WO2021060700A1

WO2021060700A1 - Apparatus and method for confirming videofluoroscopic swallowing study

Info

Publication number: WO2021060700A1
Application number: PCT/KR2020/010735
Authority: WO
Inventors: 장동진; 박성준
Original assignee: 가톨릭대학교 산학협력단
Priority date: 2019-09-24
Filing date: 2020-08-13
Publication date: 2021-04-01
Also published as: KR102094828B1

Abstract

An apparatus for confirming a videofluoroscopic swallowing study is disclosed. The apparatus for confirming a videofluoroscopic swallowing study comprises: an image input unit into which a plurality of videofluoroscopic images are input, the images being obtained by sequentially photographing motions in an anatomical position while an examinee is eating food; an anatomical position designating unit for designating the anatomical position in the plurality of videofluoroscopic images such that same corresponds to a predetermined anatomical standard position, and designating a food position with regard to the anatomical position; a coordinate setting unit for setting coordinates of the anatomical position in the plurality of videofluoroscopic images and the food position; a position tracking unit for tracking a degree of motion of the anatomical position and a movement path of food by using the coordinates of the anatomical position in the plurality of videofluoroscopic images and the food position; and a confirming unit for analyzing the degree of motion of the anatomical position and the movement path of food to confirm whether or not a swallowing disorder exists.

Description

Video-visual swallowing test reading device and method

The present invention relates to a video perspective swallowing test reading apparatus and method, and more particularly, to an artificial intelligence algorithm-based video perspective swallowing test reading device and method.

Swallowing (swallowing) is achieved by the harmonious movement of tissues in the oral cavity. In this regard, swallowing disorder (difficulty) refers to difficulty in swallowing caused by abnormalities in the muscular nervous system or structural abnormalities in the section from the oral cavity to the upper esophagus.

These swallowing disorders can occur at any age, from newborns to the elderly, and can occur as a result of various congenital anomalies, structural damage, or medical conditions. Specifically, there are various causes of swallowing disorders, in particular, there are mild causes of simple tooth abnormalities or implants, and paralysis of the muscles of the mouth, pharynx and esophagus due to nerve paralysis caused by other neurological diseases such as stroke (Stroker), There are narrowing of the pharynx or esophagus, compression of the esophagus due to malformations of the surrounding organs, convulsions due to spasm of the esophagus, and impulse to move food from the mouth to the esophagus. Symptoms include drooling or feeling that food has stopped in the esophagus. Etc. may appear.

In general, a videofluoroscopic swallowing study (VFS) is performed to determine whether there is a swallowing disorder. In particular, medical personnel with medical knowledge are reading whether airway aspiration or nasal inhalation by directly observing a plurality of VFS files composed of several frames.

However, the accuracy of the reading of the swallowing disorder by a medical practitioner is large depending on the experience or capability of the medical practitioner. In addition, there is a disadvantage in that it requires a separate time for reading.

An embodiment of the present invention is to provide a video perspective swallowing test reading apparatus and method capable of self-reading whether or not swallowing disorder from a video perspective image.

An embodiment of the present invention is to provide a video perspective swallowing test reading apparatus and method with high objectivity and accuracy.

According to an aspect of the present invention, there is provided an apparatus for reading a video fluoroscopy swallowing test, comprising: an image input unit for inputting a plurality of video fluoroscopy images sequentially photographing movements of an anatomical position while a test subject eats food; An anatomical position designating unit for designating the anatomical position to correspond to a preset anatomical standard position in the plurality of video perspective images and designating a position of food with respect to the anatomical position; A coordinate setting unit for setting coordinates of the anatomical position of the plurality of video perspective images and the position of the food; Using the anatomical position of the plurality of video perspective images and coordinates of the position of the food, a position tracking unit for tracking the moving degree of the anatomical position and the moving path of the food, and the degree of the anatomical position moving and the There is provided a video-visual swallowing test reading apparatus including a reading unit that analyzes a movement path of food and reads whether or not there is a swallowing disorder.

In this case, at least one of the image input unit, the position designating unit, the coordinate setting unit, the location tracking unit, and the reading unit may use an algorithm learned through machine learning.

In this case, the algorithm may perform the machine learning by constructing an artificial neural network.

At this time, the standard anatomical positions are tongue, hard palate, soft palate, mandibular angle, epiglottis, valleculae, laryngeal vestibule, It may be at least one of a fourth cervical vertebra (C4 vertebra), a pyriform sinus, an upper esophageal sphincter (UES), a vocal cord, an esophagus, or a trachea.

In this case, the position tracking unit may accumulate the anatomical position and the position of food included in an image divided from the video perspective image.

In this case, a result of reading whether the reading unit has a swallowing disorder may be any one of airway suction, nasal suction, or normal.

In this case, the reading unit may compare the degree of movement of the coordinates based on a predetermined clinical parameter to read whether the swallowing disorder has occurred.

In this case, the reading unit may compare the movement path of the food with a standard path model to read whether the swallowing disorder has occurred.

According to another aspect of the present invention, there is provided a method of reading a video-perspective swallowing test using a machine learning-based algorithm, the method comprising: inputting a plurality of video-perspective images to learn the algorithm; Inputting a video-perspective image in which the motion of the anatomical position is photographed while the test subject eats food; Designating the anatomical position to correspond to a preset anatomical standard position in the video perspective image and designating a position of food with respect to the anatomical position; Setting coordinates of the anatomical position and the position of the food; There is provided a method for reading a swallowing test by video perspective, including the step of tracking the degree of movement of the anatomical position and the position of food, and reading whether or not there is a swallowing disorder by analyzing the degree of movement of the anatomical position and the movement path of the food. do.

According to an embodiment of the present invention, it is possible to read whether a swallowing disorder has occurred from a video perspective image through an algorithm learned through machine learning, so that it is possible to save cost and time due to image reading for medical personnel.

According to an embodiment of the present invention, it is possible to read whether a swallowing disorder has occurred from a video perspective image through an algorithm learned through machine learning, thereby improving objectivity and accuracy of a video perspective swallowing test.

1 is a block diagram of a video perspective swallowing test reading apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a standard anatomical position designated by a coordinate setting unit of the apparatus for reading video fluoroscopy swallowing tests according to an exemplary embodiment of the present invention.

FIG. 3 is a diagram illustrating an example of accumulating a position of food in a segmented image in one image by the video perspective swallowing test reading apparatus according to an exemplary embodiment of the present invention.

4 is a diagram showing a VFS file when a video-perspective swallowing test reading apparatus normally reads according to an embodiment of the present invention.

FIG. 5 is a diagram showing a VFS file when a video-visual swallowing test reading apparatus according to an embodiment of the present invention performs airway suction reading.

FIG. 6 is a diagram illustrating a VFS file when a video-visual swallowing test reading apparatus according to an embodiment of the present invention performs nasal suction reading.

FIG. 7 is a diagram showing a standard path model prepared for reading whether a swallowing disorder is detected by a video-visual swallowing test reading apparatus according to an embodiment of the present invention.

8 is a flowchart of a method of performing a video-perspective swallowing test by a video-perspective swallowing test reading apparatus according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art may easily implement the present invention. The present invention may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and the same reference numerals are assigned to the same or similar components throughout the specification.

Some embodiments of the present disclosure may be represented by functional block configurations and various processing steps. Some or all of these functional blocks may be implemented with various numbers of hardware and/or software components that perform specific functions. For example, the functional blocks of the present disclosure may be implemented by one or more microprocessors, or may be implemented by circuit configurations for a predetermined function. In addition, for example, the functional blocks of the present disclosure may be implemented in various programming or scripting languages. Functional blocks may be implemented as an algorithm executed on one or more processors. In addition, the present disclosure may employ conventional techniques for electronic environment setting, signal processing, and/or data processing. Terms such as “mechanism”, “element”, “means” and “composition” can be used widely, and are not limited to mechanical and physical configurations.

The video-perspective swallowing test reading apparatus 1 according to an embodiment of the present invention is a device that analyzes a video-perspective image and reads whether swallowing disorders exist. In particular, by using an algorithm learned through machine learning, a video-perspective without reading by a medical person. It relates to a device for performing swallowing test.

1 is a block diagram of a video perspective swallowing test reading apparatus according to an embodiment of the present invention. FIG. 2 is a diagram showing a standard anatomical position designated by a coordinate setting unit of the apparatus for reading video fluoroscopy swallowing tests according to an exemplary embodiment of the present invention. FIG. 3 is a diagram illustrating an example of accumulating a position of food in a segmented image in one image by the video perspective swallowing test reading apparatus according to an exemplary embodiment of the present invention. 4 to 6 are diagrams each showing a VFS file when a video fluoroscopic swallowing test reading apparatus according to an embodiment of the present invention performs normal, airway suction, and nasal suction readings. FIG. 7 is a diagram showing a standard path model prepared for reading whether a swallowing disorder is detected by a video-visual swallowing test reading apparatus according to an embodiment of the present invention.

Referring to FIG. 1, a video perspective swallowing test reading device (1, hereinafter swallowing test device) according to an embodiment of the present invention includes an image input unit 110, an anatomical position designation unit 120, and a coordinate setting unit 130. , And a location tracking unit 140 and a reading unit 150.

In an embodiment of the present invention, the image input unit 110 receives a video perspective image (V) photographed while a test subject eats food. The video perspective image V is an image capable of confirming the movement of the body tissue of the test subject, and in this case, the body tissue may be at least one of an oral cavity, a pharynx, an esophagus, a larynx, or an organ. In addition, the video perspective image V may include not only an X-ray image using X-rays but also a magnetic resonance image (MRI) image.

At this time, when taking a video perspective image (V), the test subject may consume food including a contrast agent. This is to more clearly indicate the position of the food in the video perspective image (V).

The anatomical position designating unit 120 of the swallowing test apparatus 1 according to an embodiment of the present invention corresponds to the standard anatomical position St in the video perspective image V input through the image input unit 110. Anatomical location (Sp) can be specified. Here, designation may mean mapping.

At this time, referring to FIG. 2, the standard anatomical position (St) is a major body tissue that is considered important when reading whether the reading unit 150 to be described later has a swallowing disorder, and the reference when designating an anatomical position (Sp) to be described later is It means the virtual anatomical location (Sp). For example, as shown in Figure 2, the standard anatomical position (St) is tongue, hard palate, soft palate, mandibular angle, epiglottis, epiglottis (valleculae), laryngeal vestibule, fourth cervical vertebra (C4 vertebra), pyriform sinus, upper esophageal sphincter (UES), vocal cord, esophagus, or trachea Can be However, these are only examples of the standard anatomical position (St), and the standard anatomical position (St) of the swallowing examination apparatus 1 according to an embodiment of the present invention is not limited thereto.

In addition, the anatomical position (Sp) is a body tissue designated to correspond to the standard anatomical position (St) in a specific video perspective image (V), such as 10, 20, 30, 40 or 50 shown in Figs. This is an example of the standard anatomical position (St).

At this time, the designated anatomical position (Sp) is the most easily observed point among the body tissue regions corresponding to the standard anatomical position (St), the point at the center of the body tissue, and a point that can represent the body tissue. Alternatively, it can be designated as a single point, such as the point that reacts most sensitively.

The designated anatomical location Sp may be used as a spatial index indicating where food is located among body tissues when tracking the food movement path of the location tracking unit 140 to be described later.

In one embodiment of the present invention, the positioning unit 120 designates the position of food (Sf) together with the anatomical position (Sp). Unlike the anatomical position (Sp), the position (Sf) of ingested food moves along the body organs. Therefore, positions of foods in a plurality of divided images divided in chronological order may be different from each other.

As illustrated in FIG. 3, when different positions of food are accumulated in one image for each divided image, the movement path of the food can be identified. As already described above, a contrast medium is used to more clearly designate the position (Sf) of the food.

In an embodiment of the present invention, the location designation unit 120 may designate an anatomical location Sp and a food location Sf using an algorithm learned through machine learning. In other words, it is possible to automatically designate an anatomical position (Sp) on an arbitrary video perspective image using an algorithm without the aid of a medical professional. At this time, machine learning refers to an algorithm that repeatedly learns from specific sample data, finds potential features in specific sample data, and applies the learning results to new data to predict the future according to the discovered features. do.

In an embodiment of the present invention, machine learning may be based on a learning model using one or more artificial neural networks. In this case, the artificial neural network is a kind of algorithm modeled after the human neural network structure, and may include one or more nodes or one or more layers including neurons, and each node may be connected through a synapse.

In this case, the artificial neural network may be based on supervised learning in which data of a video perspective image including an anatomical location Sp designated by a medical professional with expertise is used as an input value. In addition, the artificial neural network is a method of designating an anatomical location (Sp) or food location (Sf) by discovering a feature or pattern from a plurality of video perspective images without any guidance based on unsupervised learning. You can also learn.

In an embodiment of the present invention, the coordinate setting unit 130 sets relative coordinates based on a specific point with respect to the designated anatomical location Sp and the food location Sf. This is to enable more diverse and precise data processing by quantifying the anatomical location (Sp) and the food location (Sf). For example, when analyzing the movement path of food by the location tracking unit 140 to be described later, it is possible to achieve more accurate movement path accumulation in consideration of relative coordinates.

In an embodiment of the present invention, the location tracking unit 140 may track an anatomical location Sp and a location Sf of food.

In an embodiment of the present invention, the video perspective image V may be divided into a plurality of frames. The swallowing test apparatus 1 according to an embodiment of the present invention may use only some frames among a plurality of frames for analysis. This is to save energy and time consumed in swallowing test by reducing data throughput. The location tracking unit 140 checks and analyzes according to the passage of time what spatial movement has occurred in relation to the anatomical location (Sp) and the food location (Sf) between some of the selected frames as described above. .

First, tracking the anatomical location Sp means tracking the movement of the body tissue represented by the anatomical location Sp. To this end, the location tracking unit 140 may track movement of coordinates related to the anatomical location Sp. The degree of movement of such body tissues is called a clinical parameter. The clinical parameter may be used when the reading unit 150 to be described later determines whether or not there is a swallowing disorder.

In addition, tracking of the anatomical location (Sp) may be helpful in the treatment of swallowing disorders. For example, by analyzing the degree of movement of the anatomical position Sp, it may be helpful to determine which body tissue among the body tissues causes swallowing disorder.

Next, tracking the position Sf of the food means to check the relative positional movement of the food, and it means tracking which body tissue the food is located adjacent to over time. Referring back to FIG. 3, the location tracking unit 140 individually checks the positions of foods included in some frames selected for analysis among a plurality of frames, and then comprehensively accumulates them, thereby providing an overall food movement path. have.

At this time, the movement of food between some frames and frames may be estimated by an algorithm. By filling the gaps between some of the frames through such an estimation operation, as shown in FIGS. 4 to 6, it is possible to represent a continuous moving path of food as a whole. In this case, the estimation operation may be performed by an algorithm learned through machine learning. It goes without saying that the algorithm can be designed by constructing an artificial neural network for more sophisticated estimation.

In an embodiment of the present invention, the reading unit 150 may determine whether or not there is a swallowing disorder based on the degree of movement of the anatomical position Sp obtained through the position tracking unit 140 and the movement path of the food. have. However, the anatomical location (Sp) and the movement path of food are only an example of the basis of the reading of the reading unit 150, and other information other than the anatomical location (Sp) and the movement path of food can be included to determine whether or not there is a swallowing disorder. Of course you can.

The reading unit 150 may determine whether or not there is a swallowing disorder by comparing the clinical parameter value obtained from the location tracking unit 140 with the reference value of the clinical parameter in relation to the movement of the anatomical position Sp. In addition, the clinical parameter values of each body tissue may be compared with each other to determine whether swallowing disorders exist.

In addition, the reading unit 150 may determine whether or not there is a swallowing disorder by comparing the overall food movement path obtained through the location tracking unit 140 with the standard path model of FIG. 7 in relation to the movement path of the food. 4 to 6, as shown in FIGS. 4 and 7(a), when it is confirmed that the food has moved in the order of the oral cavity 10-the pharynx 20-the esophagus 30, the reading unit 150 is normal. Can be read as. However, as shown in FIGS. 5 and 7(b), when it is confirmed that food has been moved to the larynx 40-tracheal 50 in addition to the esophagus 30, the reading unit 150 can read it as'airway suction'. have. In addition, as shown in FIGS. 6 and 7 (c), when a path through which food flows into the nasal cavity 60 is confirmed, it may be read as'nasal reflux'.

In an embodiment of the present invention, the reading of whether the reading unit 150 has a swallowing disorder may also be performed through an algorithm learned through machine learning. In more detail, the algorithm used when reading whether or not the reading unit 150 has a swallowing disorder is a swallowing disorder by comprehensively judging a variety of judgment grounds, including the above-described anatomical location (Sp), clinical parameters, and food movement paths. Whether or not can be read.

In this case, the algorithm related to the reading unit 150 may be machine learning in advance through a plurality of learning data, and may be self-updated through such a learning process. Such learning can be applied to both supervised and unsupervised learning, and of course, it can be achieved by constructing an artificial neural network.

As described above, the algorithms applied to the location designation unit 120, the location tracking unit 140, and the reading unit 150 may all be artificial intelligence algorithms based on machine learning. In this case, a plurality of algorithms may exist individually to perform the respective tasks of the position designation unit 120, the position tracking unit 140, and the reading unit 150, or the position designation unit 120 and the position tracking unit 140 ), there may be an integrated algorithm encompassing the reading unit 150.

In the swallowing test apparatus 1 according to an embodiment of the present invention, since the video-perspective swallowing test is automatically performed by an artificial intelligence algorithm as described above, it is possible to reduce labor of medical personnel and secure objectivity of reading.

Hereinafter, a method of performing a video-perspective swallowing test using the swallowing test apparatus 1 according to an embodiment of the present invention (hereinafter, a method of performing a swallowing test) will be described with different drawings.

8 is a flow chart of a method of performing a video perspective swallowing test using the swallowing test apparatus 1 according to an embodiment of the present invention.

Referring to FIG. 8, a method of performing a swallowing test according to an embodiment of the present invention includes the steps of learning an algorithm through a plurality of images (S10); Inputting a video-perspective image in which movements of body tissues are photographed during food intake (S20); Designating an anatomical position (Sp) and a position of food corresponding to the standard anatomical position (St) in the video perspective image (S30); Step of setting the anatomical position (Sp) and the coordinates of the food (S40); Including the step of tracking the degree of movement of the anatomical position (Sp) and the position of food (S50), and the step of reading whether or not there is a swallowing disorder by analyzing the degree of movement of the anatomical position (Sp) and the movement path of the food (S60). can do.

In the step of learning an algorithm through a plurality of images (S10), a plurality of video perspective images V photographed for swallowing test are provided to the algorithm as learning data. In this case, the algorithm may be learned by constructing an artificial neural network, and may be updated by itself by reflecting the results experienced during the learning process. In addition, the algorithm can perform a task by extracting features from a plurality of video perspective images V by itself.

In the step (S20) of inputting a video-perspective image (V) in which the movements of the body tissues are captured while eating food (S20), a video-perspective image (V) is input in real time from a separate video-perspective image capturing device, or a video-perspective image that has already been taken Can be entered.

In the step of designating the anatomical position (Sp) corresponding to the standard anatomical position (St) in the video perspective image (V) (S30), the anatomy selected in advance from the video perspective image that can be different from each other according to the individual's body structure Find and match the position corresponding to the enemy standard position (St). As described above, such a designation task can be performed through an algorithm learned through machine learning, thereby securing speed and objectivity.

In the step (S40) of setting the coordinates of the anatomical position (Sp), the relative coordinates of the designated anatomical position (Sp) are set through the step (S30) of designating the anatomical position (Sp) based on a specific point. The coordinates set in this way may be used when tracking the anatomical location (Sp) and the location of food (Sf) of the location tracking unit 140.

In the step (S40) of tracking the degree of movement of the anatomical position (Sp) and the position of the food (S40), the movement of the anatomical position (Sp) and the movement path of the food according to the passage of time between some of the divided frames of the video perspective image are determined. To track. In particular, by tracking the movement of the anatomical position Sp, clinical parameters to be used when analyzing the reading unit 150 to be described later can be derived.

In the step (S60) of analyzing the degree of movement of the anatomical position (Sp) and the movement path of food to read whether there is a swallowing disorder, the artificial intelligence algorithm learned through machine learning analyzes the video perspective image to read whether the swallowing disorder is can do. At this time, the reading unit 150 compares the clinical parameter obtained from the location tracking unit 140 with the clinical parameter default value or compares the entire movement route of the obtained food with the standard route model of FIG. 7 to determine whether or not there is a swallowing disorder. Can be read. However, in an embodiment of the present invention, the method of determining whether the reader 150 has a swallowing disorder is not limited thereto, and other data values extracted from a video perspective image may be considered together.

Although an embodiment of the present invention has been described above, the spirit of the present invention is not limited to the embodiment presented in the present specification, and those skilled in the art who understand the spirit of the present invention can add components within the scope of the same idea. It will be possible to easily propose other embodiments by changing, deleting, adding, etc., but it will be said that this is also within the scope of the present invention.

Claims

As a video-visual swallowing test reading device,

An image input unit for inputting a plurality of video perspective images sequentially photographing movements of an anatomical position while the test subject eats food;

An anatomical position designating unit for designating the anatomical position to correspond to a preset anatomical standard position in the plurality of video perspective images and designating a position of food with respect to the anatomical position;

A coordinate setting unit for setting coordinates of the anatomical position of the plurality of video perspective images and the position of the food;

A position tracking unit for tracking the degree of movement of the anatomical position and the movement path of the food using the coordinates of the anatomical position of the plurality of video perspective images and the position of the food, and

Comprising a reading unit for reading whether a swallowing disorder by analyzing the degree of movement of the anatomical position and the movement path of the food, video perspective swallowing test reading device.
The method of claim 1,

At least one of the image input unit, the position designation unit, the coordinate setting unit, the position tracking unit, and the reading unit uses an algorithm that is learned through machine learning.
The method of claim 2,

The algorithm constructs an artificial neural network to perform the machine learning, a video-perspective swallowing test reading device.
The method of claim 1,

The standard anatomical positions are tongue, hard palate, soft palate, mandibular angle, epiglottis, valleculae, laryngeal vestibule, and fourth. Cervical vertebra (C4 vertebra), pyriform sinus, upper esophageal sphincter (UES), vocal cord, esophagus (esophagus), or at least one of trachea (trachea), video fluoroscopic swallowing test reading device.
The method of claim 1,

The position tracking unit accumulates the anatomical position and the position of food included in an image divided from the video perspective image, a video perspective swallowing test reading device.
The method of claim 1,

The reading result of the swallowing disorder of the reading unit is any one of airway suction, nasal suction, or normal, video fluoroscopic swallowing test reading device.
The method of claim 1,

The reading unit reads whether or not the swallowing disorder by comparing the degree of movement of the coordinates based on a predetermined clinical parameter, video perspective swallowing test reading device.
The method of claim 1,

The reading unit compares the movement path of the food with a standard path model and reads whether the swallowing disorder has occurred.
As a method of reading a video perspective swallowing test using a machine learning-based algorithm,

Learning the algorithm by inputting a plurality of video perspective images;

Inputting a video-perspective image in which a motion of an anatomical position is photographed while a test subject eats food;

Designating the anatomical position to correspond to a preset anatomical standard position in the video perspective image and designating a position of food with respect to the anatomical position;

Setting coordinates of the anatomical position and the position of the food;

Tracking the degree of movement of the anatomical position and the position of the food, and

Comprising the step of reading whether or not swallowing disorder by analyzing the degree of movement of the anatomical position and the movement path of the food, video perspective swallowing test reading method.