Application of Machine Learning to Predict Weight Loss in Overweight, and Obese Patients on Korean Medicine Weight Management Program

Kim, Eun-Joo; Park, Young-Bae; Choi, Ka-Hye; Lim, Young-Woo; Ok, Ji-Myung; Noh, Eun-Young; Song, Tae Min; Kang, JiHoon; Lee, Hyangsook; Kim, Seo-Young

doi:10.13048/jkm.20015

Cited by 7 publications

(3 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, the use of machine learning models for disease classification has been developing rapidly, both because of the significant amount of data that is being generated by healthcare devices and systems and the magnitude of computational resources available for data calculation and processing ( 15 – 17 ). Obesity researchers and healthcare professionals have access to a wealth of data.…”

Section: Introductionmentioning

confidence: 99%

Age-specific risk factors for the prediction of obesity using a machine learning approach

Jeon

Lee

2023

Front. Public Health

View full text Add to dashboard Cite

Machine Learning is a powerful tool to discover hidden information and relationships in various data-driven research fields. Obesity is an extremely complex topic, involving biological, physiological, psychological, and environmental factors. One successful approach to the topic is machine learning frameworks, which can reveal complex and essential risk factors of obesity. Over the last two decades, the obese population (BMI of above 23) in Korea has grown. The purpose of this study is to identify risk factors that predict obesity using machine learning classifiers and identify the algorithm with the best accuracy among classifiers used for obesity prediction. This work will allow people to assess obesity risk from blood tests and blood pressure data based on the KNHANES, which used data constructed by the annual survey. Our data include a total of 21,100 participants (male 10,000 and female 11,100). We assess obesity prediction by utilizing six machine learning algorithms. We explore age- and gender-specific risk factors of obesity for adults (19–79 years old). Our results highlight the four most significant features in all age-gender groups for predicting obesity: triglycerides, ALT (SGPT), glycated hemoglobin, and uric acid. Our findings show that the risk factors for obesity are sensitive to age and gender under different machine learning algorithms. Performance is highest for the 19–39 age group of both genders, with over 70% accuracy and AUC, while the 60–79 age group shows around 65% accuracy and AUC. For the 40–59 age groups, the proposed algorithm achieved over 70% in AUC, but for the female participants, it achieved lower than 70% accuracy. For all classifiers and age groups, there is no big difference in the accuracy ratio when the number of features is more than six; however, the accuracy ratio decreased in the female 19–39 age group.

show abstract

Section: Introductionmentioning

confidence: 99%

Age-specific risk factors for the prediction of obesity using a machine learning approach

Jeon

Lee

2023

Front. Public Health

View full text Add to dashboard Cite

show abstract

“…Several explainability AI methods have been designed to improve the interpretability of ML models. Kim et al [ 32 ] tested an interpretability method designed for conditional recurrent neural networks to predict weight loss at 16 weeks using features collected across the 16-week study. However, this method does not apply to RF models and does not focus on early weight loss prediction.…”

Section: Introductionmentioning

confidence: 99%

An Explainable Artificial Intelligence Software Tool for Weight Management Experts (PRIMO): Mixed Methods Study

Fernandes¹,

Choi²,

Schauer³

et al. 2023

J Med Internet Res

View full text Add to dashboard Cite

Background Predicting the likelihood of success of weight loss interventions using machine learning (ML) models may enhance intervention effectiveness by enabling timely and dynamic modification of intervention components for nonresponders to treatment. However, a lack of understanding and trust in these ML models impacts adoption among weight management experts. Recent advances in the field of explainable artificial intelligence enable the interpretation of ML models, yet it is unknown whether they enhance model understanding, trust, and adoption among weight management experts. Objective This study aimed to build and evaluate an ML model that can predict 6-month weight loss success (ie, ≥7% weight loss) from 5 engagement and diet-related features collected over the initial 2 weeks of an intervention, to assess whether providing ML-based explanations increases weight management experts’ agreement with ML model predictions, and to inform factors that influence the understanding and trust of ML models to advance explainability in early prediction of weight loss among weight management experts. Methods We trained an ML model using the random forest (RF) algorithm and data from a 6-month weight loss intervention (N=419). We leveraged findings from existing explainability metrics to develop Prime Implicant Maintenance of Outcome (PRIMO), an interactive tool to understand predictions made by the RF model. We asked 14 weight management experts to predict hypothetical participants’ weight loss success before and after using PRIMO. We compared PRIMO with 2 other explainability methods, one based on feature ranking and the other based on conditional probability. We used generalized linear mixed-effects models to evaluate participants’ agreement with ML predictions and conducted likelihood ratio tests to examine the relationship between explainability methods and outcomes for nested models. We conducted guided interviews and thematic analysis to study the impact of our tool on experts’ understanding and trust in the model. Results Our RF model had 81% accuracy in the early prediction of weight loss success. Weight management experts were significantly more likely to agree with the model when using PRIMO (χ2=7.9; P=.02) compared with the other 2 methods with odds ratios of 2.52 (95% CI 0.91-7.69) and 3.95 (95% CI 1.50-11.76). From our study, we inferred that our software not only influenced experts’ understanding and trust but also impacted decision-making. Several themes were identified through interviews: preference for multiple explanation types, need to visualize uncertainty in explanations provided by PRIMO, and need for model performance metrics on similar participant test instances. Conclusions Our results show the potential for weight management experts to agree with the ML-based early prediction of success in weight loss treatment programs, enabling timely and dynamic modification of intervention components to enhance intervention effectiveness. Our findings provide methods for advancing the understandability and trust of ML models among weight management experts.

show abstract

“…Currently, the use of machine learning models for disease classification has been increasing rapidly, both because of the significant amount of data available that is being generated by healthcare devices and systems, and the magnitude of computational resources available for data calculation and processing [11][12][13] . Obesity researchers and health care professionals have access to a wealth of data.…”

Section: Introductionmentioning

confidence: 99%

Age-specific risk factors for the prediction of obesity using a machine learning approach

Jeon

Lee

2022

Preprint

View full text Add to dashboard Cite

Machine Learning is a powerful tool to discover hidden features in various data driven research fields. Obesity involves extremely complex factors, such as biological, physiological, psychological, and environmental factors. A machine learning framework can provide a successful approach to revealing essential risk factors of the complex obesity phenomenon. Over the last two decades, the obesity population (BMI of above 23) in Korea has been rapidly growing. In this work, we assess obesity prediction by utilizing eight Machine Learning algorithms, and identify risk factors of obesity based on the Korea National Health and Nutrition Examination Survey (KNHANES) data 2016-2019. We explore age-specific and gender-specific risk factors of obesity for adults (19-79 years old). Our findings show that the risk factors for obesity are sensitive to age and gender under different Machine Learning algorithms. Both male and female 19-39 age groups show the highest performance of over 70% accuracy and ROC while the 60-79 group shows around 65% accuracy and ROC. Both male and female 40-59 age groups achieved the highest performance of over 70% in ROC, but the female achieved lower 70% in accuracy. Our results highlight that the top four significant features in all age gender groups for prediciting obesity are Triglyceride, ALT(SGPT), Glycated hemoglobin, and urine acid. For the accuracy ratio of the classifiers and age groups, there is no big difference in accuracy when the number of features is more than six, except the accuracy ratio decreased in the female 19-39 age group.

show abstract

Application of Machine Learning to Predict Weight Loss in Overweight, and Obese Patients on Korean Medicine Weight Management Program

Cited by 7 publications

References 35 publications

Age-specific risk factors for the prediction of obesity using a machine learning approach

Age-specific risk factors for the prediction of obesity using a machine learning approach

An Explainable Artificial Intelligence Software Tool for Weight Management Experts (PRIMO): Mixed Methods Study

Age-specific risk factors for the prediction of obesity using a machine learning approach

Contact Info

Product

Resources

About