Method and Apparatus for Non-Invasive Detection of Chronic Kidney Disease by
Monitoring Saliva Urea Concentration
1. Field of Invention
This invention relates to a method and apparatus for the non-invasive detection and monitoring of chronic kidney disease by monitoring the saliva urea concentration.
2. Back ground of the invention
It is estimated that nearly 323 million people are affected by kidney disease globally. Kidney disease is viewed as one of the most dangerous disease because once it is affected, it is not possible to repair the damage caused to kidney. Many people die of renal failure mainly due to limited medical facilities.
Commonly, blood test is adopted for detecting chronic kidney disease. Blood test involves an invasive procedure, which causes pain and discomfort to patients especially children. Normal blood test for detecting kidney disease measures the level of urea and creatinine in serum.
Saliva as a potential diagnostic tool is seen attracting lot of research activities nowadays due to its various advantages over other body fluids. Saliva contains various components that can be used for the analysis and detection of various diseases. Recent studies have opened up a new window of opportunity in exploring possibilities of disease detection using salivary biomarkers.
According to recent research findings, saliva based tests could help ascertain whether or not dialysis is necessary, and if so, when it should be administered in patients. So a salivary test can be used as a replacement for blood test for detecting and monitoring chronic kidney disease. A non-invasive extraction method of salivary test is a relatively stress-free process and for that reason, commands greater acceptability among subjects.
Urea and creatinine are the most accepted biomarkers of kidney disease. The primary objective of this invention is to detect chronic kidney disease by monitoring urea concentration in saliva. Urea is a waste product which is produced from protein breakdown. Ammonia is formed when proteins are broken

down to amino acids and deaminated. Ammonia is then converted to urea, which is a non-toxic compound. This conversion is done by the liver and the generated urea is eliminated in urine. According to medical reports, in a healthy individual, saliva urea value ranges from 12 to 70 mg/dL.
There are a number of techniques available for detecting the urea concentration in the body fluids. But most of these are laboratory diagnosis techniques based on diacetyl monoxime method or enzymatic methods. These methods use big devices like chemical analysers which are not suitable for real time applications. Accordingly, there is need for small detection devices to measure urea concentration in body fluids like saliva. Human breath also contains ammonia but its concentration is very less. Highly sensitive sensors are required to measure minute concentrations which make breath based analysis less efficient when compared to saliva based diagnosis for detecting kidney disease.
3. Summary of the Invention
Embodiments of the present invention provide methods and systems for non-invasive monitoring of kidney disease using salivary diagnosis. Urea is considered as potential biomarkers to detect chronic kidney disease. Saliva test can be used as an alternative to blood test to detect chronic kidney disease as saliva contains the chronic kidney disease biomarker urea. Study reports show that there is a positive correlation between blood urea and saliva urea levels in the body.
The main function of kidney is to remove the waste products from the body. So when kidneys don't function effectively, concentration of urea in the body increases. This increase is reflected in saliva urea too. Therefore, by monitoring the urea levels in the saliva, it is possible to analyse the kidney functioning.
One aspect of the present invention is a method for detecting chronic kidney disease non-invasively. The present invention explains a novel method for real time detection of chronic kidney disease by monitoring the urea concentration in saliva. The present invention provides advantages over the commonly used blood test by providing more rapid and stress-free non-invasive method.
Another aspect of the present invention is a method for detecting urea concentration in saliva. Urea in the sample is converted to ammonia using enzymatic method and the ammonia thus produced is measured using a gas sensor which is selective towards ammonia. Sensitivity of the sensor is improved

by selecting suitable toad resistance. The whole conversion process is been carried out in a specially designed detection apparatus gas chamber.
4. Brief Description of the Drawings
Preferred representations of the non-invasive system of the present invention are described in detail below with reference to the drawings wherein:
FIG. I is a diagram illustrating the embodiment of the system.
FIG. 2 is a schematic diagram of gas sensor test circuit.
FIG. 3 is a schematic diagram of gas chamber.
FrG. 4 depicts the response of the sensor for a sample test input for 1000 samples.
FIG. 5 is a flow chart delineating the steps involved in the overall process.
FIG. 6 illustrates the detection results of machine learning algorithm.
5. Detailed Description of the Invention
The methods adopted in the invention are capable of detecting chronic kidney disease, non-invasively. The invention also discloses a reliable method for detecting kidney disease using simple hardware setup and supervised learning technique. One aspect of the present invention is that urea concentration can be detected using simple setup.
Referring to FIG. 1 the overall system consists of an ammonia gas sensor, gas sensing chamber and a processor board. The sensor is mounted inside the gas chamber. A gas sensor which is highly sensitive to ammonia gas can measure the ammonia gas generated inside the chamber.
Referring to FIG. 2, the sensor requires a circuit voltage, a heater voltage and a load resistor. Detection voltage to the sensor is provided by the circuit voltage. Heater voltage provides the working temperature

Next, the output of the sensor was applied to the machine learning algorithms for checking the diagnostic ability of the system to detect the disease. The overall process is illustrated in FIG. 5.
The following examples are offered to claim the current invention.
In a first example of the present invention, the sensor was tested before testing the real samples. Different concentrations of ammonia gas were injected into the chamber through the inlet valve and the response of the sensor was recorded on a digital signal oscilloscope. The output voltage of the sensor increased with increase in ammonia gas concentration inside the gas chamber.
A second example of the present invention, a machine learning algorithm was used for classifying the sensor output as Healthy and Abnormal samples. Support Vector Machine (SVM) learning model was used for the analysis. Initially SVM is trained with some sample data sets. Two features were selected from the sensor output response. These were maximum output voltage and area under the response curve. The main concept of SVM is to separate the data by defining a hyper plane with maximum margin. After training the classifier, all the input samples are tested and classified. FIG. 6 illustrates the detection results of the classifier. The algorithm successfully classified the samples as healthy and abnormal samples with an accuracy of 96.45%.
A benefit of the current invention is that it is a non-invasive process. Moreover, the hardware part used here is portable. Using the methods and apparatus of the present invention it is possible to detect chronic kidney disease in real time.