Previously submitted to: JMIR Medical Informatics (no longer under consideration since Dec 05, 2020)
Date Submitted: Nov 5, 2020
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Predicting complications of diabetes mellitus through machine learning based on topic modeling: study design
ABSTRACT
Background:
Predicting the complications of diabetes mellitus from an early stage would be beneficial for its management. Topic modeling is a posterior procedure to estimate semantic objects in a dataset through a statistical approach. The topic model can play the role of a feature set for supervised classification.
Objective:
: We performed a study to predict diabetic retinopathy (DMR), diabetic nephropathy (DMN), and non-alcoholic fatty liver disease (NAFLD) from clinical notes using semi-supervised classification based on topic modeling.
Methods:
: We applied four types of machine learning algorithms for classification: random forest (RF), gradient boosting machine (GBM), support vector machine (SVM), and fully connected artificial neural network (ANN) We reviewed the topic models through statistical analysis to determine whether these topic models are clinically plausible.
Results:
F1 scores were above 0.8 when predicting all kinds of target diseases with all types of classification methods, and above 0.9 using RF or GBM. Hypertension and dyslipidemia seem to be statistically associated with DMR, DMN, and NAFLD. They may be important clues with which we can predict DMR, DMN, and NAFLD.
Conclusions:
This study showed that complications of diabetes mellitus that are likely to occur later in life can be predicted from the clinical notes of outpatient departments. We believe that this kind of predictive model could be utilized by patients and physicians in outpatient departments as a useful tool, similar to clinical decision support systems.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.