Using Machine Learning Classifiers to Assist Healthcare-Related Decisions: Classification of Electronic Patient Records

Abstract

Surveillance Levels (SLs) are categories for medical patients (used in Brazil) that represent different types of medical recommendations. SLs are defined according to risk factors and the medical and developmental history of patients. Each SL is associated with specific educational and clinical measures. The objective of the present paper was to verify computer-aided, automatic assignment of SLs. The present paper proposes a computer-aided approach for automatic recommendation of SLs. The approach is based on the classification of information from patient electronic records. For this purpose, a software architecture composed of three layers was developed. The architecture is formed by a classification layer that includes a linguistic module and machine learning classification modules. The classification layer allows for the use of different classification methods, including the use of preprocessed, normalized language data drawn from the linguistic module. We report the verification and validation of the software architecture in a Brazilian pediatric healthcare institution. The results indicate that selection of attributes can have a great effect on the performance of the system. Nonetheless, our automatic recommendation of surveillance level can still benefit from improvements in processing procedures when the linguistic module is applied prior to classification. Results from our efforts can be applied to different types of medical systems. The results of systems supported by the framework presented in this paper may be used by healthcare and governmental institutions to improve healthcare services in terms of establishing preventive measures and alerting authorities about the possibility of an epidemic.

Learning from a Class Imbalanced Public Health Dataset: a Cost-based Comparison of Classifier Performance

“…The authors chose to represent the WEKA cost matrix in the ratio of 1:10, i.e., The cost of FN is ten times more than the cost of the FP. The widely used stratified 10-fold cross-validation was deployed for the testing of the classifiers, due to its relatively low bias and variance [7], [17]. The core classifiers were compared in terms of total cost and true positive rate.…”

“…In the medical diagnosis domain, classifiers have been built to predict diseases such as breast cancer, insomnia, thyroid disease, Parkinson"s disease and even recommend medication [1][2][3][4][5][6]. Pollettini et al [7] propose a classifier which automatically classifies patients into surveillance levels based on associations among patient features and health. Classifiers have also been designed to predict the cost of healthcare services, to predict intensive care unit readmission, mortality rate and life expectancy rate [1], [8].…”

Medical health data analysis based on Spark Mllib

“…EHR (Electronic health records) is the electronic records of people directly formed in health care activities [3].In this paper, the electronic health record data mainly includes sex, age, height, weight and other basic information and heart rate, blood pressure, blood sugar, family history and other biological information [4].At the same time, exercise and sleep health or other health data from the pedometer bracelet are also uploaded as an important feature selection. As shown in the …”

A Health Surveillance Software Framework to deliver information on preventive healthcare strategies

“…To identify children with developmental problems, we created the Automatic-SL system, which aims to assist healthcare professionals in making decisions [2]. Using machine-learning classifiers, this system automatically assigns Surveillance Levels (SLs) to patients based on the patient's information after each medical appointment at primary healthcare centers.…”

“…The manual assignment of SLs is a laborious task that demands trained people and personalized evaluation. Our system has been successfully used in pediatric care [2].…”