Publication: Automatic diagnosis of venous thromboembolism risk based on machine learning
0
0
Issued Date
2021
Resource Type
Language
eng
File Type
application/pdf
Other identifier(s)
2-s2.0-85106713467
Rights Holder(s)
Scopus
Bibliographic Citation
2021 IEEE International IOT, Electronics and Mechatronics Conference, IEMTRONICS 2021 - Proceedings. Vol , No. (2021)
Suggested Citation
Sukperm A., Rojnuckarin P., Akkawat B., Sa-Ing V. Automatic diagnosis of venous thromboembolism risk based on machine learning. 2021 IEEE International IOT, Electronics and Mechatronics Conference, IEMTRONICS 2021 - Proceedings. Vol , No. (2021). doi:10.1109/IEMTRONICS52119.2021.9422638 Retrieved from: https://hdl.handle.net/20.500.14740/7829
Author(s)
Abstract
Venous thromboembolism (VTE) is an important disease to increase the number of patients because of lacking awareness in Thailand to block the blood flow in the vein. In addition, an effective assessment model of VTE risk is the most important for medical doctors to diagnose. So, this paper represents an automatic diagnosis model by using effective machine learning to predict the important risk factors of VTE from collecting patient data of the medical ward at King Chulalongkorn Memorial Hospital. This research prepares the 83, 850 raw data and investigates the missing values for transforming the data to ready import into each model and then separates the adjusted data for training and testing in the ratio of 70:30. The experimental results were compared to the effectiveness of three machine learning algorithms that consist of the decision tree, logistic regression, and neural network. From the experimental result of the decision tree, this model represents the best assessment model with an accuracy of 96.6% by adjusting the balance data with the class weight method for assisting diagnose the medical doctor. © 2021 IEEE.
Subject(s)
Decision trees
Diagnosis
Hospital data processing
Internet of things
Learning algorithms
Logistic regression
Metadata
Risk assessment
Trees (mathematics)
Assessment models
Automatic diagnosis
Medical doctors
Missing values
Patient data
Training and testing
Venous thromboembolism
Weight methods
Machine learning
Diagnosis
Hospital data processing
Internet of things
Learning algorithms
Logistic regression
Metadata
Risk assessment
Trees (mathematics)
Assessment models
Automatic diagnosis
Medical doctors
Missing values
Patient data
Training and testing
Venous thromboembolism
Weight methods
Machine learning