A non-invasive ultra-wide band based system using artificial intelligence to determine blood glucose level
Diabetes is a serious health concern and declared as global epidemic by WHO due to its rapidly increasing incidence. It is a major cause of mortality worldwide. For a diabetic patient maintenance of blood glucose level within the physiological range is essential to lead a healthy life. The frequent...
Saved in:
| Format: | Thesis |
|---|---|
| Language: | English |
| Subjects: | |
| Online Access: | http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77983/1/Page%201-24.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77983/2/Full%20text.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/77983/4/Md%20Shawkat.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Diabetes is a serious health concern and declared as global epidemic by WHO due to its rapidly increasing incidence. It is a major cause of mortality worldwide. For a diabetic patient maintenance of blood glucose level within the physiological
range is essential to lead a healthy life. The frequent monitoring of blood glucose is an important part of diabetic management specially for type-1 diabetes. A laboratory test or self-test with a small device uses a blood sample collected from a body part with a needle. In extreme cases a diabetic patient needs to undergo this painful process several times a day. To reduce this suffering, a non-invasive (without any blood sample) and patient friendly way of measurement is crucial. Unique advantageous features of UWB technology has demonstrated the widely use of biomedical applications, specially for early breast cancer detection. In the field of exploring potential non-invasive solutions to diabetes detection one promising alternative can be UWB based system using artificial intelligence technique. This relies on variation of dielectric properties (permittivity and conductivity) of target
tissues or cells in a given frequency. Initially the experimental setup was prepared with different types of homemade antennas to select the appropriate antenna type, perfect measurable body place, and to confirm the proof of concept. In integrated
system a rectangular patch antenna was fixed with a transceiver to generate 4.3 GHz frequency and pass through the earlobe. Received discriminated scattered signal was processed and discrete values were reduced to use as input of artificial neural
network (ANN). Number of experiment was conducted to construct an optimal ANN module where actual blood glucose was used as target. The final network output was used to obtain the blood glucose reading from a given scattered signal value. |
|---|