D-Care: A Non-Invasive Glucose Measuring Technique For Monitoring A Diabetes Patient

Abstract

In the context of 21st century, Diabetes is one of the most lethal disease from which millions of people are suffering of different age level. It is imperative for a diabetes patient to maintain the blood glucose level all the time as there is no stable solution for this disease. However, the infants and the elderly suffering from diabetes might not have the notion of maintaining the level of sugar in their blood. So constant monitoring of the patient has become the most feasible ‘solution’. But the traditional method for measuring blood glucose level is not a viable option for constant blood glucose monitoring. Even if this invasive method is more accurate, it is painful and also there is a risk of infection. That’s why the non-invasive method seems to be the only reasonable option for constant monitoring of blood glucose. For constant remote collection and monitoring of the data obtained from the diabetes patient, an IOT based monitoring system has been proposed in this paper.

Keywords: Noninvasive glucose measure, near infrared, photodiode, GSM module, microcontroller (Arduino UNO)

Introduction

Diabetes is one kind of disease of pancreas which is responsible for producing a hormone called insulin. Diabetes happens when the pancreas gland is no longer able to produce insulin or the body can't properly utilize the insulin delivered by pancreas resulting the level of blood glucose level in blood stream. However human body need some glucose as glucose provides the necessary energy for human body to work. The prevalence of diabetes was about 8% in 2011 which most likely to rise to 10% by 2030. The alarming news is 80% of the diabetes affected people are from low income or middle income countries. In Bangladesh about 10%, in China 9% and in India 8% of the population are affected by the disease. Deaths occurred from Diabetes mellitus in Bangladesh has reached 40,142 which is nearly 5. 09% of total deaths according to WHO.

As the statistics shows, Diabetes is a key reason for death in Bangladesh in recent years. A statistics shows that prevalence of diabetes in Bangladesh has increased, from 4% in 1995 to 2000 and 5% in 2001 to 2005 and 9% from 2006 to 2010. The prevalence of diabetes will reach up to 13% by 2030 according to International Diabetes Federation. The most alarming thing is diabetes affected infants and aged people are less cautious about their condition than the adults and middle aged people. Supposedly middle aged people are more likely cautious about their condition and seems that they are capable of understanding their medical condition better than the aged people and the young ones. Non-invasive method for blood glucose monitoring is becoming more of interest to researchers as it has become a must for constant monitoring of glucose level in blood. In this paper a design has been presented to monitor blood glucose level using near infrared (NIR) light source which is of 940 nm wavelength. Intensity of light passed through the finger is used to calculate blood glucose level. Arm, finger and ear lobe; these three different probes were used to calculate blood glucose using 940 nm NIR LED. The data obtained from the sensor through the microcontroller then is sent to a web server and also to an observer (any relative/someone taking care of the patient) if the level of glucose in blood stream is too low or too high. That way a person can always be aware of any serious condition if such condition arrives.

Related works

In their works, the main objective of their proposed method is to measure the glucose level and the proper insulin dose. R. A. Buda and M. Mohd. Addi et al., in their research work developed non-invasive blood glucose monitoring device to observe glucose concentration in blood, to show the glucose level and the proper insulin dose, resembling to the body mass index (BMI) of the user. They showed 4% - 16% accuracy in glucose detection. M. A. Aizat Rahmat et al., in their research work proposed IoT based non-invasive glucose monitoring technique where they put fingertip into NIR LED to calculate blood glucose and the glucose concentration of the blood. They displayed glucose reading on LCD display and showed it to the family and the doctor via SMS and Android Application for monitoring patient remotely. They discussed about the result of their work had 7. 20% percentage error of glucose reading by comparing their proposed method with invasive method.

Methodology and system architecture

Block Diagram of proposed work

Noninvasive glucose have been measured by near infrared sensor. NIR with 800-1700nm wavelength is suitable for measuring continuous glucose level. NIR sensor transmits continuous wave and a photodiode with 1550nm wavelength receives these continuous wave. After noise filtering and amplifying these wave converted into a suitable voltage value and the microcontroller converted these voltage value with equivalent glucose value. This glucose value is displayed with a LCD monitor. Microcontroller also send this value to a server using GSM module (sim 808). Observer can observe this glucose level obtained from the patient by using a smartphone. And in critical situation emergency SMS is also send to observer.

Conclusion

The non-invasive method using NIR is becoming the only viable option for constant glucose monitoring. As it’s not painful and there’s no risk for infection unlike the traditional invasive methods, more people are willing to utilize the system. This proposed system is a multidimensional health monitoring and a system that combines the monitoring system with notifying the patient and the observer for the patient which ensures the ultimate wellbeing of the patient. The constant remote collection of the patient’s glucose readings will also help any doctor to take decision based on patient’s health if the patient or the observer decides to show the data to the doctor.

Future work

Ongoing research, the obtained result might contain some error. So further we’ll developed our system that will fixed all error also give more accurate value. We also add diet plan for diabetes patient based on their glucose level.