首页|Aalborg University Hospital Reports Findings in Type 1 Diabetes(Explainable Mac hine-Learning Models to Predict Weekly Risk ofHyperglycemia, Hypoglycemia, and Glycemic Variability in PatientsWith Type 1 Diabetes Based on Continuous Glucos e ...)
Aalborg University Hospital Reports Findings in Type 1 Diabetes(Explainable Mac hine-Learning Models to Predict Weekly Risk ofHyperglycemia, Hypoglycemia, and Glycemic Variability in PatientsWith Type 1 Diabetes Based on Continuous Glucos e ...)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews Daily News – New research on Nutritional and Metabo lic Diseases and Conditions - Type 1 Diabetes isthe subject of a report. Accord ing to news originating from Aalborg, Denmark, by NewsRx correspondents,researc h stated, “The aim of this study was to develop and validate explainable predict ion models basedon continuous glucose monitoring (CGM) and baseline data to ide ntify a week-to-week risk of CGM keymetrics (hyperglycemia, hypoglycemia, glyce mic variability). By having a weekly prediction of CGM keymetrics, it is possib le for the patient or health care personnel to take immediate preemptive action. ”Our news journalists obtained a quote from the research from Aalborg University Hospital, “We analyzed,trained, and internally tested three prediction models ( Logistic regression, XGBoost, and TabNet)using CGM data from 187 type 1 diabete s patients with long-term CGM monitoring. A binary classificationapproach combi ned with feature engineering deployed on the CGM signals was used to predict hyperglycemia, hypoglycemia, and glycemic variability based on consensus targets (t ime above range 5%, timebelow range 4%, coefficient o f variation 36%). The models were validated in two independent coho rtswith a total of 223 additional patients of varying ages. A total of 46 593 w eeks of CGM data were includedin the analysis. For the best model (XGBoost), th e area under the receiver operating characteristic curve(ROC-AUC) was 0.9 [95% confidence interval (CI) = 0.89-0.91], 0.89 [95% CI = 0.88-0.9], and 0. 8 [95% CI = 0.79-0.81] for p redicting hyperglycemia, hypoglycemia, and glycemic variability in the intervalvalidation, respectively. The validation test showed good generalizability of th e models with ROC-AUC of0.88 to 0.95, 0.84 to 0.89, and 0.80 to 0.82 for predic ting the glycemic outcomes. Prediction models basedon real-world CGM data can b e used to predict the risk of unstable glycemic control in the forthcomingweek. ”
AalborgDenmarkEuropeCyborgsEmerg ing TechnologiesGlucose Metabolism DisordersHealth and MedicineHyperglycem iaHypoglycemiaInsulin DependentDiabetes MellitusMachine LearningMetabol ic Diseases and ConditionsNutritional and MetabolicDiseases and ConditionsR isk and PreventionType 1 Diabetes
NETL
NSTL
