Nutritional intelligence in the food system: Combining food, health, data and AI expertise

Nutrition Bulletin, Год журнала: 2025, Номер unknown

Опубликована: Янв. 12, 2025

Transformative change is needed across the food system to improve health and environmental outcomes. As food, nutrition, data are generated beyond human scale, there an opportunity for technological tools support multifactorial, integrated, scalable approaches address complexities of dietary behaviour change. Responsible technology could act as a mechanistic conduit between research, policy, industry society, enabling timely, informed decision making action by all stakeholders system. Domain expertise in nutrition should always be integrated into both development continuous deployment AI-powered nutritional intelligence (NI) ensure it responsible, accurate, safe, useable effective. Dietary behaviours complex improving diet-related outcomes requires socio-cultural-demographic considerations within design NI tools. This article describes existing examples future opportunities. Human-in-the-loop with experts involved at stages including acquisition, processing, output validation ongoing quality assurance essential evidence-based practice. The same ethical apply this domain any other (e.g. privacy, inclusivity, robustness, transparency accountability) responsible practice must encompass rigorous standards alignment regulatory frameworks. Critical today accessibility appropriate high-quality compositional sets, which include up-to-date information on commercially available products that reflect constantly evolving landscape realise potential AI help challenges.

Язык: Английский

---

Machine Learning Approaches for Stroke Risk Prediction: Findings from the Suita Study

Journal of Cardiovascular Development and Disease, Год журнала: 2024, Номер 11(7), С. 207 - 207

Опубликована: Июль 1, 2024

Stroke constitutes a significant public health concern due to its impact on mortality and morbidity. This study investigates the utility of machine learning algorithms in predicting stroke identifying key risk factors using data from Suita study, comprising 7389 participants 53 variables. Initially, unsupervised k-prototype clustering categorized into clusters, while five supervised models including Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), Light Boosted (LightGBM) were employed predict outcomes. incidence disparities among identified clusters method are substantial, according findings. Supervised learning, particularly RF, was preferable option because higher levels performance metrics. The Shapley Additive Explanations (SHAP) age, systolic blood pressure, hypertension, estimated glomerular filtration rate, metabolic syndrome, glucose level as predictors stroke, aligning with findings approach high-risk groups. Additionally, previously unidentified such elbow joint thickness, fructosamine, hemoglobin, calcium demonstrate potential for prediction. In conclusion, facilitated accurate predictions highlighted biomarkers, offering data-driven framework assessment biomarker discovery.

Язык: Английский

---

A machine learning-based risk score for prediction of mechanical ventilation in children with dengue shock syndrome: A retrospective cohort study

PLoS ONE, Год журнала: 2024, Номер 19(12), С. e0315281 - e0315281

Опубликована: Дек. 6, 2024

Background Patients with severe dengue who develop respiratory failure requiring mechanical ventilation (MV) support have significantly increased mortality rates. This study aimed to a robust machine learning-based risk score predict the need for MV in children shock syndrome (DSS) developed acute failure. Methods single-institution retrospective was conducted at tertiary pediatric hospital Vietnam between 2013 and 2022. The primary outcome DSS. Key covariables were predetermined by LASSO method, literature review, clinical expertise, including age (< 5 years), female patients, early onset day of DSS (≤ 4), large cumulative fluid infusion, higher colloid-to-crystalloid infusion ratio, bleeding, transaminitis, low platelet counts 20 x 10 9 /L), elevated hematocrit, high vasoactive-inotropic score. These analyzed using supervised models, Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM), k-Nearest Neighbor (KNN), eXtreme Gradient Boosting (XGBoost). Shapley Additive Explanations (SHAP) analysis used assess feature contribution. Results A total 1,278 patients included, median patient 8.1 years (IQR: 5.4–10.7). Among them, 170 (13.3%) required ventilation. fatality rate observed group than that non-MV (22.4% vs. 0.1%). RF SVM models showed highest model discrimination. SHAP explained significant predictors. Internal validation predictive consistency predicted data, good slope calibration training (test) sets 1.0 (0.934), Brier 0.04. Complete-case construct Conclusions We estimate hospitalized

Язык: Английский

---

The association of lifestyle with cardiovascular and all-cause mortality based on machine learning: a prospective study from the NHANES

BMC Public Health, Год журнала: 2025, Номер 25(1)

Опубликована: Янв. 24, 2025

Lifestyle and cardiovascular mortality all-cause have been exhaustively explored by traditional methods, but the advantages of machine learning (ML) over methods may lead to different or more precise conclusions. The aim this study was evaluate effectiveness learning-based lifestyle factors in predicting compare results obtained methods. A prospective cohort conducted using a nationally representative sample adults aged 40 years older, drawn from US National Health Nutrition Examination Survey 2007 2010. participants underwent comprehensive in-person interview medical laboratory examinations, subsequently, their records were linked with Death Index for further analysis. Extreme gradient enhancement, random forest, support vector other are used build prediction model. Within comprising 7921 participants, spanning an average follow-up duration 9.75 years, total 1911 deaths, including 585 cardiovascular-related recorded. model predicted area under receiver operating characteristic curve (AUC) 0.862 0.836. Stratifying into distinct risk groups based on ML scores proved effective. All behaviors associated reduced mortality. As age increases, effects dietary sedentary time become pronounced, while influence physical activity tends diminish. We develop predict developed offers valuable insights assessment individual lifestyle-related risks. It applies individuals, healthcare professionals, policymakers make informed decisions.

Язык: Английский

---

A review of six bioactive compounds from preclinical studies as potential breast cancer inhibitors

Molecular Biology Reports, Год журнала: 2025, Номер 52(1)

Опубликована: Фев. 5, 2025

Язык: Английский

---

Machine learning: An effective tool for monitoring and ensuring food safety, quality, and nutrition

Food Chemistry, Год журнала: 2025, Номер 477, С. 143391 - 143391

Опубликована: Фев. 12, 2025

Язык: Английский

---

Prediction of depressive disorder using machine learning approaches: findings from the NHANES

BMC Medical Informatics and Decision Making, Год журнала: 2025, Номер 25(1)

Опубликована: Фев. 17, 2025

Depressive disorder, particularly major depressive disorder (MDD), significantly impact individuals and society. Traditional analysis methods often suffer from subjectivity may not capture complex, non-linear relationships between risk factors. Machine learning (ML) offers a data-driven approach to predict diagnose depression more accurately by analyzing large complex datasets. This study utilized data the National Health Nutrition Examination Survey (NHANES) 2013–2014 using six supervised ML models: Logistic Regression, Random Forest, Naive Bayes, Support Vector (SVM), Extreme Gradient Boost (XGBoost), Light Boosting (LightGBM). Depression was assessed Patient Questionnaire (PHQ-9), with score of 10 or higher indicating moderate severe depression. The dataset split into training testing sets (80% 20%, respectively), model performance evaluated accuracy, sensitivity, specificity, precision, AUC, F1 score. SHAP (SHapley Additive exPlanations) values were used identify critical factors interpret contributions each feature prediction. XGBoost identified as best-performing model, achieving highest highlighted most significant predictors depression: ratio family income poverty (PIR), sex, hypertension, serum cotinine hydroxycotine, BMI, education level, glucose levels, age, marital status, renal function (eGFR). We developed models for interpretation. identifies key associated depression, encompassing socioeconomic, demographic, health-related aspects.

Язык: Английский

---

An interpretable machine learning model with demographic variables and dietary patterns for ASCVD identification: from U.S. NHANES 1999–2018

BMC Medical Informatics and Decision Making, Год журнала: 2025, Номер 25(1)

Опубликована: Март 3, 2025

Current research on the association between demographic variables and dietary patterns with atherosclerotic cardiovascular disease (ASCVD) is limited in breadth depth. This study aimed to construct a machine learning (ML) algorithm that can accurately transparently establish correlations variables, habits, ASCVD. The dataset used this originates from United States National Health Nutrition Examination Survey (U.S. NHANES) spanning 1999–2018. Five ML models were developed predict ASCVD, best-performing model was selected for further analysis. included 40,298 participants. Using 20 population characteristics, eXtreme Gradient Boosting (XGBoost) demonstrated high performance, achieving an area under curve value of 0.8143 accuracy 88.4%. showed positive correlation male sex ASCVD risk, while age smoking also exhibited associations risk. Dairy product intake negative correlation, lower refined grains did not reduce risk Additionally, poverty income ratio calorie non-linear disease. XGBoost significant efficacy, precision determining relationship characteristics participants U.S. NHANES 1999–2018

Язык: Английский

---

Machine Learning and SHAP Value Interpretation for Predicting Comorbidity of Cardiovascular Disease and Cancer with Dietary Antioxidants

Redox Biology, Год журнала: 2024, Номер 79, С. 103470 - 103470

Опубликована: Дек. 16, 2024

Язык: Английский

---

Machine Learning Approaches for Stroke Risk Prediction: Findings from the Suita Study

Опубликована: Май 14, 2024

Stroke constitutes a significant public health concern due to its impact on mortality and morbidity. This study investigates the utility of machine learning algorithms in predicting stroke identifying key risk factors using data from Suita study, comprising 7,389 participants 53 variables. Initially, unsupervised K-prototype clustering categorized into clusters, while five supervised models including Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM), eXtreme Gradient Boosting (XGBoost), Light Boosted (Light-GBM) were employed predict outcomes. incidence disparities among identified clusters method are substantial, according findings. Supervised learning, particularly RF was preferable option because higher levels performance metrics. The Shapley Additive Explanations (SHAP) age, systolic blood pressure, hypertension, estimated glomerular filtration rate, metabolic syndrome, glucose level as predictors stroke, aligning with findings approach high-risk groups. Additionally, previously unidentified such elbow joint thickness, fructosamine, hemoglobin, calcium demonstrate potential for prediction. In conclusion, facilitated accurate predictions highlighted biomarkers, offering data-driven framework assessment biomarker discovery.

Язык: Английский

---