DESIGN AND REALIZATION OF NON-INVASIVE BLOOD GLUCOSE MEASUREMENT SYSTEM PROTOTYPE

• Main Author: JAYAMULIA RUSLI (NIM 23204011); Pembimbing: Prof. Dr. Ir. Soegijardjo Soegijoko, ADI
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/19802
• Institution: Institut Teknologi Bandung

<p align="justify">This thesis describes the completed design and realization of microcontroller-based non-invasive blood glucose measurement system. According to the American Diabetes Association, there are about 20.8 million people in United States, or 7% of the population, who have diabetes. Diabetes is also known as one of the world's topmost diseases that cause death and disability. However, in death report diabetes was rarely considered as the major cause of death.<p align="justify"><p>For many years, conventional invasive method was used to measure patient's blood glucose level. The measurement method used a strip test that collects a sample of patient's blood to measure their blood glucose level. The blood sample was obtained by using a lancet that pricks a finger or other region of the patient's body. This blood sample collecting method can cause trauma to several types of patients. Besides, if patients are suggest to measure their blood glucose level up to 4-6 times a day, this method can increase the infection risk or lead to skin tissue damage. The cost of blood glucose measurement were also expensive.<p align="justify"><p>The non-invasive method used in this thesis is based on differential body temperature measurement. It uses two thermistor-based temperature sensors to detect the temperature of two regions of the ear, tragus and antihelix regions. The comparison of differential temperature is reference values taken from the root mean square of fasting blood glucose and HbA1c value; and differential temperature which taken at the same time with the measurement of blood glucose. If the differential temperature decreases, the blood glucose increases for about 1 mg/dl on every 0.024 oC differential temperature decrease, and vice versa. The reference values have been taken before the measurement and stored in the external memory of the system.<p align="justify"><p>The temperature difference is measured using two thermistors, each with its associated signal conditioning circuit (a Wheatstone bridge and an instrumentation amplifier). The output from the two instrumentation amplifiers are then connected to a multiplexer, 12-bit analog-to-digital converter and ATMEGA microcontroller. The analog multiplexer select analog signal from the instrumentation amplifiers to be convert to digital signal by analog-to-digital converter. The microcontroller process the digital signal from analog-to-digital converter, and stored the measurement result and patient's parameters in the EEPROM. A keypad is used as controller of the system, and LCD module to display the measurement results and other information, such as patient's name and measurement parameters. System design consists of two section, hardware and software design. Hardware design includes sensor's bridge circuit and instrumentation amplifier interface design, analog-to-digital circuit, microcontroller minimum system and peripheral interface such as power supply management, LCD module and keypad interface. Software design includes the software for microcontroller programming, such as measurement system, data acquisition, input process, external memory access, etc.<p align="justify"><p>Experiments on five patients, diabetic or non-diabetic patients have been completed. In general, the results show that this new method is sufficiently effective to detect the tendencies of blood glucose changes. The correlation between differential temperature changes that effect blood glucose level is also exist. But, the system can only measure a slight changes of blood glucose from blood glucose reference, since it will be saturated in certain conditions where differential temperature between tragus and antihelix region is minimum.<p align="justify"><p>For further development, a number of improvements are still required. For instance, personal data and measurement results on some patients can be stored in a single memory chip. Moreover, system integration to PC with additional software to monitor the changes in patient's blood glucose via interactive graphics and charts.