Harnessing Machine Learning to Personalize Web-Based Health Care Content

Diagnosis Based on Population Data versus Personalized Data: The Evolving Paradigm in Laboratory Medicine

The diagnosis of diseases is a complex process involving the integration of multiple parameters obtained from various sources, including laboratory findings. The interpretation of laboratory data is inherently comparative, necessitating reliable references for accurate assessment. Different types of references, such as reference intervals, decision limits, action limits, and reference change values, are essential tools in the interpretation of laboratory data. Although these references are used to interpret individual laboratory data, they are typically derived from population data, which raises concerns about their reliability and consequently the accuracy of interpretation of individuals' laboratory data. The accuracy of diagnosis is critical to all subsequent steps in medical practice, making the estimate of reliable references a priority. For more precise interpretation, references should ideally be derived from an individual's own data rather than from population averages. This manuscript summarizes the current sources of references used in laboratory data interpretation, examines the references themselves, and discusses the transition from population-based laboratory medicine to personalized laboratory medicine.

Socioeconomic disparities and regional environment are associated with cervical lymph node metastases in children and adolescents with differentiated thyroid cancer: developing a web-based predictive model

Purpose

To establish an online predictive model for the prediction of cervical lymph node metastasis (CLNM) in children and adolescents with differentiated thyroid cancer (caDTC). And analyze the impact between socioeconomic disparities, regional environment and CLNM.

Methods

We retrospectively analyzed clinicopathological and sociodemographic data of caDTC from the Surveillance, Epidemiology, and End Results (SEER) database from 2000 to 2019. Risk factors for CLNM in caDTC were analyzed using univariate and multivariate logistic regression (LR). And use the extreme gradient boosting (XGBoost) algorithm and other commonly used ML algorithms to build CLNM prediction models. Model performance assessment and visualization were performed using the area under the receiver operating characteristic (AUROC) curve and SHapley Additive exPlanations (SHAP).

Results

In addition to common risk factors, our study found that median household income and living regional were strongly associated with CLNM. Whether in the training set or the validation set, among the ML models constructed based on these variables, the XGBoost model has the best predictive performance. After 10-fold cross-validation, the prediction performance of the model can reach the best, and its best AUROC value is 0.766 (95%CI: 0.745-0.786) in the training set, 0.736 (95%CI: 0.670-0.802) in the validation set, and 0.733 (95%CI: 0.683-0.783) in the test set. Based on this XGBoost model combined with SHAP method, we constructed a web-base predictive system.

Conclusion

The online prediction model based on the XGBoost algorithm can dynamically estimate the risk probability of CLNM in caDTC, so as to provide patients with personalized treatment advice.

The characteristics and acceptance of Technology-Enabled diabetes prevention programs (t-DPP) amongst individuals with prediabetes: A scoping review

AIM AND OBJECTIVE

As rising global prevalence of diabetes burdens an overstrained healthcare system, it would be prudent to employ primary prevention strategies. This review aims to detail characteristics of technology-enabled diabetes prevention programs (t-DPP) and the technology acceptance amongst prediabetic individuals.

DESIGN

A scoping review.

REVIEW METHODS

Summative and direct content analysis.

DATA SOURCES

Seven electronic databases-PubMed, Cochrane, Embase, CINAHL, Scopus, PsycINFO and Web of Science-were searched from inception till 9 June 2022 for primary studies conducted on t-DPP. Initial search identified 2412 unique articles. Removal of duplicates and irrelevant articles resulted in 58 full text articles screened and 17 articles meeting the eligibility criteria. There was no limitation to study type or year of publication, but language was limited to English.

RESULTS

Common t-DPP characteristics include physical activity (n = 17), diet control (n = 16), coaching (n = 12), social support (n = 9) and skills acquisition (n = 12). Technological acceptance of t-DPPs were generally positive as participants found them useful (n = 5) and easy to use (n = 4), with majority of the participants interested (n = 5) and engaging well with it (n = 13). However, personal-, design- and technological-level factors were found to negatively influence t-DPPs acceptance.

CONCLUSION

This review reported a generally positive technological acceptance. The result encourages remote delivery of diabetes prevention programs, offering researchers a guide to t-DPP development. However, it also highlights the need for integration of behavioural change theories and socio-cultural considerations, with gaps in knowledge amongst men and young adults.

IMPLICATIONS FOR NURSING

The success of t-DPP can reinforce clinical advice and sustain health behaviours advocated by nurses. Involvement of diabetes-trained nurses would enable continual risk assessment, monitoring and timely intervention to prevent diabetes and potential complications.

REPORTING METHOD

PRISMA-ScR checklist.

Subtypes and Mechanisms of Hypertrophic Cardiomyopathy Proposed by Machine Learning Algorithms

Hypertrophic cardiomyopathy (HCM) is a relatively common inherited cardiac disease that results in left ventricular hypertrophy. Machine learning uses algorithms to study patterns in data and develop models able to make predictions. The aim of this study is to identify HCM subtypes and examine the mechanisms of HCM using machine learning algorithms. Clinical and laboratory findings of 143 adult patients with a confirmed diagnosis of nonobstructive HCM are analyzed; HCM subtypes are determined by clustering, while the presence of different HCM features is predicted in classification machine learning tasks. Four clusters are determined as the optimal number of clusters for this dataset. Models that can predict the presence of particular HCM features from other genotypic and phenotypic information are generated, and subsets of features sufficient to predict the presence of other features of HCM are determined. This research proposes four subtypes of HCM assessed by machine learning algorithms and based on the overall phenotypic expression of the participants of the study. The identified subsets of features sufficient to determine the presence of particular HCM aspects could provide deeper insights into the mechanisms of HCM.