DESC:FIELD OF THE INVENTION
The present invention relates to voltage intensity based system for non-invasive glucose monitoring using near-infrared radiation and method thereof.

BACKGROUND OF THE INVENTION
According to the World Health Organization (WHO), there will be more than 300 million diabetic people by the end of 2025. Diabetes is a state where the human body does not produce the amount of insulin as required to regulate the normal blood glucose level. Diabetes can lead to major complications including heart failure, blindness and amputation. In fact diabetes can be considered as one of the major contributors of premature illness and death among non-communicable diseases. Normally all diabetic patients are required to monitor and regulate their blood glucose levels through proper diet and by injecting insulin when needed.

Current methods for measuring blood glucose include pricking the finger, piercing the vein and collecting the blood for testing it for unusual presence of blood glucose. One of the major drawbacks of the existing method is that it is not suitable for the continuous monitoring of blood glucose level. The difficulties associated with this can be removed by making the use of the non-invasive way of measurement.

¬A number of researchers have worked on developing reliable tools, system and methods for performing noninvasive measurements of glucose. And based on noninvasive blood glucose measurement there have been several experimental attempts to develop noninvasive measurements of glucose.
One research paper titled “Design of a Dielectric Spectroscopy Sensor for Continuous and Non-Invasive Blood Glucose Monitoring” published in IJAET in May 2012, describes an architechure based on dielectric spectroscopy for non-invasive continuous glucose monitoring using dielectric spectroscopy through the change in tissue permittivity. The electrical circuit, schematic and the PCB design were made in such a manner that a dielectric constant is used for the measurement of the frequency of both the amplitude and the phase, which in turn indicates the blood glucose level. Thus, blood glucose level is measured using dielectric spectroscopy through the change in tissue permittivity.

Another research paper titled “A novel method for blood glucose measurement by non invasive technique using Laser” published in “World Academy of Science , Engineering and Technology” by Ashok et. al describes a method for assessment of blood glucose level of diabetic and non-diabetic patients using Trans illuminated laser beam. The laser used here is an atomic gas laser also called a He-Ne laser, operating at a 632.8 nm wavelength. Here a single mode laser was used as a monochromatic light source to eliminate mode interference noise. Making use of this method it was able to monitor blood glucose level of the diabetic patient continuously and noninvasively. The result obtained shows that blood flow is directly proportional to blood glucose level.

Wang et. al describes a method based on metabolic heat conformation (MHC) and developed a prototype. In this method glucose level is estimated from the quantity of heat dissipation, local tissue blood flow rate and degree of blood oxygen saturation. Clinical tests were conducted and correlation coefficient of the blood flow rate between this method and the Doppler blood flow meter were found to be equal to 0.914. This result is closer, yet still not acceptable for clinical use.

Amir et. al developed a successful method which was published in Jounal of diabetes, Science and technology under title “Continous glucose monitoring techonolgy based on occlusion spectroscopy” acceptable for clinical use. Amir et. al describes a noninvasive method for detecting blood glucose using the technology of occlusion spectroscopy, developed using a NBM device by Orsense Ltd. In this experiment a light is passed through the finger and the amount of light present on the other side of the finger is measured as photons. The presence of glucose blocks the light from passing through the finger. Therefore, the blood glucose can be measured as light intensity varies.

Despite presence of experimental studies which predict the satisfactory results for the glucose level in the blood. There is still a need of system and method for noninvasive measuring of the glucose level of blood which is reliable, accurate and easy to use.

Accordingly the present invention relates to voltage intensity based system for non-invasive glucose monitoring in the blood by using NIR radiations at body site as per user defined condition. The system comprising of at least one sensor unit, at least one signal conditioning unit, at least one microcontroller unit and at least one communicating device connected either wire or wirelessly to the microcontroller unit. The sensor unit is capable of reflecting the NIR radiation with or without occluding the blood from the dermis tissue of the body site and capable of detecting said reflected NIR radiations having varied intensity characteristics of glucose level in the blood of said dermis tissue as analog voltage signals. The signal unit capable of amplifying and filtering the detected analog voltage signals in the sensor unit. The micro-controller unit comprises of at least one ADC unit, at least one memory unit, at least one processing unit. The ADC unit is capable of converting said analog voltage signal into digital signal and then storing the same in its buffer memory. The memory unit having at least one list of preset values of the glucose level corresponding to the different voltage signals which is acquired by long term machine analysis. The process unit capable of obtaining the glucose level as per user defined condition by performing intensity variation analysis on said stored digital voltage signal in said ADC unit by making the use of the preset data values in the memory unit.

The one Input/ output mean having at least one audio or one visual or one audiovisual means to obtain the glucose level in either audio or visual or audio visual form or inputting the user defined conditions into said memory of said microcontroller unit.

The sensor unit is reflected optical sensor. The sensor unit operates in the range of 800 nm-1100 nm. The reflected optical sensor comprising of at least one infrared emitter for reflecting the NIR radiation from the dermis tissue of the skin and one photodetector unit for detecting the reflected NIR radiations having varies intensity.

The intensity variation analysis is the measuring of the blood glucose level by comparing the voltage intensity difference before and after occlusion.

As per another aspect of the invention, the present invention also provides a method for performing for non-invasive glucose monitoring in the blood by using NIR at body site. The method may comprise the step of setting the non-invasive system in predefined mode as per user defined conditions selected from pre-prandial, post prandial and random condition. The method may further comprise the steps of detecting the said reflected NIR radiations having shifted intensity characteristics of the glucose level in the blood of the dermis tissue as analog voltage then amplifying and filtering said detected analog voltage signals in the signal conditioning unit. The method further may comprise the steps of converting said detected analog voltage signal into digital signal using the ADC unit of the Microcontroller unit and storing the same in ADC buffer and then performing the intensity variation analysis of the store voltage signals by the processing unit of the microcontroller unit as per predefined condition and obtaining the results such as glucose level and diabetic condition of the patient by making the use of the dataset which is based on the long term machine learning analysis and communicating it to the output device through wired or wireless communication means.

As per another aspect of the invention the method of machine learning is provided in which the list of preset data values of the glucose level corresponding to different voltage signals is generated heuristically over a period of time by simultaneously analysing data collected from any known method and results are obtained by the non-invasive system over a period of time from the users of different age group corresponding to different user defined conditions.

OBJECT OF THE INVENTION
In order to obviate the drawback in the prior art , the main object of the present invention is to provide voltage intensity based system for non invasive glucose monitoring using near-infrared radiation.

Another object of the present invention is to provide the system for non glucose monitoring which is highly reliable and cost effective.

SUMMARY OF THE INVENTION
Accordingly the present invention provides the system and method for voltage intensity based non invasive glucose monitoring using near-infrared radiation (hereinafter NIR).

The system uses the concept of analysing reflected signal from the capillary vessel of the dermis tissue in the skin, as glucose content is correlated with the blood glucose in the same way as it correlates to interstitial fluid. It is well-known in the art that the human skin comprises of 3 layers namely the epidermis or the outermost layer, the middle layer i.e. the dermis which is rich in blood carrying capilliaries and the innermost sub-cutaneous layer which comprises of fatty globules which cusion and insulate the inner parts of the body. While the non-invasive devices of the prior art obtain data from the tissues in its entirety, the present invention focuses on the blood in the middle layer or the dermis thereby making the final reading of blood glucose more accurate.

The non invasive glucose monitoring system of the present invention comprises of reflected optical sensor, signal conditioning unit at least one microcontroller unit and atleast one a remote communication device connected wirelessly and a power source which may be integrated into the system. The reflected optical sensor comprises of an infrared emitter and photodetector. The signal conditioning unit comprises of preamplifier and filtering section, an analog to digital converter (ADC) with buffer memory, processing unit to perform intensity based analysis and memory having data set which is based on long term machine learning analysis. The remote communication device may be an audio-video device which is capable of displaying the information such as a smart phone. The power source may be a DC battery.

When user places a finger tip over the reflective optical sensor, the emitted NIR light from the infrared emitter pass through the fingertip and reflects back from the dermis layer of the finger tip before the occlusion and after the occlusion i.e. before and after blocking the blood flow. The photodetector of the reflective optical sensor which is adjacently placed to the infrared emitter detects these reflected signal as the intensity of light falling on said photo detector increases and converts it into voltage depending on the intensity of the signal.

The photodetector may be phototransistor, where the output is taken across the collector of the photransistor and fed into the microcontroller to perform voltage variation analysis. This reflected recieved voltage signal from reflective optical sensor is amplified and filtered in the preamplifier and filter section of the microcontroller, following which, the amplified and filtered signal fed into ADC unit of the microcontroller for the conversion of the said amplified and filtered signal which is an analog signal, to digital signal and stored in the ADC buffer memory.

When buffer memory is filled then the intensity variation analysis is performed by the processor of the microcontroller The analysis is done based on the the data set developed heuristically on the long term machine learning process.

The results thus obtained is displayed on audio/visual device and/ or sent to remote device wirelessly. For example the results can be sent to a remotely located smart phone using Bluetooth or through Internet where the individual would be intimated about the diabetic status and health condition.

The reflective optical sensor usually operates at a wavelength between 800 nm to 1100 nm. One such non-limiting example is that of the reflective optical sensor such as TCRT 5000 which operates at a wavelength of 980 nm.

In the machine learning process, a lot of data is acquired by analyzing the patients using the glucose meter and with the system set-up. For example in a specific interval of time, multiple sets of reading is taken from the non invasive glucose system and glucose meter. Thus data set of “voltage intensity difference” reading corresponding to specific glucose levels is obtained.

The present invention also provides for a method of operating the non-invasive glucose monitoring system. The system may be operated in three modes / phases namely, fasting or pre-parandial mode, post prandial mode and random mode. The readings from the user’s fingertip are taken before occlusion and after occlusion. Meanwhile, the memory of the glucose monitoring system has a list of preset threshold values corresponding to the glucose levels. For each of the three condtions namely fasting/pre-parandial, post-parandial and random reading, there are different threshold values.

When the user switches on the device and selects the mode/phase of choice, the device provides instructions to user for the next steps, said next steps being placing of the finger tip over the reflected optical sensor. The NIR signals falls on the fingertip and reflects back from the dermal layer of the fingertip to the photodetector. The reflected signal at the reflected optical sensor is sent to the microcontroller where voltage conversion takes place. The same procedure is repeated after occlusion, after blocking the flow of blood in the finger. Thereafter the device calculates the difference between the voltage for both cases after occlusion and before occlusion, which is displayed on the visual display.

Meanwhile, this difference in the voltage before and after occlusion is also checked against the dataset stored in the memory. The result obtained by this comparison and validation method, is displayed on the display device of the glucose monitoring system and is further communicated with a remote device wirelessely which enables the users to analyse their current diabetic condition.

Hence present invention provides innovative idea by providing system and method in which correlation between the intensity variation and glucose level is used to determine the glucose level in the blood which provide much reliable, which present systems are facing with current technology.

BRIEF DESCRIPTION OF THE DRAWINGS:
Fig 1 depicts the schematic architecture of the voltage intensity based system for non-invasive glucose monitoring
Fig 2. illustrates the circuit diagram of the system.
Fig 3. depicts the gluck check algorithm used by the microcontroller unit of the present system.
Fig. 4. provides the method of using the voltage intensity based system for non-invasive glucose monitoring
Fig. 5 provides the one example of implemention of method and system of present invention.
Fig 6. shows the plot between Standard deviation and glucometer reading.
Fig 7 shows the plot between the Bias and actual data.

DETAIL DESCRIPTION OF THE INVENTION WITH ILLUSTRATING EXAMPLES:
The present invention relates to system and method for voltage intensity based non invasive glucose monitoring using near-infrared radiation.

The term “occulusion” in the specification is an act of occluding or the state of being occluded : an obstruction to the flow of blood through the blood vessel.

The system uses the concept of analysing reflected signal from the capillary vessel of the dermis tissue in the skin, as glucose content is correlated with the blood glucose in the same way as it correlates to interstitial fluid. It is well-known in the art that the human skin comprises of 3 layers namely the epidermis or the outermost layer, the middle layer i.e. the dermis which is rich in blood carrying capilliaries and the innermost sub-cutaneous layer which comprises of fatty globules which cusion and insulate the inner parts of the body. While the non-invasive devices of the prior art obtain data from the tissues in its entirety i.e, the present invention focuses on the blood in the middle layer or the dermis thereby making the final reading of blood glucose more accurate.

The present invention is based on the concept that the intensity of the reflected infrared radiation from the dermis tissues corresponds to the glucose molecules in the blood of the middle layer dermis tissue. By analyzing the intensity of the reflected signal from the blood dermis tissue as a variations in the analog voltage signals, the glucose level in the blood glucose level and current diabetic condition of the user is computed .

Fig 1 illustrates the block diagram of the non-invasive glucose monitoring system (100) while Fig. 2 shows the schematic circuit diagram of the system. As shown in Fig 1, the non-invasive glucose monitoring system (100) of the present invention comprises of at least one reflected optical sensor (110), at least one signal conditioning unit (120), and at least one microcontroller unit (120), at least one remote communication device connected wired or wirelessly (130) and a power source which may be integrated into the system.

The NIR sensor unit which is reflective optical sensor comprises of at least one infra red emitter and at least one photodetector unit.

The infrared emitter of the optical sensor unit is capable of reflecting the NIR radiations from the dermis tissue of the body site, while the photodetector unit of the optical sensor is capable of detecting the variation in the intensity of NIR radiations.as ananalog voltage signals

The NIR sensor works on the principle that the intensity of the light is inversely proportional to the voltage. When the light intensity falling on the photodetector increases, the conduction through the photodetector is also increases as it is now in ON state. The output is then taken across the collector of the photo detector. As the conduction increases i.e. intensity increases the output voltage across the the collector decreases. This output voltage is then transmitted to the digital end circuit . The NIR sensor usually operates in range of 980 nm.

The digital end circuit is essentially a microcontroller unit (130). Inside the microcontroller unit (130), the ADC unit (131) converts the analog voltage signals which it receives from the signal conditional unit (120) into digital voltage signals such that they can be read by the processing unit (133) of the microcontroller unit (130). The digital voltage signals are bieng stored in the memory unit (132) which is also ADC buffer of the microcontroller unit (130). Once the ADC buffer of the microcontroller unit (130) is filled with the digital voltage signals the processing unit (133) performs the intensity variation.

The Intensity variation analysis is calculated from the difference in voltages obtained before and after the occlusion of blood vessel.

During intensity variation analysis the system use the data set which is being developed heuristically on the long term machine learning process. The results thus obtained is displayed on audio/visual device and/ or sent to remote device wirelessly. For example the results can be sent to a remotely located smart phone using Bluetooth or through Internet where the individual would be intimated about the diabetic status and health condition.

Figure 4 illustrates the method of using non-invasive glucose monitoring system. The user switches on the device and selects the mode/phase of choice in step (401), the device provides instructions to the user for the next steps, said next step being placing of the finger tip over the reflected optical sensor.
The NIR signals falls on the fingertip and reflects back from the dermis layer of the fingertip to the photodetector in step (402). The photodetector of the optical sensor unit detects the variation in the intensity of the reflected signal as analog voltage signal and transfer it to the signal conditioning unit for amplification and filtering purposes in the step (404).
In the step 405, said detected analog voltage signal is converted into digital signal by the ADC unit of the microcontroller unit and stored in its buffer. Once the ADC buffer is filled, then the processor of the microcontroller unit preforms intensity variation analysis on the stored digital data by making the use of the preset data based on the long term machine learning analysis. (406).
In the step 407, the system obtains the results on the output unit such as LCD display of the microcontroller unit in which user glucose level with diabetic condition and further communicated wired or wirelessly to the device which is connected wire or wirelessly to it e.g. Bluetooth.
In one of the embodiment, the method for machine learning process is also provided, in which the list of preset data value corresponding to the voltage signals is generated by simultanously analysing the data collected from any known method and results obtained by said noninvasive system over a period of time from the users of different age group and for different user defined conditions.

During maching learning process a large amount of data is stored and analysed simultaneously by invasive system or glucometer as well as non-invasive system, to create a list of preset data values, from the individual of various age group such that during system operation these preset data values can be utilized by processing unit of the microcontroller as described below to predict the glucose level and diabetic condition of any random user under different conditions such as post prandial or pre- prandial.

PROCESS OF CONDUCTING GLUECHECK
The processing unit of the microcontroller unit performs the analysis on the collected data by employing the gluck check algorithm (shown in Fig 3). The gluck check algorithm provides the provision in which different threshold are set for each condition i.e., fasting , pre-prandial and random condition.

As shown in Fig 3, algorithm checks whether system is set or not (1). When it is set in the correct position, it reads the sensor values which it obtains from the optical sensor unit, before occlusion (4) and calculate the average sensor values (6) and display them for predetermined count values (7). Similiar process is repeated for the occlusion condition (9,10,11,12).

So after calculating the voltages in both conditions i.e. before occlusion and after occlusion, the two sensed voltages are compared with the data set values (17 to 22). In the data set for each corresponding voltage there is glucose values. For each condition i.e. fasting (23), prenadial(30) and random (37) there are threshold values corresponding there is set flag which is set by user predetermined by user according to user defined conditions.

So after reterving the glucose values from the data set, algorithm compares it with the threshold level of glucose corresponding to set flag. By this comparison and validation method, obtained result is displayed on an LCD and futher communicated to the smartphone using Bluetooth.

The system may be operated in three modes / phases namely, fasting or pre-parandial mode, post prandial mode and random mode. The readings from the user’s fingertip are taken before occlusion and after occlusion.

The non-limiting example to explain the present invention are below:

EXAMPLE 1:
Table 1 depicts the specification of the component being used in the present system.
Component Specification Dimension (L*W*H)mm Operating voltage
NIR SENSOR Vishay TCRT 5000 10.2*5.8 3.3-5.0
Microcontroller unit Atmega 328 75*53.5*15 1.8-5.5
Bluetooth Module HC-06 28*15*2.35 3.0-4.2

The power to switch on the device is driven from a DC battery source; same power is used to power the Bluetooth device also. Since the Bluetooth works in 3 V level converter is used. Three switches are used which helps the user to select in which phase the individual need to test. A reset button is also provided in case device hanged.

Fig 5 illustrates the one example of the system and method of implementing the present invention. As depicted in Figure 5 the reflected signal from Finger tip of the NIR sensor (510) is sent to the microcontroller unit after amplification in the differential amplifier (520) and filtered in signal conditioning unit which may be part of microcontroller unit (530) wherein the voltage conversion of the signal is performed.

The microcontroller unit (530) computes the voltage values and caliberates the results by checking the same with the preset dataset using GLUE Check algorithm. The Bluetooth of the smartphone(531) is also switched on and paired with microcontroller unit () such that whenever the full process is over the approximate glucose level along with current diabetic condition is displayed on the screen of smartphone(600) by way of application. For example for healthy person ssytem displays person is “healthy” along with glucose level while for diabetic person the system display “diabetic” along with the glucose value.

The present system has been tested and the means, standard deviation have been found to be consistent and in close correlation with the standard results obtained by known invasive methods.

The present system has been tested for all kind of individuals. Invasive method has been used to estimate the level of glucose level along with the present system. The age limit of individuals were in the range of 22-60.

Table 2 shows one example in which analysis has been performed on an individual at two different periods, i.e, fasting and post-prandial conditions. As shown in table the observations has been made with in an interval of 1 minute and the consistancy of the measured readings has been analyzed in terms of mean and variance. The reference values obtained using glucose meter during fasting and post-prandial was 101 mg/dl and 110 mg/dl. The readings obtained during fasting and post-prandial conditions using Gluck check is as shown below

Table 2
From the table it infers that the proposed system gives the average value of glucose content in an individual during fasting as 95.6 with a standard deviation of 1.95. The average value of glucose content in an individual during post-prandial as 108.4 with a standard deviation 1.85. It also can be inferred from the table that the average difference of person’s glucose content at fasting and post prandial is 12.8 with a standard deviation of 3.42.

An individual representative has been subjected to an experimental analaysis where the blood glucose reading was calculated within an interval of 1min(5readings taken) and system has the inference in blood glucose levels with a deviation in blood glucose level of approx. 3.42mg/dl these are the readings captured during the experimental analysis performed on individuals with different blood glucose levels who fall under different age groups, According to the machine learning processed values from the table 2 the varation in glucose level during fasting can vary to an approx. of +- 1.95mg/dl and during post-prandial it can vary to an approx. of +- 1.85mg/dl
.
Fig 7 shows the plot between the standard deviation and actual data while Fig 8 shows the plot between actual bias and preset data values. The analysis shows that the bias as well as standard deviation decreases as the concenteration of glucose level increase.

Hence present invention provides innovative idea by providing system and methods in which the correlation between the intensity variation and glucose level is used to determine the glucose level in the blood which provide much reliable, which present systems are facing with current technology.
,CLAIMS:We claim
1. Voltage intensity based system for non-invasive glucose monitoring in the blood by using NIR radiations at body site as per user defined condition , said system (100) comprising of
- at least one sensor unit (110), said sensor unit capable of reflecting the NIR radiations with or without occluding the blood from the dermis tissue of the body site and capable of detecting said reflected NIR radiations having varied intensity characterstics of glucose level in the blood of said dermis tissue as analog voltage signals,
- at least one signal conditioning unit (120) is capable of amplifiying and filtering said detected analog voltage signals,
- at least one microcontroller unit (130), said microcontroller unit comprising of
- at least one ADC unit (131) capable of converting said analog voltage signals into digital signals and then storing the same in its buffer memory,
- at least one memory unit (132) having at least one list of preset data values of the glucose level corresponding to different voltage signals which is acquired by long term machine learning analysis of said system ,
- at least one processing unit (133) capable of obtaining the glucose level as per user defined condition by performing intensity variation analysis on said stored digital voltage signal in said ADC unit (131) by making the use of said preset data values in said memory unit (132),
- at least one Input/ output mean having at least one audio or one visual or one audiovisual means to obtain the said glucose level in either audio or visual or audio visual form or inputting the user defined conditions into said memory of said microcontroller unit,
- at least one communicating device (140) connected either wired or wirelessely to said microcontroller for obtaining the user glucose level at remote location.

2. Voltage intensity based system for non-invasive glucose monitoring in the blood by using infra radiations at body site as claimed in claim 1 wherein said user defined condition is selected from fasting, pre-parandial or any random condition.

3. Voltage intensity based system for non-invasive glucose monitoring in the blood by using infra radiations at body site , wherein said sensor unit (100) is reflected optical sensor, said reflected optical sensor comprising of at least one infrared sensor for reflecting the near infrared radiation and at least one photo detector unit for detecting the reflected radiations with varied intensity.

4. Voltage intensity based system for non-invasive glucose monitoring in the blood by using NIR radiations at body site as claimed in claim 1 wherein said intensity variation analysis is the performed by comparing the voltage intensity difference before and after occlusion of the blood.

5. Voltage intensity based system for non-invasive glucose monitoring in the blood by using NIR radiations at body site as claimed in claim1, wherein said sensor(110) operates in the range of 800 nm-1100 nm.

6. Voltage intensity based system for non-invasive glucose monitoring in the blood by using NIR radiations as claimed in claim 1 wherein said body site is finger.

7. A method for performing for non-invasive glucose monitoring in the blood by using NIR at body site comprising steps of
- setting the noninvasive system in predefined mode as per user defined conditions selected from preparandial, post parandial and random conditions,
- reflecting the NIR radiations from with or without blocking blood from the darmis tissue of the body site,
- detecting the said reflected NIR radiations having shifted intensity characterstics of the glucose level in the blood of said dermis tissue as analog voltage signal,
- amplifiying and filtering said detected analog voltage signals in the signal conditioning unit,
- converting said detected analog voltage signals into digital signals using the ADC unit of the Microcontroller unit and storing the same in its buffer,
- performing tintensity variation analysis of the stored digital voltage signals by the processing unit of the microcontroller unit as per predefined condition and obtaining the results by making the use of the dataset which are based on the long term machine learning analysis,
- Obtaining the results which is glucose level on the input/output device and communicating the same to the remote monitoring device output device through wired or wire less communication means for further analysis.

8. A method for performing for non-invasive glucose monitoring in the blood by using NIR at body site as claimed in claim 7, where in the machine learning analysis is a process in which list of preset data values of the glucose level corresponding to different voltage signals generated heuristically over a period of time by simultanously analysing data collected from any known method and results obtained by said noninvasive system (100) over a period of time from the users of different age group corresponding to different user defined conditions.