TY - JOUR
T1 - A Collaborative Framework for Sensing Abnormal Heart Rate Based on a Recommender System
T2 - Semantic Recommender System for Healthcare
AU - Guzmán, Giovanni
AU - Torres-Ruiz, Miguel
AU - Tambonero, Vianney
AU - Lytras, Miltiadis D.
AU - López-Ramírez, Blanca
AU - Quintero, Rolando
AU - Moreno-Ibarra, Marco
AU - Alhalabi, Wadee
N1 - Publisher Copyright:
© 2018, Taiwanese Society of Biomedical Engineering.
PY - 2018/12/1
Y1 - 2018/12/1
N2 - According to different studies, people are not able to identify the physiological symptoms related to high risk cardiovascular condition that could require medical attention. In consequence, when they see a medical doctor, the heart damage could be quite advanced. Moreover, there are several studies focused on applying the Framingham or systematic coronary risk evaluation indexes; however, the combination with other physiological variables such as lifestyle, current activity, and maximal heart rate has not been deeply studied in the state-of-the-art. This paper proposes a collaborative framework for sensing physiological variables to determine possible high risk cardiovascular conditions, it will also provide a weighted ranking list of medical speciality centers. The framework will consist of two stages: in the first one, an ubiquitous heart rate monitoring by using an ID3 decision tree is applied to classify sensed-data for identifying the presence of a high risk cardiovascular condition. The second stage proposes a recommender system leading towards extracting and clustering of a set of hospitals, in which the medical specialities are defined in an application ontology. The clustering process matches the hospital attention factor, in order to estimate the number of possible medical doctors and the required cardiovascular medical speciality. In conclusion, the proposal applies different decision trees such as ID3, J48, NBTree, and BFTree in order to evaluate and compare the classification performance. The effectiveness of the ID3 decision tree was 85.71%.
AB - According to different studies, people are not able to identify the physiological symptoms related to high risk cardiovascular condition that could require medical attention. In consequence, when they see a medical doctor, the heart damage could be quite advanced. Moreover, there are several studies focused on applying the Framingham or systematic coronary risk evaluation indexes; however, the combination with other physiological variables such as lifestyle, current activity, and maximal heart rate has not been deeply studied in the state-of-the-art. This paper proposes a collaborative framework for sensing physiological variables to determine possible high risk cardiovascular conditions, it will also provide a weighted ranking list of medical speciality centers. The framework will consist of two stages: in the first one, an ubiquitous heart rate monitoring by using an ID3 decision tree is applied to classify sensed-data for identifying the presence of a high risk cardiovascular condition. The second stage proposes a recommender system leading towards extracting and clustering of a set of hospitals, in which the medical specialities are defined in an application ontology. The clustering process matches the hospital attention factor, in order to estimate the number of possible medical doctors and the required cardiovascular medical speciality. In conclusion, the proposal applies different decision trees such as ID3, J48, NBTree, and BFTree in order to evaluate and compare the classification performance. The effectiveness of the ID3 decision tree was 85.71%.
KW - Bluetooth sensor device
KW - Decision tree
KW - Healthcare
KW - Medical ontology
KW - Monitoring
KW - Recommender system
UR - http://www.scopus.com/inward/record.url?scp=85056418844&partnerID=8YFLogxK
U2 - 10.1007/s40846-018-0421-y
DO - 10.1007/s40846-018-0421-y
M3 - Artículo
SN - 1609-0985
VL - 38
SP - 1026
EP - 1045
JO - Journal of Medical and Biological Engineering
JF - Journal of Medical and Biological Engineering
IS - 6
ER -