Diabetes and Prediabetes Classification Using Glycemic Variability Indices From Continuous Glucose Monitoring Data

Acciaroli, Giada; Sparacino, Giovanni; Hakaste, Liisa; Facchinetti, Andrea; Nunzio, Giorgio Maria Di; Palombit, Alessandro; Tuomi, Tiinamaija; Gabriel, Rafael; Aranda, J López; Vega, Saturio; Cobelli, Claudio

doi:10.1177/1932296817710478

Cited by 34 publications

(56 citation statements)

References 41 publications

(58 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We demonstrate that IGT and T2D patients could be distinguished with accuracy levels as high as 87.1%; this improvement was significant when compared to the baseline of 61.3% established by implementing a logistic regression-based strategy similar to that proposed in Ref. [21].…”

Section: Introductionmentioning

confidence: 73%

“…CGM-acquired signals were initially processed to extract a set of 37 GV indices (including the 25 indices used previously used in Ref. [21] and, for purposes of compressed representation, in Refs. [19,20]).…”

Section: Glycaemic Variability Indicesmentioning

confidence: 99%

“…To the best of our knowledge, Ref. [21] presents the first investigation of a basic classification problem, such as distinguishing between healthy subjects and patients affected by impaired glucose tolerance (IGT) or type 2 diabetes (T2D), using a large subset of GV indices. In that work, the authors implemented a cascade of two logistic regression steps to sequentially determine which of the 102 subjects were healthy and subsequently pinpoint the exact altered metabolic state (IGT or T2D) of the remaining subjects.…”

Section: Introductionmentioning

confidence: 99%

“…In the present paper, which considers a dataset of 62 IGT and T2D subjects evaluated previously by Acciaroli et al [21], we discuss the use of a machine learning approach to further develop the previous work, with the aim of improving the distinction between IGT and T2D. Specifically, we explored two possible avenues of improvement: first, we attempted to increase the quantity of information available for classification by considering a wider subset of GV indices in the literature and complementing these tools using basic information such as age, sex, body mass index (BMI), and waist circumference; second, we tested the performance of more sophisticated techniques-namely, two support vector machine (SVM) variants-which might be more suitable for more challenging classification problems.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Glycaemic variability-based classification of impaired glucose tolerance vs. type 2 diabetes using continuous glucose monitoring data

Longato

Acciaroli

Facchinetti

et al. 2018

Computers in Biology and Medicine

Self Cite

View full text Add to dashboard Cite

Many glycaemic variability (GV) indices extracted from continuous glucose monitoring systems data have been proposed for the characterisation of various aspects of glucose concentration profile dynamics in both healthy and non-healthy individuals. However, the inter-index correlations have made it difficult to reach a consensus regarding the best applications or a subset of indices for clinical scenarios, such as distinguishing subjects according to diabetes progression stage. Recently, a logistic regression-based method was used to address the basic problem of differentiating between healthy subjects and those affected by impaired glucose tolerance (IGT) or type 2 diabetes (T2D) in a pool of 25 GV-based indices. Whereas healthy subjects were classified accurately, the distinction between patients with IGT and T2D remained critical. In the present work, by using a dataset of CGM time-series collected in 62 subjects, we developed a polynomial-kernel support vector machine-based approach and demonstrated the ability to distinguish between subjects affected by IGT and T2D based on a pool of 37 GV indices complemented by four basic parameters-age, sex, BMI, and waist circumference-with an accuracy of 87.1%.

show abstract

Section: Introductionmentioning

confidence: 73%

Section: Glycaemic Variability Indicesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Glycaemic variability-based classification of impaired glucose tolerance vs. type 2 diabetes using continuous glucose monitoring data

Longato

Acciaroli

Facchinetti

et al. 2018

Computers in Biology and Medicine

Self Cite

View full text Add to dashboard Cite

show abstract

“…The same algorithm is used to diagnose diabetes, particularly diabetes retinopathy through extracted features from heart rate (HR) [19]. Another study aims to monitor blood glucose using a sensor of continuous glucose monitoring (CGM) [20]. Classification Logistic regression model and two-step binary logistic regression model are used to analyze the feasibility of using CGM-based GV indices from impaired glucose tolerance (IGT) or T2DM and only IGT.…”

Section: Review Machine Learning Methods For Diabetes Predictionmentioning

confidence: 99%

Machine Learning Methods for Diabetes Prediction

Dzakiyullah¹,

Burhanuddin²,

Ikram³

et al. 2019

IJITEE

View full text Add to dashboard Cite

Machine Learning is one of the methods used for task prediction. In the diabetic’s research field, the application of machine learning is emerging since the advantages of approximation on the prediction technique has significantly given insight for many health practitioners. Machine Learning is utilized in order to handle the uncontrollable risk factor by finding a relation between such a risk factor trough prediction. This study aims to review recent machine learning models that have been used in diabetes prediction with respect to the risk factors in order to prevent diabetes. This study compares the performance of the model by justified the accuracy as the baseline to evaluate the model. The result of this review shows that the Random Forest and Support Vector Machine are the most popular technique among researcher. Moreover, from this study, it can be seen that Type 2 Diabetes Mellitus (T2DM) has been a concern by researchers since the incidence of diabetes was increasing in worldwide today that happened from an uncontrollable risk factor

show abstract

Wie profitieren Menschen mit Diabetes von Big Data und künstlicher Intelligenz?

Kulzer

2021

Diabetologe

View full text Add to dashboard Cite

In diesem Beitrag-Einstellung zu künstlicher Intelligenz in der Medizin -Anwendungen von künstlicher Intelligenz bei Diabetes Prävention des Diabetes • Diagnostik • Präzisionsmedizin • Entscheidungssysteme • Steuerung von Glukoseverläufen • Information, Schulung • Robotik -Ethik QR-Code scannen & Beitrag online lesen Zusammenfassung Die meisten Menschen befürworten die fortschreitende Digitalisierung, den Einsatz von künstlicher Intelligenz (KI), im Gesundheitswesen. Während in der Diabetologie viele Jahre v. a. die KI-Grundforschung dominierte, gibt es mittlerweile eine große Anzahl an KI-Anwendungen, die für den Einsatz in der klinischen Praxis geeignet sind. Diese betreffen insbesondere die Themenbereiche Prävention des Typ-2-Diabetes, Diagnostik und Prognose von Krankheitsverläufen, klinische Entscheidungssysteme, Informationsgewinnung und Schulung. Künstliche Intelligenz spielt bei der Entwicklung von adaptiven, personalisierten Steuerungsalgorithmen bei künftigen möglichst selbststeuernden Systemen zur automatischen Insulindosierung (AID-Systeme ["automated insulin delivery"]) eine wichtige Rolle. Trotz großer Fortschritte gilt es, auch die Limitationen von KI bei Diabetes zu sehen und ethische Standards, wie sie kürzlich von der WHO (Weltgesundheitsorganisation) definiert wurden, in der Praxis einzuhalten.

show abstract

Diabetes and Prediabetes Classification Using Glycemic Variability Indices From Continuous Glucose Monitoring Data

Abstract: Even with a basic classification strategy, CGM-based GV indices show good accuracy in classifying healthy and subjects with diabetes. The classification into IGT or T2D seems, not surprisingly, more critical, but results encourage further investigation of the present research.

Cited by 34 publications

References 41 publications

Glycaemic variability-based classification of impaired glucose tolerance vs. type 2 diabetes using continuous glucose monitoring data

Glycaemic variability-based classification of impaired glucose tolerance vs. type 2 diabetes using continuous glucose monitoring data

Machine Learning Methods for Diabetes Prediction

Wie profitieren Menschen mit Diabetes von Big Data und künstlicher Intelligenz?

Contact Info

Product

Resources

About