KR20200058657A - Vfss based auromatic dysphagia detection system and the method therof - Google Patents

Abstract

According to the present invention, disclosed are a videofluoroscopic swallowing study-based (VFSS-based) automatic dysphagia detection system and a detection method thereof. According to an embodiment of the present invention, the VFSS-based automatic dysphagia detection system comprises: an image extraction unit; an area detection unit; a food detection unit; a calculation unit; and a determination unit. According to the present invention, a ratio of a residual food area in vallecula epiglottica and piriform sinus areas is measured, and thus severity of dysphagia can be determined through the measured ratio.

Description

VFSS-based automatic dysphagia detection system and its detection method {VFSS BASED AUROMATIC DYSPHAGIA DETECTION SYSTEM AND THE METHOD THEROF}

The present invention relates to an automatic detection system for dysphagia based on VFSS and a detection method thereof, and more specifically, to detect a dysphagia based on the detection of the ratio of epiglottis and abnormal food areas contained in the region and laryngeal dysphagia. It relates to a system and its detection method.

Swallowing refers to a series of functions that deliver food from the oral cavity to the gastrointestinal tract, which involves many head and neck muscles, including the sphincter muscles dominated by the nerves of the central and peripheral nervous system, as well as skeletal and other soft tissues. When an abnormality occurs in any part of the swallowing pathway, it is called dysphagia or swallowing disorder. The main causes include muscle and neurological abnormalities related to dietary intake, and other psychiatric diseases or drugs. The back may also cause dysphagia. In particular, 100,000 stroke patients occur annually in Korea, of which 51 to 73 (%) suffer from dysphagia, which is difficult to eat.

Difficulty dysphagia shows various symptoms such as delay in swallowing reflexes, retention of pharyngeal food, and food inflow into the larynx, as well as weakening of the control of the lips and tongue. In addition, dysphagia frequently occurs not only in stroke patients, but also in patients with central nervous system damage such as head / cervical injury due to traffic accidents, brain tumors, and the like.

In order to diagnose such dysphagia patients, conventionally, a method of determining the severity of dysphagia was used by a medical staff directly checking the area of food left in the epiglottis (Vallecular) and abnormal sinus (Pyriform sinus) on a radiographic image. .

This method has a problem in that it is impossible to accurately diagnose the ratio of food remaining in the epiglottis and the epiglottis.

In order to solve the problems of this method, a study on a system capable of improving the efficiency and accuracy of diagnosis by dimensioning the proportion of food left in the epiglottis and the epiglottis is essential.

KR Registered Patent Publication No. 10-1793816 (published Sep. 6, 2017)

The present invention provides a VFSS based swallowing dysfunction automatic detection system and a method for detecting the severity of dysphagia through the measured ratio by measuring the ratio of the epiglottis and the areas of abnormal food and residual food in the region.

The present invention provides a VFSS-based swallowing dysfunction automatic detection system and a detection method capable of detecting accurate epiglottis and abnormalities and regions by performing deep learning training on the shape and position of the epiglottis and abnormalities.

The present invention provides a VFSS-based automatic dysphagia automatic detection system and its detection method that can more clearly check the location of the hyoid bone, epiglottis, abnormal and residual food areas by performing contrast correction on the extracted VFSS image.

VFSS-based swallowing dysfunction automatic detection system according to an embodiment of the present invention, an image extraction unit for extracting the image located at the bottom of the hyoid bone among a plurality of Videofluoroscopic Swallowing Study (VFSS) images, the image extraction Region detection unit to detect epiglottis and abnormalities and regions from images extracted from the part, food detection unit to detect residual food areas within the epiglottis and abnormalities and regions detected by the region detection unit, epiglottis and abnormalities and regions detected from the region detection unit The calculation unit calculates the area of the area, calculates the area of the residual food area detected by the food detection unit, and calculates the ratio of the area occupied by the residual food area among the areas in the epiglottis and abnormal areas and the calculation area and the calculation unit It may include a judgment unit for determining the severity of dysphagia in contrast to the proportion of the remaining food and the predetermined value.

According to an aspect of the present invention, the contrast correction to correct the contrast so that the numerical value of the contrast of the hyoid area, the epiglottis area, the abnormal area and the residual food area has a predetermined difference from the average contrast value of the VFSS image. Further, the contrast correction unit may have a difference in contrast by emphasizing whiteness of the epiglottis and abnormal areas and regions through gamma correction, and emphasizing black of a bone region of a subject in the VFSS image.

According to an aspect of the present invention, the food detection unit, when the food containing barium solution displayed on the VFSS image taken by the subject remains in the epiglottis and abnormalities and areas, the image extracted from the image extraction unit It can be detected as the remaining food area.

According to an aspect of the present invention, when the ratio of the residual food area occupies less than the first value, it is determined as non- swallowing disorder, and when the ratio of the remaining food area occupies less than the first value and less than the second value, the If it is determined as a disorder, and the proportion of the remaining food area occupies more than the second value, it can be determined as a severe dysphagia.

According to an embodiment of the present invention, a method for automatically detecting dysphagia based on VFSS extracts an image in which the location of the hyoid bone is located at the bottom of a plurality of VFSS images, and a boundary is displayed in the hyoid region, the epiglottis in the extracted image Detecting a grain and an abnormality and an area, and displaying the boundary between the detected epiglottis and an abnormality and an area, and detecting a residual food area in the image in which the boundary between the epiglottis and the abnormality and the area is displayed, and detecting the boundary of the detected residual food area. Is displayed, calculating the area of the epiglottis area, the abnormal area and the residual food area, calculating the proportion of the residual food area in the epiglottis area and the abnormal area and the proportion of the residual food area. And determining the severity of dysphagia in contrast to the set value.

According to an embodiment of the present invention, by measuring the ratio of the epiglottis and abnormal areas and food areas remaining in the region, the VFSS-based swallowing dysfunction automatic detection system and its detection method capable of determining the severity of dysphagia through the measured ratio Is provided.

According to an embodiment of the present invention, by performing deep learning training on the shape and location of the epiglottis and abnormalities, a VFSS-based swallowing dysfunction automatic detection system and a method for detecting the correct epiglottis and abnormalities and regions are provided. do.

According to an embodiment of the present invention, by performing a contrast correction operation on the extracted VFSS image, the VFSS-based swallowing dysfunction automatic detection system that can more clearly check the location of the hyoid bone, epiglottis, abnormal orbital and residual food areas, and A detection method is provided.

1 is a block diagram showing the configuration of a VFSS-based swallowing disorder automatic detection system according to an embodiment of the present invention.
2 is a view showing an example in which the ratio of the remaining food area according to an embodiment of the present invention is calculated.
3 is a view showing an example of contrast correction through gamma correction according to an embodiment of the present invention.
4 is a flow chart showing the flow of the VFSS-based swallowing dysfunction automatic detection method according to an embodiment of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the contents described in the accompanying drawings. However, the present invention is not limited or limited by the embodiments. The same reference numerals in each drawing denote the same members.

Conventionally, an image processing program used to determine whether dysphagia is difficult has a problem in that an accurate diagnosis is impossible because a doctor visually checks an image taken by a subject and manually draws and diagnoses it.

In addition, there was also a problem that the diagnosis result differed according to differences in opinions among doctors.

The present invention is an invention designed to solve the problems of the prior art, and the configuration of the present invention will be described in detail below.

1 is a block diagram showing the configuration of a VFSS-based swallowing difficulty automatic detection system 100 according to an embodiment of the present invention.

Referring to FIG. 1, the VFSS-based automatic dysphagia automatic detection system 100 of the present invention includes an image extraction unit 110, an area detection unit 120, a food detection unit 130, a calculation unit 140, and a determination unit 150. It may include.

The image extraction unit 110 may extract an image in which the location of the hyoid bone is located at the bottom of a plurality of VFSS images.

That is, the hyoid bone 210 is included in every VFSS image and photographed, and the location of the hyoid bone 210 for each image may be expressed by a Y coordinate value. Therefore, through comparison between multiple VFSS images, an image having the smallest Y-coordinate value of the hyoid bone 210 is extracted and used as an image for detecting the epiglottis 220 and anomalous 230 region, which will be described later. have.

This is because, when the hyoid bone 210 is positioned at the bottom, the epiglottis 220 and the abnormal orbit 230 are maximally expanded, so it is most suitable to observe the epiglottis 220 and the abnormal orbit 230 region. .

The region detection unit 120 may detect a region of the epiglottis 220 and the abnormality 230 from the image extracted from the image extraction unit 110. Here, when the detected areas of the epiglottis 220 and the abnormality 230 are detected, the outer boundary may be separately displayed in a distinguishable color.

At this time, the color of the boundary line of the outer boundary of the epiglottis 220 region and the abnormality 230 region may be displayed in different colors.

The food detection unit 130 detects an area of the residual food 240 in the epiglottis 220 and the abnormal 230 region detected by the region detection unit 120, and an outer boundary of the detected food residue 240 region Can be displayed.

The method of detecting the area of the residual food 240 is when the subject ingests food (ex. Water, pudding, mime, etc.) containing barium solution that can be displayed on the VFSS image, the epiglottis 220 and The food containing barium solution remaining in the abnormal orbital region 230 may be displayed on the VFSS image to detect the remaining food 240 region.

In other words, if there is dysphagia in general food intake, residual food 240 will be present in the epiglottis 220 and the abnormal gyrus 230, so the food containing the barium solution is consumed and the food containing the barium solution is the epiglottis It is possible to predict in advance that there will be residual food 240 in the future through what remains in the song 220 and the Sangwawa 230.

At this time, the outer boundary of the epiglottis 220 region and the outer boundary of the residual food 240a region within the epiglottis 220 region may be displayed in the same color, and the boundary between the abnormality and the 230 region and the abnormality The outer border of the area of the residual food 240b in the area may be displayed with the same color.

In addition, the outer boundary may be displayed as a boundary line, the entire area may be painted and displayed, and the region of interest and the non-interested region may be corrected and displayed to have a difference in contrast.

The method of correcting the region of interest and the non-interested region to have a difference in contrast will be described in detail in a contrast correction unit (not shown), which will be described later.

In one embodiment, when an outer boundary in the region of the epiglottis 220 is indicated by a boundary line, the region of the epiglottis 220 is detected by the region detection unit 120, and the boundary of the outer region of the epiglottis 220 is detected. Borders can be displayed in green.

The food detection unit 130 detects the area of the residual food 240a in the region of the epiglottis 220 detected by the area detection unit 120, and a border line may be displayed in green on the outer boundary of the area of the residual food 240a. have.

In another embodiment, when the boundary of the epiglottis 220 region is to be painted and expressed in the entire region, the region detection unit 120 detects the epiglottis 220 region, and the detected epiglottis 220 ) The entire area can be painted in green to express the border.

The food detection unit 130 detects the area of the residual food 240a within the region of the epiglottis 220 detected by the area detection unit 120, and paints the entire area of the residual food 240a in green to express a boundary. Can be.

At this time, the color of the color painted in the region of the epiglottis 220 and the area of the residual food 240a is the same, but the contrast is different to allow the user to distinguish the epiglottis 220 and the residual food 240a. have.

As in the above-described embodiments, a doctor can easily diagnose the patient by marking a boundary line on the outer boundary of the epiglottis 220 and the residual food 240a, or by painting the entire region of interest. The boundary-marked image or the entire region of interest is painted, and the severity of dysphagia can be easily explained.

In addition, a method of displaying a boundary may be equally applied to the abnormality 230 region and the residual food 240b region in the abnormality 230 region.

Meanwhile, deep learning technology is applied to the image extraction unit 110, the area detection unit 120, and the food detection unit 130, the hyoid bone 210, the epiglottis 220, the abnormal vortex 230, and the remaining food 240 The detection accuracy of the region can be improved.

The calculation unit 140 calculates the area of the epiglottis 220 and the abnormal orbital region 230 detected from the region detection unit 120, and the area of the residual food 240 region detected by the food detection unit 130. The ratio of the area occupied by the area of the residual food 240 among the areas in the epiglottis 220 and the abnormal area 230 may be calculated.

In addition, the calculator 140 calculates the area of the residual food 240a region in the epiglottis 220 region and the residual food 240b region in the abnormal 230 region, respectively, as a numerical value on the VFSS image. Can be displayed.

That is, the calculation unit 140 calculates the area of the epiglottis 220 region, the abnormal gyrus 230 region, and the remaining food 240 region, and the calculated ratio according to the area is displayed on the image. By making the diagnosis with the correct dimensions, the accuracy of the diagnosis can be improved.

The determination unit 150 may determine the severity of dysphagia by comparing the proportion of the remaining food 240 calculated by the calculation unit 140 with a preset value.

A criterion for determining the severity of dysphagia in comparison with a predetermined value in the determination unit 150 may be determined as a non- swallowing disorder when the proportion occupied by the area of the residual food 240 is less than a first value. (240) When the proportion of the area occupied by the first value or more and less than the second value may be determined as mild dysphagia, and when the proportion of the remaining food 240 area occupies more than the second value, it may be determined as severe dysphagia. .

The criterion for determining the proportion occupied by the residual food 240 area is compared with the predetermined area by comparing the ratio of the area occupied by the residual food 240 area to the total area of the epiglottis 220 and the abnormal 230 area. The severity of the disorder can be determined.

In addition, the ratio occupied by the residual food 240a region in the epiglottis 220 region and the ratio occupied by the residual food 240b region in the abnormal 230 region are more occupied by the residual food 240a, 240b. The severity of dysphagia can be determined by contrasting the larger proportions with the predetermined values.

For example, if the proportion of the remaining food 240 area to the total area of the epiglottis 220 and the abnormal area 230 area is 25%, the residual food relative to the total area of the epiglottis 220 and the abnormal area 230 area ( 240) It can be compared with the ratio occupied by the area and the predetermined value (ex. 20% or more and less than 30%, it is considered to have mild dysphagia).

In this case, since the predetermined value was exceeded, it can be determined that there is a mild dysphagia.

As another example, if the proportion of the remaining food 240a in the epiglottis 220 region is 40% and the proportion of the remaining food 240b in the abnormal 230 region is 5%, the epiglottis 22 ) Since the proportion (40%) of the area of the residual food (240a) in the area is greater than the proportion (5%) of the area of the residual food (240b) in the area (230), the residual food in the epiglottis (220) area (240a) The ratio of the area, 40%, can be compared with a predetermined value (ex. If it is more than 40%, it is determined that there is severe dysphagia).

In this case, since the predetermined value was exceeded, it can be determined that there is severe dysphagia.

An embodiment of the VFSS-based dysphagia automatic detection system 100 will be described in detail with reference to FIG. 2 below.

2 is a view showing an example in which the ratio of the remaining food area according to an embodiment of the present invention is calculated.

Referring to Figure 2, the image extraction unit 110, the area detection unit 120, the hyoid bone 210, the epiglottis 220, the abnormal orbital 230 and the residual food 240 area detected by the food detection unit 130 It can be displayed by painting on the VFSS image.

The hyoid bone 210 area is painted and displayed in red, and the residual food 240a area in the epiglottis 220 and the epiglottis 220 area is painted and displayed in green, and the abnormalities 230 and the The area of the residual food 240b in the area of the abnormal phase 230 may be displayed by painting in blue.

At this time, the region of the epiglottis 220 and the area of the remaining food 240a in the region of the epiglottis 220 are painted and displayed in the same color, but can be distinguished by different contrasts, and the region of the abnormality 230 Although the area of the food 240b remaining in the abnormal area 230 is also painted and displayed in the same color, the contrast may be distinguished by different contrasts.

The ratio of the remaining food (240a) area in the epiglottis 220 region and the proportion of the remaining food (240b) region in the abnormal (230) region are analyzed by the calculator 140 and displayed on the VFSS image. Can be.

As a result of analysis, the proportion of the remaining food 240a area in the region of the epiglottis 220 is calculated as 16.12% and displayed on the VFSS image, and the proportion of the remaining food 240b area in the abnormal 230 region is calculated. It can be calculated as 1.31% and displayed on the VFSS image.

The severity of the swallowing disorder is determined by the determination unit 150 by comparing the calculated ratio with a predetermined value.

At this time, as illustrated above, by extracting the values of the areas of the epiglottis 220 and the abnormal portion 230 where the remaining food area is larger, it is possible to determine whether swallowing disorders by comparing the extracted values with preset values. .

That is, the area ratio occupied by the residual food 240a in the epiglottis 220 region is 16.12%, and the area ratio occupied by the residual food 240b in the abnormal 230 region is 1.31%, whichever is greater. It can be judged by extracting 16.12% having the area ratio as the comparison area ratio and comparing it with a preset value (ex. The first value of the preset value is set to 5%, and the second value to 20%). have.

As a result of the determination, 16.12%, which is the ratio occupied by the area of the residual food 240a in the region of the epiglottis 220, is distributed within the first value (5%) or more and less than the second value (20%). 150) determines the severity of dysphagia in patients who have taken VFSS images as mild dysphagia.

As described above, when the VFSS-based swallowing dysfunction automatic detection system 100 of the present invention is used, it is possible to determine the severity of dysphagia according to a preset value even without a doctor's diagnosis.

The VFSS-based automatic dysphagia detection system 100 may further include a contrast compensation unit (not shown).

The contrast correction unit (not shown) has an average contrast value of the VFSS image in which the contrast values of the hyoid bone 210 region, the epiglottis 220 region, the anomaly 230 region, and the residual food 240 region are measured. The contrast can be corrected to have a preset difference with.

At this time, the method of correcting the contrast may correct the contrast through gamma correction.

An embodiment of correcting contrast through gamma correction will be described in detail with reference to FIG. 3 below.

3 is a view showing an example of contrast correction through gamma correction according to an embodiment of the present invention.

Referring to FIG. 3, the contrast correction unit (not shown) emphasizes the whiteness of the epiglottis 220 and the abnormal orbital region 230 through gamma correction, and emphasizes the blackness of the bone region of the subject in the VFSS image. By doing so, it is possible to clearly observe the area (ex. Epiglottis, abnormal gyrus, etc.) required when determining dysphagia.

At this time, emphasizing the white may be interpreted as means to correct close to white, and emphasizing the black may be interpreted to mean correcting close to black.

That is, the contrast compensation unit (not shown) is detected in the area detection unit 120 and the food detection unit 130, the hyoid bone 210 region, the epiglottis 220 region, the abnormal vortex 230 region and the residual food 240 By emphasizing the white of the region of interest, such as the region, and by emphasizing the non-interesting region that is not the region of interest, the contrast between the region of interest and the non-interesting region is maximized so that the user can observe the region of interest clearly do.

Hereinafter, an automatic detection method for VFSS dysphagia will be described in detail with reference to FIG. 4.

The VFSS-based dysphagia automatic detection method is performed by the above-described VFSS-based dysphagia automatic detection system, and all detailed configurations applied to the VFSS-based dysphagia automatic detection system can be applied to the VFSS-based dysphagia automatic detection method, Hereinafter, a description will be made focusing on contents not mentioned in the configuration of the VFSS-based swallowing dysfunction automatic detection system.

4 is a flow chart showing the flow of the VFSS-based swallowing dysfunction automatic detection method according to an embodiment of the present invention.

Referring to FIG. 4, the VFSS-based swallowing dysfunction automatic detection method extracts an image in which the location of the hyoid bone 210 is located at the bottom of a plurality of VFSS images, and a boundary is displayed in the hyoid bone 210 region (S110). , In the extracted image, detecting the epiglottis 220 and abnormal region 230 region, and displaying the boundary between the detected epiglottis 220 and abnormal region 230 region (S120), the epiglottis 220 ) And the image of the boundary between the area of the abnormality 230 and the residual food 240 is detected, and the boundary of the detected residual food 240 area is displayed (S130), the epiglottis 220 area, Calculating the area of the area of the abnormal orbital (230) and the area of residual food (240), and calculating the proportion of the area of residual food (240) in the epiglottis (220) and abnormal orbital area (230) (S140) and the It may include a step (S150) of determining the severity of dysphagia by comparing the proportion occupied by the remaining food 240 area with a predetermined value.

In the step of detecting an image in which the location of the hyoid bone 210 is located at the bottom of the plurality of VFSS images (S110), the location of the hyoid bone 210 is expressed as a Y coordinate value, and the hyoid bone among them is compared through comparison between multiple VFSS images. An image having the smallest Y-coordinate value of 210 may be extracted.

The reason for extracting the image with the smallest Y-coordinate value of the hyoid bone 210 is that the epiglottis 220 and the abnormal orbit 230 expand to the maximum when the hyoid bone 210 is positioned at the bottom, so this case is the epiglottis. This is because it is most suitable for observing the region of the song 220 and the abnormal vortex 230.

The region of the epiglottis 220 and the abnormality 230 is detected, a step of displaying a boundary (S120), and the region of residual food 240 in the region of the detected epiglottis 220 and the abnormality 230 is detected, In the step of displaying the boundary (S130), by applying a deep learning technique, it is possible to improve the accuracy of detecting the epiglottis 220 region, the abnormal gyrus 230 region, and the remaining food 240 region, by displaying the boundary , The doctor can make the diagnosis easily, and the doctor can show the patient with a bordered image and easily explain the severity of dysphagia.

In addition, as a method of displaying the boundary, a method of displaying an outer boundary as a boundary line, a method of painting and displaying the entire region of interest, a method of correcting and displaying the region of interest and a non-interested region to have a difference in contrast can be used. have.

In addition, in the step (S130) of detecting the region of the epiglottis 220 and the residual food 240 in the region of the abnormality 230, when the subject ingests food containing barium solution that can be displayed on the VFSS image, The food that contains barium dragons remaining in the epiglottis 220 and the abnormal orbital region 230 may be displayed on the VFSS image to detect the remaining food 240 region.

In the step (S140) of calculating the area of the area of the residual food 240 in the region of the epiglottis 220 and the abnormal and 230, and calculating the proportion occupied by the region of the residual food 240, the epiglottis 220 and Residual food (240a, 240b) areas within the area of the orbital (230) are calculated, respectively, and a ratio occupied by the remaining food (240a, 240b) areas is respectively dimensioned and displayed on the image.

This can improve the accuracy of diagnosis by dimensioning the proportions of the remaining foods 240a and 240b, respectively, and displaying them on the image.

The criterion for determining the severity of swallowing disorder in comparison with the preset value in step S150 is determined by comparing the ratio occupied by the area of the dimensioned residual food 240 with the preset value and the predetermined value. When the proportion of the remaining food 240 area occupies less than the first value, it can be determined as non- swallowing disorder, and when the proportion is greater than or equal to the first value and less than the second value, it can be determined as mild dysphagia, and the proportion is second. If it is above the value, it can be judged as severe dysphagia.

According to an embodiment of the present invention as described above, by measuring the proportion of the food area remaining in the epiglottis and abnormal areas and the VFSS-based swallowing dysfunction automatic detection system capable of determining the severity of dysphagia through the measured ratio and A detection method is provided.

According to an embodiment of the present invention, by performing deep learning training on the shape and location of the epiglottis and abnormalities, a VFSS-based swallowing dysfunction automatic detection system and a method for detecting the correct epiglottis and abnormalities and regions are provided. do.

According to an embodiment of the present invention, by performing a contrast correction operation on the extracted VFSS image, the VFSS-based swallowing dysfunction automatic detection system that can more clearly check the location of the hyoid bone, epiglottis, abnormal orbital and residual food areas, and A detection method is provided.

In addition, according to an embodiment of the present invention, the VFSS-based swallowing dysfunction automatic detection system and its detection method may be recorded on a computer-readable medium including program instructions for performing various computer-implemented operations. The computer-readable medium may include program instructions, data files, data structures, or the like alone or in combination. The media may be program instructions specially designed and constructed for the present invention, or may be known and usable by those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs, DVDs, and magnetic media such as floptical disks. -Hardware devices specifically configured to store and execute program instructions such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, etc., as well as machine language codes produced by a compiler.

As described above, although one embodiment of the present invention has been described by a limited embodiment and drawings, one embodiment of the present invention is not limited to the above-described embodiment, which is a general knowledge in the field to which the present invention pertains. Various modifications and variations can be made by those who have this description. Accordingly, one embodiment of the present invention should be understood only by the claims set forth below, and all equivalents or equivalent modifications thereof will be said to fall within the scope of the spirit of the present invention.

100: VFSS based dysphagia automatic detection system
110: image extraction unit
120: area detection unit
130: food detection unit
140: calculation unit
150: judgment unit
210: hyoid bone
220: epiglottis
230: Wow
240: residual food

Claims (5)

An image extraction unit for extracting an image in which the location of the hyoid bone is at the bottom of a plurality of VFSS (Videofluoroscopic Swallowing Study) images;
A region detection unit that detects epiglottis, abnormalities, and regions from the image extracted from the image extraction unit;
A food detection unit that detects the epiglottis and abnormalities detected by the region detection unit and the residual food region within the region;
Calculate the area of the epiglottis and abnormalities and areas detected by the area detection unit, calculate the area of the residual food area detected by the food detection unit, and the ratio of the area occupied by the residual food area among the areas of the epiglottis and abnormalities and areas Calculation unit for calculating; And
A determination unit to determine the severity of dysphagia by comparing the proportion of the remaining food and beverage calculated by the calculation unit with a predetermined value;
VFSS-based swallowing difficulty automatic detection system characterized in that it comprises a. According to claim 1,
Further comprising a contrast correction for correcting the contrast so that the numerical value of the contrast of the hyoid area, the epiglottis area, the abnormal area and the residual food area has a predetermined difference from the average contrast value of the VFSS image,
The contrast compensation unit,
VFSS-based swallowing dysfunction automatic detection system characterized by having a difference in contrast by emphasizing whiteness of the epiglottis and abnormal areas and areas through gamma correction, and emphasizing black of a bone area of a subject in the VFSS image. According to claim 1,
The food detection unit,
When the food containing barium solution displayed on the VFSS image taken by the subject remains in the epiglottis and abnormal areas, the image is extracted from the image extraction unit and detected as a residual food region. VFSS-based dysphagia automatic detection system. According to claim 1,
The determination unit,
If the proportion occupied by the residual food area is less than the first value, it is determined as non- swallowing disorder,
When the proportion of the remaining food area occupies less than the first value or less than the second value, it is determined as mild dysphagia,
VFSS-based dysphagia automatic detection system, characterized in that when the proportion of the remaining food area occupies more than the second value, it is determined as severe dysphagia. Extracting an image in which the location of the hyoid bone is located at the bottom of a plurality of VFSS images, and displaying a boundary in the hyoid region;
Detecting an epiglottis and anomalies and regions in the extracted image, and displaying a boundary between the detected epiglottis and anomalies and regions;
Detecting a residual food area in the image in which the boundary between the epiglottis and the abnormal area is displayed, and displaying a boundary of the detected remaining food area;
Calculating an area of the epiglottis area, anomalous area, and residual food area, and calculating a proportion of the residual food area in the epiglottis area and anomaly area; And
Determining the severity of dysphagia by comparing the proportion of the remaining food area with a predetermined value;
VFSS-based swallowing dysfunction automatic detection method comprising a.

Images