Data Fault Detection in Medical Sensor Networks
Medical body sensors can be implanted or attached to the human body to monitor the physiological parameters of patients all the time. Inaccurate data due to sensor faults or incorrect placement on the body will seriously influence clinicians’ diagnosis, therefore detecting sensor data faults has bee...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2015-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | http://www.mdpi.com/1424-8220/15/3/6066 |
| _version_ | 1818007480408473600 |
|---|---|
| author | Yang Yang Qian Liu Zhipeng Gao Xuesong Qiu Luoming Meng |
| author_facet | Yang Yang Qian Liu Zhipeng Gao Xuesong Qiu Luoming Meng |
| author_sort | Yang Yang |
| collection | DOAJ |
| description | Medical body sensors can be implanted or attached to the human body to monitor the physiological parameters of patients all the time. Inaccurate data due to sensor faults or incorrect placement on the body will seriously influence clinicians’ diagnosis, therefore detecting sensor data faults has been widely researched in recent years. Most of the typical approaches to sensor fault detection in the medical area ignore the fact that the physiological indexes of patients aren’t changing synchronously at the same time, and fault values mixed with abnormal physiological data due to illness make it difficult to determine true faults. Based on these facts, we propose a Data Fault Detection mechanism in Medical sensor networks (DFD-M). Its mechanism includes: (1) use of a dynamic-local outlier factor (D-LOF) algorithm to identify outlying sensed data vectors; (2) use of a linear regression model based on trapezoidal fuzzy numbers to predict which readings in the outlying data vector are suspected to be faulty; (3) the proposal of a novel judgment criterion of fault state according to the prediction values. The simulation results demonstrate the efficiency and superiority of DFD-M. |
| first_indexed | 2024-04-14T05:16:07Z |
| format | Article |
| id | doaj.art-0857ee5200c34702a079a063264aec05 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-14T05:16:07Z |
| publishDate | 2015-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-0857ee5200c34702a079a063264aec052022-12-22T02:10:21ZengMDPI AGSensors1424-82202015-03-011536066609010.3390/s150306066s150306066Data Fault Detection in Medical Sensor NetworksYang Yang0Qian Liu1Zhipeng Gao2Xuesong Qiu3Luoming Meng4State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, No.10 Xitucheng Road, Haidian District, Beijing 100876, ChinaState Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, No.10 Xitucheng Road, Haidian District, Beijing 100876, ChinaState Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, No.10 Xitucheng Road, Haidian District, Beijing 100876, ChinaState Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, No.10 Xitucheng Road, Haidian District, Beijing 100876, ChinaState Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, No.10 Xitucheng Road, Haidian District, Beijing 100876, ChinaMedical body sensors can be implanted or attached to the human body to monitor the physiological parameters of patients all the time. Inaccurate data due to sensor faults or incorrect placement on the body will seriously influence clinicians’ diagnosis, therefore detecting sensor data faults has been widely researched in recent years. Most of the typical approaches to sensor fault detection in the medical area ignore the fact that the physiological indexes of patients aren’t changing synchronously at the same time, and fault values mixed with abnormal physiological data due to illness make it difficult to determine true faults. Based on these facts, we propose a Data Fault Detection mechanism in Medical sensor networks (DFD-M). Its mechanism includes: (1) use of a dynamic-local outlier factor (D-LOF) algorithm to identify outlying sensed data vectors; (2) use of a linear regression model based on trapezoidal fuzzy numbers to predict which readings in the outlying data vector are suspected to be faulty; (3) the proposal of a novel judgment criterion of fault state according to the prediction values. The simulation results demonstrate the efficiency and superiority of DFD-M.http://www.mdpi.com/1424-8220/15/3/6066fault detectionmedical sensorlocal outlier factorfuzzy number |
| spellingShingle | Yang Yang Qian Liu Zhipeng Gao Xuesong Qiu Luoming Meng Data Fault Detection in Medical Sensor Networks Sensors fault detection medical sensor local outlier factor fuzzy number |
| title | Data Fault Detection in Medical Sensor Networks |
| title_full | Data Fault Detection in Medical Sensor Networks |
| title_fullStr | Data Fault Detection in Medical Sensor Networks |
| title_full_unstemmed | Data Fault Detection in Medical Sensor Networks |
| title_short | Data Fault Detection in Medical Sensor Networks |
| title_sort | data fault detection in medical sensor networks |
| topic | fault detection medical sensor local outlier factor fuzzy number |
| url | http://www.mdpi.com/1424-8220/15/3/6066 |
| work_keys_str_mv | AT yangyang datafaultdetectioninmedicalsensornetworks AT qianliu datafaultdetectioninmedicalsensornetworks AT zhipenggao datafaultdetectioninmedicalsensornetworks AT xuesongqiu datafaultdetectioninmedicalsensornetworks AT luomingmeng datafaultdetectioninmedicalsensornetworks |