Evaluation of blood glucose change by human emitted volatile organic compounds
Faiem, Nabid (2022)
Faiem, Nabid
2022
Bioteknologian ja biolääketieteen tekniikan maisteriohjelma - Master's Programme in Biotechnology and Biomedical Engineering
Lääketieteen ja terveysteknologian tiedekunta - Faculty of Medicine and Health Technology
This publication is copyrighted. Only for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2022-08-03
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202206305941
https://urn.fi/URN:NBN:fi:tuni-202206305941
Tiivistelmä
Continuous blood glucose management is critical for diabetic patients to keep blood glucose in range. The invasive nature of blood glucose monitoring imposes physical and mental chal lenges on both the healthy subject and diabetic patients. Over time, inadequate blood glucose management can damage major organs such as the brain, heart and kidneys. Healthy people can also benefit from continuous blood glucose monitoring to improve their diet and behavioural health.
Detection of metabolism or disease-associated volatile organic compounds (VOCs) such as acetone, ethanol, and methyl nitrate emitted from skin or breath could be a potential alternative to continuous blood glucose monitoring. Detecting these low-concentrated VOCs introduces challenges due to environment and biological interferences.
Metal oxide semiconductor (MOS) sensors with different temperature cycling can be used to detect low concentrated target VOCs from environmental and biological interferences. Tempera ture cycle operation (TCO) increases the sensitivity, cross-selectivity and reliability of sensors to detect sub-ppm level VOCs.
Lakka Health Oy developed a MOS sensor-based VOC detection device called Sokru to de tect low blood glucose-related VOCs using TCO. This device could be a potential alternative to an invasive continuous blood glucose monitor.
In this thesis, the primary aim was to find signal characteristics of Sokru sensor for skin and breath excreted VOCs during fasting and after glucose intake. The association between fasting and glucose-related VOCs and blood glucose was observed. Glucose measurements were con ducted on a single obese healthy subject following a specific protocol where the subject fasted approximately 16 hours before each measurement and drank glucose solution. The second aim of this thesis was to find signal characteristics in a high relative humidity (RH) environment. Dif ferent signal pre-processing methodologies, such as the Savitzky-Golay convolution (SG) filter, were applied to eliminate or reduce high-frequency noise. The normalization methods were to reduce sensor drifting. Blood glucose and VOCs were measured from a Type 1 diabetes mellitus (T1DM) patient. The measurement was conducted in-home environment where the patient moved and performed regular tasks.
The SG filter smooths the cycle signal without altering the shape and transient of the cycle if the cycle resistance and shape remain in the normal range. The normalization methods reduce the drifting of sensors as these normalization methods depend on the cycle's mean. The extracted features from different parts of each cycle are strongly influenced by RH.
Exhalation significantly changes the cycle's shape and resistance level of the sensor. No sig nificant difference was observed between the fasting and after glucose intake, as the fasting glu cose was relatively higher for the subject in all measurements. More study is required regarding the cycle shape changes due to exhalation.
Sokru sensor shows a potential association between the sensor resistance and hypoglycae mia in the T1DM patient. The sensor resistance and shape of Sokru during hypoglycaemia changed significantly compared to normal subjects.
Detection of metabolism or disease-associated volatile organic compounds (VOCs) such as acetone, ethanol, and methyl nitrate emitted from skin or breath could be a potential alternative to continuous blood glucose monitoring. Detecting these low-concentrated VOCs introduces challenges due to environment and biological interferences.
Metal oxide semiconductor (MOS) sensors with different temperature cycling can be used to detect low concentrated target VOCs from environmental and biological interferences. Tempera ture cycle operation (TCO) increases the sensitivity, cross-selectivity and reliability of sensors to detect sub-ppm level VOCs.
Lakka Health Oy developed a MOS sensor-based VOC detection device called Sokru to de tect low blood glucose-related VOCs using TCO. This device could be a potential alternative to an invasive continuous blood glucose monitor.
In this thesis, the primary aim was to find signal characteristics of Sokru sensor for skin and breath excreted VOCs during fasting and after glucose intake. The association between fasting and glucose-related VOCs and blood glucose was observed. Glucose measurements were con ducted on a single obese healthy subject following a specific protocol where the subject fasted approximately 16 hours before each measurement and drank glucose solution. The second aim of this thesis was to find signal characteristics in a high relative humidity (RH) environment. Dif ferent signal pre-processing methodologies, such as the Savitzky-Golay convolution (SG) filter, were applied to eliminate or reduce high-frequency noise. The normalization methods were to reduce sensor drifting. Blood glucose and VOCs were measured from a Type 1 diabetes mellitus (T1DM) patient. The measurement was conducted in-home environment where the patient moved and performed regular tasks.
The SG filter smooths the cycle signal without altering the shape and transient of the cycle if the cycle resistance and shape remain in the normal range. The normalization methods reduce the drifting of sensors as these normalization methods depend on the cycle's mean. The extracted features from different parts of each cycle are strongly influenced by RH.
Exhalation significantly changes the cycle's shape and resistance level of the sensor. No sig nificant difference was observed between the fasting and after glucose intake, as the fasting glu cose was relatively higher for the subject in all measurements. More study is required regarding the cycle shape changes due to exhalation.
Sokru sensor shows a potential association between the sensor resistance and hypoglycae mia in the T1DM patient. The sensor resistance and shape of Sokru during hypoglycaemia changed significantly compared to normal subjects.