Method and diagnostic apparatus for determining constipation using machine learning model
A method for determining whether constipation is present by using a machine learning model. The method includes a process of analyzing a mixture of a gut-derived substance collected from a subject and a gut environment-like composition, a process of extracting multiple microbial data based on an analysis result of the mixture, a process of selecting microbe-related features to be used in the machine learning model from the multiple microbial data based on a predetermined feature selection algorithm, a process of training the machine learning model with the microbe-related features, and a process of inputting, to the trained machine learning model, the microbial data collected from the subject to be tested and determining whether constipation is present.
1. A method for diagnosing the presence or absence of constipation by using a machine learning model, the method comprising:
a process of analyzing a mixture of a gut-derived substance collected from a subject and a gut environment-like composition;
a process of extracting microbial data based on an analysis result of the mixture;
a process of selecting microbe-related features to be used in the machine learning model from the microbial data based on a feature selection algorithm;
a process of training the machine learning model with the microbe-related features; and
a process of inputting, to the trained machine learning model, the microbial data collected from the subject to be tested and determining whether the constipation is present,
wherein the microbe-related features include an amount of one or more microbes selected from genera included in families of Lachnospiraceae, Erysipelatoclostridiaceae, Pseudomondaeae, Prevotellaceae, Desulfovibrionaceae, Clostridiaceae, Gemellaceae, Bacteroidaceae, Streptococcaceae, Anaerofustaceae, Monoglobaceae, and RF39, and
wherein the gut environment-like composition includes one or more of sodium chloride (NaCl), sodium carbonate (NaHCO 3 ), potassium chloride (KCl) and hemin.
2. The method of claim 1 ,
wherein the number of the microbe-related features to be used in the machine learning model is 12 to 20.
3. The method of claim 1 ,
wherein the process of analyzing the mixture includes:
a process of culturing the mixture for 18 to 24 hours under anaerobic conditions; and
a process of analyzing a culture in which the mixture has been cultured.
4. The method of claim 3 ,
wherein the process of analyzing the culture includes:
a process of centrifuging the culture to separate a supernatant and a precipitate and analyzing the supernatant and the precipitate.
5. The method of claim 3 ,
wherein the microbial data include at least one of an amount, a concentration and a kind of a substance contained in the culture, and changes in a kind, a concentration, an amount and a diversity of bacteria included in microbiota, and
the substance contained in the culture includes at least one of endotoxins, hydrogen sulfides, short-chain fatty acids (SCFAs) and microbiota-derived metabolites.
6. The method of claim 1 ,
wherein the feature selection algorithm includes at least one of a Boruta algorithm and a recursive feature elimination (RFE) algorithm.
7. The method of claim 1 ,
wherein the machine learning model includes at least one of a linear regression analysis (LRA) model, a random forest model, a generalized linear (GLM) model, a gradient boosting model, and an extreme gradient boosting (XGB) model.
8. The method of claim 1 ,
wherein the microbe-related features include an amount of one or more microbes selected from species included in genera of Eubacterium, Ruminococcus, Pseudomonas, Prevotella, Bilophila, Clostridium, Gemella, Bacteroides, Streptococcus , Anaerofustis, Monoglobus, and RF39.
9. An apparatus of diagnosing the presence or absence of constipation by using a machine learning model, comprising:
at least one processor; and
at least one memory for storing a computer program code,
wherein the at least one processor, when the computer program code is executed, is configured to:
extract microbial data based on an analysis result of a mixture of a gut-derived substance collected from a subject and a gut environment-like composition;
select microbe-related features to be used in the machine learning model from the microbial data based on a feature selection algorithm;
train the machine learning model with the microbe-related features; and
input, to the trained machine learning model, the microbial data collected from the subject to be tested and diagnose the constipation,
wherein the microbe-related features include an amount of one or more microbes selected from genera included in families of Lachnospiraceae, Erysipelatoclostridiaceae, Pseudomondaeae, Prevotellaceae, Desulfovibrionaceae, Clostridiaceae, Gemellaceae, Bacteroidaceae, Streptococcaceae, Anaerofustaceae, Monoglobaceae, and RF39, and
wherein the gut environment-like composition includes one or more of sodium chloride (NaCl), sodium carbonate (NaHCO 3 ), potassium chloride (KCl) and hemin.
10. The apparatus of claim 9 ,
wherein the number of the microbe-related features to be used in the machine learning model is 12 to 20.
11. The apparatus of claim 9 ,
wherein the microbial data include at least one of an amount, a concentration and a kind of a substance contained in a culture in which the mixture has been cultured for 18 to 24 hours under anaerobic conditions, and changes in a kind, a concentration, an amount and a diversity of bacteria included in microbiota, and
the substance contained in the culture includes at least one of endotoxins, hydrogen sulfides, short-chain fatty acids (SCFAs) and microbiota-derived metabolites.
12. The apparatus of diagnosing the presence of absence of claim 9 ,
wherein the feature selection algorithm includes at least one of a Boruta algorithm and a recursive feature elimination (RFE) algorithm.
13. The apparatus of claim 9 ,
wherein the machine learning model includes at least one of a linear regression analysis (LRA) model, a random forest model, a generalized linear (GLM) model, a gradient boosting model, and an extreme gradient boosting (XGB) model.
14. The apparatus of claim 9 ,
wherein the microbe-related features include an amount of one or more microbes selected from species included in genera of Eubacterium, Ruminococcus, Pseudomonas, Prevotella, Bilophila, Clostridium, Gemella, Bacteroides, Streptococcus , Anaerofustis, Monoglobus, and RF39.
15. The method of claim 1 ,
wherein the gut environment-like composition further includes L-cysteine hydrochloride and mucin,
wherein the L-cysteine hydrochloride is at a concentration of 0.001% w/v to 5% w/v, and the mucin is at a concentration of 0.01% w/v to 5% w/v.
16. The apparatus of claim 9 ,
wherein the gut environment-like composition further includes L-cysteine hydrochloride and mucin,
wherein the L-cysteine hydrochloride is at a concentration of 0.001% w/v to 5% w/v, the mucin is at a concentration of 0.01% w/v to 5% w/v.