, minimum value, mean value, etc.) are used as features. In addition, the control unit 21 recognizes the cardiac region from a predetermined frame image of the acquired dynamic image (e.g., a frame image of the maximum inspiratory position) using known image processing such as template matching or machine learning, and calculates the cardiothoracic ratio. The control unit 21 associates the calculated feature quantities with patient information, examination information, and dynamic images, and stores them in the storage unit 22.

[0060] On the other hand, in pneumothorax, the lung field collapses within the pleural cavity, so it is necessary to recognize the lung field area within the pleural cavity. However, in general lung field area recognition, the area within the contour of the pleural cavity is recognized as the lung field area. Therefore, the control unit 21 applies frequency enhancement processing to each frame image of the dynamic image to emphasize high-frequency components and generates an image with the ribs removed. From the generated image, it recognizes the first lung field region (the region surrounded by the contour of the thoracic cavity) by edge detection, etc. (see R1 in Figure 12). The control unit 21 also performs edge detection, etc., on the recognized first lung field region to recognize the second lung field region surrounded by the visceral pleura (see R2 in Figure 12). Then, it calculates information regarding the size or change of each lung field, for example, the ratio of the area of ​​the second lung field region to the area of ​​the first lung field region (the area within the contour of the thoracic cavity) recognized from the frame images at maximum expiratory position and maximum inspiratory position (referred to as the lung ratio), for each lung, and uses this as a feature quantity.

[0061] Next, the control unit 21 acquires characteristic features of predetermined structures related to lung complications from dynamic images taken before the ventilator was attached and dynamic images taken previously while the ventilator was attached (step S24). The control unit 21 reads dynamic images of the subject H taken before and during the use of a ventilator from the memory unit 22, calculates feature quantities related to pneumonia, atelectasis, pulmonary edema, and pleural effusion as described in step S23 for the read dynamic images, and obtains feature quantities related to pneumonia, atelectasis, pulmonary edema, and pleural effusion for predetermined structures related to lung complications in past dynamic images taken before and during the use of a ventilator. If feature quantities related to pneumonia, atelectasis, pulmonary edema, and pleural effusion have already been calculated from past dynamic images and stored in the memory unit 22, then in step S24, the feature quantities are retrieved from the memory unit 22.

[0062] Next, the control unit 21 generates information regarding the presence or absence of lung complications based on the feature quantities calculated in steps S23 and S24 (step S25). For example, the control unit 21 generates information showing the temporal changes in at least one of the signal values ​​of the lung field region, lung field area, and cardiothoracic ratio before and during mechanical ventilation, as information regarding the presence or absence of pneumonia, atelectasis, pulmonary edema, and pleural effusion. Images from before and during mechanical ventilation (past) and images (video or representative frame images) acquired in step S22 may be placed side by side to generate information regarding the presence or absence of pneumonia, atelectasis, pulmonary edema, and pleural effusion. Furthermore, information is generated by applying frequency-enhanced processing to the frame images of the maximum expiratory and maximum inspiratory positions acquired in at least step S22 and arranging them in a comparable manner, and / or information regarding the size of the lung field or the amount of change therein, as information regarding the presence or absence of pneumothorax.

[0063] Then, the control unit 21 displays an evaluation screen 241 on the display unit 24 that shows information regarding the presence or absence of lung complications (step S26), and terminates the lung complication evaluation process.

[0064] Figure 11 shows an example of the evaluation screen 243 displayed on the display unit 24 in step S26. Figure 11 shows an example of the evaluation screen 243 displaying information regarding the presence or absence of pneumonia, atelectasis, pulmonary edema, and pleural effusion. As shown in Figure 11, the evaluation screen 243 displays, for example, patient information 243a of the patient being evaluated, dynamic images 243b (which may be representative frame images or videos) acquired before or during (in the past) placement of a ventilator, dynamic images 243c taken this time, graphs 243d and 243e showing the time-dependent changes in signal values ​​of the left and right lung fields, a table 243f showing the time-dependent changes in the cardiothoracic ratio, and a table 243g showing the time-dependent changes in lung field area.

[0065] Thus, the evaluation screen 243 displays, in a comparative manner, information regarding the presence or absence of lung complications, including the changes over time in images and characteristic quantities of predetermined structures related to lung complications (in this case, signal values ​​of lung fields, lung area, and cardiothoracic ratio) from before mechanical ventilation to the present. This allows the user to easily understand the presence or absence of lung complications, specifically pneumonia, atelectasis, pulmonary edema, or pleural effusion. Furthermore, if a complication is being treated, the user can easily understand whether or not the treatment is effective.

[0066] Figure 12 shows an example of an evaluation screen 244 displayed on the display unit 24 in step S26. Figure 12 shows an example of an evaluation screen 244 displaying information regarding the presence or absence of pneumothorax. As shown in Figure 12, the evaluation screen 244 displays, for example, patient information 244a of the patient being evaluated, a frequency-enhanced dynamic image 244b generated in step S23, a frame image 244c of its maximum inspiratory position, a frame image 244d of its maximum expiratory position, and a table 244e of the lung percentages of the right and left lungs at their respective maximum expiratory and maximum inspiratory positions.

[0067] Thus, the evaluation screen 244 displays frame images of the maximum expiratory position and maximum inspiratory position while the patient is on a ventilator, allowing for side-by-side comparison of information regarding the presence or absence of lung complications. It also displays the lung percentages for the right and left lungs at their respective maximum expiratory and inspiratory positions, making it easy for the user to identify the presence or absence of pneumothorax as a lung complication.

[0068] In the above-described process for evaluating pulmonary complications, we explained that dynamic imaging of the chest is performed and the resulting dynamic images are analyzed to generate information regarding the presence or absence of pulmonary complications. However, it is also possible to perform still imaging of the chest and analyze the resulting still images to generate information regarding the presence or absence of pulmonary complications. For example, still images may be taken at the maximum inspiratory position (when taking a deep breath) and the maximum expiratory position (when exhaling completely), and these still images may be analyzed.

[0069] Furthermore, in the above-mentioned pulmonary complication evaluation process, information regarding the presence or absence of complications is generated for all items related to lung complications, including pneumothorax, pneumonia, atelectasis, pulmonary edema, and pleural effusion. However, the system may be configured so that the user can select which items to generate information for from the control unit 23.

[0070] Furthermore, an alert may be output (display, audio output, etc.) if the difference between the features calculated before or during mechanical ventilation and the features calculated this time exceeds a predetermined threshold, or if the features calculated this time exceed a predetermined threshold.

[0071] Furthermore, information regarding the presence or absence of pneumothorax may be generated using images taken within a specified time after extubation from the ventilator.

[0072] Furthermore, in the above-described pulmonary complication evaluation process, as a preferred example, information showing the time-dependent changes between feature quantities calculated from images taken before and while the patient was on a ventilator and feature quantities calculated from the currently acquired image was used as information regarding the presence or absence of complications. However, information showing the time-dependent changes between feature quantities calculated from either images taken before or while the patient was on a ventilator and feature quantities calculated from the currently acquired image may also be used as information regarding the presence or absence of complications.

[0073] (Cardiac complication assessment process) The cardiac complication assessment process is a process for generating information regarding the presence or absence of cardiac complications. One cardiac complication is heart failure. Heart failure as a complication occurs while on mechanical ventilation. Therefore, in this embodiment, cardiac complication evaluation processing is performed while the ventilator is attached, and information regarding cardiac complications is generated.

[0074] Figure 13 is a flowchart showing the flow of the cardiac complication assessment process performed in console 2. The cardiac complication assessment process is performed through the cooperation of the control unit 21 and the program stored in the memory unit 22. The cardiac complication assessment process will be described below. In the explanation of Figure 13, we describe a case where dynamic imaging is performed to acquire dynamic images, and information regarding the presence or absence of cardiac complications is generated based on the acquired dynamic images. Furthermore, in the dynamic analysis system 100, it is assumed that dynamic images acquired by dynamic imaging of the subject H's chest before the attachment of the ventilator, and dynamic images and feature quantities acquired in a cardiac complication evaluation process previously performed while the subject was on a ventilator, are stored in the memory unit 22.

[0075] First, the control unit 21 receives input from the operation unit 23 of patient information (name, age, sex, disease, etc.) and examination information (area to be examined (here, for example, chest), type of imaging (here, dynamic imaging), time of imaging (here, while wearing the device), evaluation items (here, for example, cardiac complications)) (step S31).

[0076] Next, the control unit 21 controls the radiation irradiation control unit 12 and the FPD cassette 4 based on the input examination information to perform dynamic imaging of the subject H's chest in response to the pressing of the exposure switch 13, and acquires a dynamic image (step S32). Dynamic chest imaging may be used in conjunction with the assessment of pulmonary complications.

[0077] Next, the control unit 21 calculates characteristic features of predetermined structures related to cardiac complications from the acquired dynamic images (step S33).

[0078] The heart acts like a pump, sending blood throughout the body, but heart failure occurs when this function deteriorates and the body can no longer supply the necessary blood. In cases of heart failure, the lungs also do not receive enough blood, resulting in areas with impaired blood flow. Additionally, the cardiothoracic ratio increases in cases of heart failure.

[0079] Therefore, in step S33, for example, the control unit 21 performs blood flow analysis on the acquired dynamic image and calculates the blood flow feature quantities for each sub-region (each pixel or each set of pixels) of the lung field region as feature quantities of predetermined structures related to cardiac complications, and also calculates the cardiothoracic ratio as a feature quantity of predetermined structures related to cardiac complications. Alternatively, it may generate only one of these.

[0080] As a method for blood flow analysis, for example, the difference (absolute value of the difference) between the signal value of each sub-region of the lung field in each frame image of the dynamic image and the signal value of the corresponding sub-region in the reference frame image (the frame image with the highest signal value (i.e., the frame image with the least blood flow)) is calculated as a feature quantity indicating the blood flow rate for each sub-region of the lung field in each frame image. Alternatively, the difference between the signal value of each sub-region of the lung field in each frame image of the dynamic image and the corresponding sub-region in a frame image adjacent in the time direction may be calculated as a feature quantity indicating the blood flow rate for each sub-region in each frame image. If the dynamic image is taken while the patient is breathing, it is preferable to filter the time change of signal value for each corresponding sub-region between frame images with a time-direction high-pass filter (e.g., cutoff frequency of 0.7 Hz) before calculating the difference value.

[0081] Alternatively, as described in Japanese Patent Publication No. 2012-239796, for example, the cardiac region may be recognized from a dynamic image, a signal value waveform of the cardiac region may be generated as a heart rate signal waveform, and a signal value waveform may be generated for each sub-region of the lung field region of the dynamic image, and the cross-correlation coefficient with the heart rate signal waveform may be calculated by shifting the generated signal value waveforms by one frame interval (shifting them in the time direction), and the calculated cross-correlation coefficient may be used as a feature quantity related to blood flow for each frame image of each sub-region.

[0082] The control unit 21 associates the calculated feature quantities with patient information, examination information, and dynamic images, and stores them in the storage unit 22.

[0083] Next, the control unit 21 acquires characteristic features of predetermined structures related to cardiac complications from dynamic images taken before the ventilator was attached and dynamic images taken previously while the ventilator was attached (step S34). The control unit 21 reads dynamic images of the subject H taken before and during the use of a ventilator from the memory unit 22, and performs the same processing on the read dynamic images as described in step S33 to obtain characteristic quantities of predetermined structures related to cardiac complications in past dynamic images taken before and during the use of a ventilator. Furthermore, if characteristic quantities of predetermined structures related to cardiac complications have already been calculated from previously acquired dynamic images, the characteristic quantities are retrieved from the memory unit 22.

[0084] Next, the control unit 21 generates information regarding the presence or absence of lung complications (step S35). For example, in each dynamic image, the maximum value of the blood flow feature calculated for each frame image is combined into a single image, and an analysis result image is generated with a color corresponding to the maximum value. Information is then generated by arranging these analysis result images over time. Additionally, a table or graph showing the change in the cardiothoracic ratio over time is generated. It is also acceptable to generate only one of these.

[0085] Then, the control unit 21 displays an evaluation screen 245 on the display unit 24 that shows information regarding the presence or absence of complications related to the generated heart (step S36), and terminates the cardiac complication evaluation process.

[0086] Figure 14 shows an example of the evaluation screen 245 displayed on the display unit 24 in step S36. As shown in Figure 14, the evaluation screen 245 displays, for example, patient information 245a of the patient being evaluated, analysis result images 245b and 245c generated before and during mechanical ventilation (in the past), analysis result image 245d generated by the current imaging, and a table 245f comparing the feature quantities (cardiothoracic ratio) calculated from dynamic images acquired before and during mechanical ventilation (in the past) and the feature quantities (cardiothoracic ratio) calculated from the dynamic images taken this time.

[0087] As shown in Figure 14, the evaluation screen 245 displays information regarding the presence or absence of cardiac complications, including the time-series changes in image and characteristic quantities of predetermined structures related to lung complications (in this case, blood flow characteristics and cardiothoracic ratio) from before mechanical ventilation to the present, allowing users to easily understand the presence or absence of cardiac complications, such as heart failure. Furthermore, if complications are being treated, users can easily understand whether or not the treatment is effective. For example, in Figure 14, a blood flow defect (indicated as A in Figure 14) is seen in the analysis result image from 9 / 8, but no significant blood flow defect is seen in the current analysis result image, indicating improvement due to treatment.

[0088] In the cardiac complication evaluation process described above, we explained that dynamic imaging of the chest is performed and the resulting dynamic images are analyzed to generate information regarding the presence or absence of cardiac complications. However, the cardiothoracic ratio can also be generated by taking still images of the chest at its maximum expiratory position (when taking a deep breath) and analyzing the resulting still images.

[0089] Furthermore, an alert may be output (display, audio output, etc.) if the difference between the features calculated before or during the use of a ventilator and the features calculated this time exceeds a predetermined threshold, or if the features calculated this time exceed a predetermined threshold.

[0090] Furthermore, in the cardiac complication evaluation process described above, as a preferred example, information showing the time-dependent changes between feature quantities calculated from images taken before and while the patient was on a ventilator and feature quantities calculated from the image taken this time was used as information regarding the presence or absence of complications. However, information showing the time-dependent changes between feature quantities calculated from either images taken before or while the patient was on a ventilator and feature quantities calculated from the image taken this time may also be used as information regarding the presence or absence of complications.

[0091] Furthermore, while it is preferable to perform all of the above-mentioned airway complication assessment processes, pulmonary complication assessment processes, and cardiac complication assessment processes, it is also acceptable to perform only one or two of them.

[0092] Although embodiments of the present invention have been described above, the descriptions in the above embodiments are merely preferred examples of the dynamic analysis system according to the present invention and are not limited thereto.

[0093] For example, in the above embodiment, the dynamic analysis system was described as a system for ward rounds, but the present invention is also applicable to a dynamic analysis system that takes images in a shooting room and performs analysis on the obtained dynamic images.

[0094] Furthermore, while the above description discloses examples using hard disks, semiconductor non-volatile memory, etc., as computer-readable media for the program according to the present invention, the invention is not limited to these examples. Portable recording media such as CD-ROMs can also be used as other computer-readable media. In addition, carrier waves can be used as a medium for providing the data of the program according to the present invention via a communication line.

[0095] Furthermore, the detailed configuration and operation of each device constituting the dynamic analysis system can also be modified as appropriate, without departing from the spirit of the invention. [Explanation of symbols]

[0096] 100 Dynamic Analysis System 1. Radiation generating device 11 Radiation source 12 Radiation irradiation control unit 13 Exposure switch 2 Console 21 Control Unit 22 Memory section 23 Control section 24 Display 25 Communications Department 26 connectors 27 Bus 3 Access points 4 FPD Cassettes

Claims

1. An acquisition unit that acquires radiographic images including multiple frame images obtained by performing dynamic imaging on a subject while on a ventilator, A generation unit generates information regarding the presence or absence of complications related to the ventilator of the subject based on a predetermined frame image from among the radiographic images, which include the plurality of frame images acquired. Equipped with, The generation unit is an image analysis device that calculates information regarding the size of the left and right lung fields based on images obtained by performing dynamic imaging of the subject's chest while the ventilator is attached, and generates the calculated information as information regarding the presence or absence of complications related to the ventilator in the subject.

2. The aforementioned subject is a subject in the intensive care unit. The image analysis apparatus according to claim 1.

3. The aforementioned radiographic image was taken using a mobile medical unit. The image analysis apparatus according to claim 1 or 2.

4. The predetermined frame image includes at least one of an inspiratory frame image and an expiratory frame image. The image analysis apparatus according to any one of claims 1 to 3.

5. The image analysis device according to any one of claims 1 to 4, wherein the aforementioned complications include complications relating to the lungs.

6. The image analysis device according to claim 5, wherein the lung complications include at least one of pneumothorax, pneumonia, atelectasis, pulmonary edema, and pleural effusion.

7. The image analysis device according to claim 6, wherein the generation unit generates information as information regarding the presence or absence of pneumothorax in the subject, by applying frequency enhancement processing to at least images of the maximum expiratory position and images of the maximum inspiratory position obtained by performing dynamic imaging of the subject's chest while the ventilator is attached, and arranging them in a comparable manner.

8. The image analysis device according to claim 6 or 7, wherein the generation unit calculates at least one piece of information, such as the signal value of the lung field region, the cardiothoracic ratio, and the lung field area, from each of the images obtained by performing dynamic imaging on the chest of the subject while the ventilator is attached, and generates information showing the change over time between the calculated information and the information calculated based on images obtained before or previously while the ventilator was attached and dynamic imaging was performed on the chest of the subject, as information regarding the presence or absence of pneumonia, atelectasis, pulmonary edema, or pleural effusion in the subject.

9. Computers An acquisition unit that acquires radiographic images including multiple frame images obtained by performing dynamic imaging on a subject while on a ventilator. A generation unit generates information regarding the presence or absence of complications related to the ventilator of the subject based on a predetermined frame image from among the acquired multiple frame images of the radiographic image. To make it function as, The generation unit is a program that calculates information regarding the size of the left and right lung fields based on images obtained by performing dynamic imaging of the subject's chest while the ventilator is attached, and generates the calculated information as information regarding the presence or absence of complications related to the ventilator in the subject.

Images