KR-20260063966-A - IMAGE-BASED BILIRUBIN LEVEL ESTIMATION METHOD, SYSTEM, AND COMPUTER PROGRAM FOR JAUDICE DIAGNOSTIC AID

KR20260063966AKR 20260063966 AKR20260063966 AKR 20260063966AKR-20260063966-A

Abstract

A bilirubin level estimation system according to one embodiment of the technical concept of the present disclosure comprises a memory for storing at least one instruction and at least one processor for executing and processing said at least one instruction, wherein the at least one processor extracts a scleral color feature from a scleral image of an eye, said scleral image includes a region including the eye and a region corresponding to a color patch including a plurality of colors, corrects the scleral color feature based on a comparison result between the color feature extracted from the color patch and the color feature extracted from a reference color patch, and estimates a bilirubin level based on the corrected scleral color feature.

Inventors

구형일
최소미
권경범

Assignees

아주대학교산학협력단

Dates

Publication Date: 20260507
Application Date: 20241031

Claims (15)

A bilirubin level estimation system for jaundice diagnosis assistance comprising at least one computing device, Memory for storing at least one instruction; and It includes at least one processor that executes and processes the above at least one instruction, and The above-mentioned at least one processor is, Scleral color features are extracted from an image of the sclera of an eye, - the sclera image includes an area including the eye and an area corresponding to a color patch including a plurality of colors -, Correcting the scleral color features based on the comparison result between the color features extracted from the above color patch and the color features extracted from the reference color patch, and Estimating bilirubin levels based on corrected scleral color features, Bilirubin level estimation system.
In paragraph 1, A photographing unit further comprising a photographing unit that acquires an image of the sclera by photographing the eye including the color patch above. Bilirubin level estimation system.
In paragraph 1, The above at least one processor is, A scleral region mask corresponding to the scleral region of the eye is obtained from the above sclera image, and A scleral region image is extracted by applying the scleral region mask to the above scleral image, and Extracting scleral color features of the scleral region image from the pixel values of the extracted scleral region image, Bilirubin level estimation system.
In paragraph 3, The above at least one processor is, Extracting the scleral color features from the scleral region image using a Gaussian Mixture Model (GMM), Bilirubin level estimation system.
In paragraph 4, The above scleral color features include the mean, covariance, and weight for each color extracted from the color space, Bilirubin level estimation system.
In paragraph 1, The above at least one processor is, Extracting the first color feature of the color patch included in the above sclera image, and Extracting the second color feature of the above reference color patch, and Establish parameters for color transformation of the sclera color feature based on the difference between the first color feature and the second color feature, and Applying estimated parameters to the above scleral color features and obtaining corrected scleral color features based on the application result, Bilirubin level estimation system.
In paragraph 1, The above-mentioned at least one processor is, Acquiring a bilirubin value corresponding to the corrected scleral color feature using a neural network trained to output a bilirubin value from a scleral color feature, Bilirubin level estimation system.
A method for estimating bilirubin levels for jaundice diagnosis assistance performed by at least one computing device, Step of extracting scleral color features from an image of the eye's sclera - the sclera image includes an area including the eye and an area corresponding to a color patch including a plurality of colors -; A step of correcting the scleral color feature based on the result of comparison between the color feature extracted from the color patch and the color feature extracted from the reference color patch; and A step comprising estimating bilirubin levels based on corrected scleral color features, Method for estimating bilirubin levels.
In paragraph 8, Prior to the step of extracting the above-mentioned sclera color features, the method further comprises the step of obtaining the sclera image by photographing the eye including the above-mentioned color patch. Method for estimating bilirubin levels.
In paragraph 8, The step of extracting the above sclera color features is, A step of obtaining a scleral region mask corresponding to the scleral region from the above scleral image; A step of extracting a scleral region image by applying the scleral region mask to the scleral image; and A step comprising extracting scleral color features of the scleral region image from the pixel values of the extracted scleral region image, Method for estimating bilirubin levels.
In Paragraph 10, The step of extracting scleral color features of the scleral region image from the pixel values of the scleral region image is: A method comprising the step of extracting scleral color features from pixel values of the scleral region image using a Gaussian Mixture Model (GMM). Method for estimating bilirubin levels.
In Paragraph 11, The above scleral color features include the mean, covariance, and weight for each color extracted from the color space, Method for estimating bilirubin levels.
In paragraph 8, The step of correcting the above-mentioned sclera color characteristics is, A step of extracting a first color feature of a color patch included in the above sclera image; A step of extracting a second color feature of the above reference color patch; A step of estimating a parameter for color transformation of the sclera color feature based on the difference between the first color feature and the second color feature; and A method comprising the step of applying estimated parameters to the scleral color features and obtaining corrected scleral color features based on the application results. Method for estimating bilirubin levels.
In Paragraph 13, The step of estimating the bilirubin level above is, A method comprising the step of obtaining a bilirubin value corresponding to the corrected scleral color feature using a neural network trained to output a bilirubin value from the scleral color feature, Method for estimating bilirubin levels.
A computer program stored on a computer-readable recording medium for executing a method for estimating bilirubin levels according to any one of paragraphs 8 through 14 on a computer.

Description

Image-based bilirubin level estimation method, system, and computer program for jaundice diagnostic aid The technical concept of the present disclosure relates to an image-based bilirubin level estimation method, system, and computer program to assist in the diagnosis of jaundice. Various methods are used in the field of medical diagnosis to effectively diagnose jaundice. One of the primary methods for diagnosing jaundice is to assess its severity by measuring bilirubin levels in the blood. Additionally, the condition of the liver can be evaluated through various blood tests, including liver function tests. Tests such as ultrasound and computed tomography (CT) are used to check the condition of organs related to the liver. In some cases, a liver biopsy is performed to obtain tissue samples for further analysis. Furthermore, cerebrospinal fluid (CSF) analysis can diagnose liver problems that are causing the jaundice. These diverse diagnostic methods help to comprehensively identify the underlying causes and severity of jaundice. Jaundice is primarily assessed through bilirubin levels, which are metabolites of hemoglobin, a component of red blood cells. If bilirubin is not excreted from the body, it accumulates in the blood, potentially causing jaundice. The normal range for total bilirubin levels in adults is 0.2 to 1.2 mg/dL. For example, a bilirubin level of 2.4 mg/dL may be diagnosed as jaundice. As described above, conventional methods for diagnosing jaundice require invasive procedures on the human body to obtain bilirubin levels. Here, "invasiveness" is a term primarily used in biology and medicine, referring to the process by which external substances or microorganisms enter the body and exert influence. For example, bacteria penetrating the skin and entering the body can be considered invasiveness. However, since such invasion is the process of foreign substances or microorganisms entering the body, it can have negative effects such as infection or inflammation. For example, if an invasive method is used to diagnose jaundice, the patient faces a higher risk of wound infection. Meanwhile, a smartphone-based total bilirubin measurement system for jaundice screening is being developed as a technology that is safer than existing serum-based diagnostics and can be used even without medical facilities. Although this system allows for bilirubin measurement without visiting a hospital separately, it still has limitations, such as the invasive blood collection and the requirement of a separate bilirubin measurement cartridge every time, as shown in Fig. 1. A brief description of each drawing is provided to help to better understand the drawings cited in the present disclosure. Figure 1 is a schematic diagram illustrating a bilirubin measurement cartridge of a smartphone-based bilirubin measurement system for conventional jaundice screening tests. FIG. 2 is a diagram showing the schematic configuration of an image-based bilirubin level estimation system to assist in the diagnosis of jaundice according to an exemplary embodiment of the present disclosure. Figure 3 is a diagram showing the image and the schematic configuration of a patch for assisting in jaundice diagnosis when the sclera is photographed in different shooting environments in the image-based bilirubin level estimation system for assisting in jaundice diagnosis shown in Figure 2. FIGS. 4a and 4b are drawings that exemplarily illustrate the schematic configuration of a scleral color feature extraction unit and the image feature extraction results of an image-based bilirubin level estimation system for assisting in the diagnosis of jaundice according to an exemplary embodiment of the present disclosure. FIG. 5 is a diagram showing the schematic configuration of a scleral color feature correction unit of an image-based bilirubin level estimation system to assist in the diagnosis of jaundice according to an exemplary embodiment of the present disclosure. FIGS. 6a and 6b are drawings showing the color characteristics of a scleral color characteristic correction unit of an image-based bilirubin level estimation system to assist in the diagnosis of jaundice according to an exemplary embodiment of the present disclosure, before and after correction. FIG. 7 is a diagram showing an exemplary configuration of a machine learning unit of a bilirubin level estimation system according to an exemplary embodiment of the present disclosure. FIG. 8 is a flowchart for explaining a method for estimating bilirubin levels according to an exemplary embodiment of the present disclosure. FIG. 9 is a schematic block diagram of a device for performing a bilirubin level estimation system according to an exemplary embodiment of the present disclosure. Exemplary embodiments according to the technical concept of the present disclosure are provided to more fully explain the technical concept of the present disclosure to those skilled in the art, and the following embodiments may