CHAPTER 5 A CASE OF DIABETES ONTOLOGY

Transcription

1 CHAPTER 5 A CASE OF DIABETES ONTOLOGY 5.1 About Diabetes The functioning of OntoSri tool is tested with a case of Diabetes ontology. This chapter presents the basics of diabetes, ontology development and use case description of OntoSri with a case of diabetes ontology Glucose is essential for providing energy for normal body functions. In diabetes the blood glucose levels are increased due to relative or absolute deficiency of insulin. Insulin is a hormone. A Hormone is a chemical secreted by one of the glands in our body. This gland is situated in abdomen and is known as pancreas. Insulin acts as a gatekeeper that allows entry of glucose into the cell. If the amount of insulin is abnormal or the function of insulin is at fault excess of glucose accumulates in the body with harmful effects on the cells of various organs. Diabetes is a metabolic disorder in which body is unable to handle glucose for its energy requirements. Insulin is essential for entry of glucose in to the cell that is why 1) Cells cannot adequately utilize glucose, so 2) Body tries to produce more glucose (gluconeogenisis). This is how blood glucose keeps on rising even if there is no food intake. There are two major types of diabetes. The first type i.e. insulin dependent diabetes. It can be treated only with insulin as the main drug. This is common in younger age group of patients. The second type is non-insulin dependent diabetes (Type II). This is more common in India. This usually affects people above 40 years of age. This group can be managed with diet, exercises and oral medications. Some patients in this group may require insulin sometime later in life. The comparison of two types of diabetes is shown in following table

2 Table 5.1: Comparison of Two Types of Diabetes FEATURE TYPE I (IDDM) TYPE II (NIDDM) Age of Onset Younger age group above years Insulin level in blood Negligible Decreased or High Bodyweight Underweight or Normal Overweight or Normal Usual Treatment Insulin Oral and Insulin Inheritance Less Significant More Significant Uncontrolled diabetes can lead to deleterious effects on every part of the body if not looked after well at the right time. Even if one may not, in actual sense suffer the consequences of the disease today, it needs to be adequately treated. Complacency and neglect is always dangerous. The disease has to be well controlled as it is critical in leading normal life. Poorly controlled diabetes can be a major aggravating factor for later diabetes complications. Following table 5.2 shows some of these complication risks. One should know how important it is to control diabetes and not to scare one about this disease. Table 5.2 Complication Risks Complication Retinopathy leading to blindness Kidney disease Foot problems Paralysis Heart disease Frequency, in comparison to non-diabetics 25 times higher times higher times higher 2 times higher 2-4 times higher One should understand how this disease starts making its impression on your life. Usually, disease creeps in silently. One comes to know of it generally during routine blood sugar test for getting licenses, insurance etc. Sometimes it is detected only during treatment of other concomitant illness or before surgery. 135

3 Following are the common manifestations of diabetes. Report as a surprise Increased appetite and thirst Significant weight loss Infections in folds Increased urination Exhaustion Testing Themselves for Diabetes: 1. Everybody with above mentioned symptoms. 2. Close relatives of Diabetes Patients. For Type 1 diabetes hereditary factor is less prominent. If a father is diabetic then the risk to the child is 5-10%. If a mother is diabetic then the risk is 5%. Type 2 or NIDDM has significant genetic predisposition. If one parent has diabetes the risk is 40%. If both parents are diabetic the risk is 90%. 3. Pregnant women Diabetes is not uncommon due to various hormonal and Metabolic changes Uring pregnancy. 4. Those having high blood pressure, obesity, heart disease and paralysis etc. 5. A lady who delivered a baby weighing more than 4.5 kg 6. All above 45 years of age should check their sugar level once in a year. Diabetes Detection: Estimation and interpretation of blood glucose is the only test for diagnosis of diabetes. World Health Organization (WHO) has laid down guidelines for diagnosis. 136

4 Criteria for Diagnosis of Diabetes: Fasting Venous plasma glucose > 140 mg/100ml and Two hours after 75 gm ingestion of glucose > 200 mg /100ml Recently, American Diabetes Association has accepted the fasting plasma glucose values as >126mg/100ml for diagnosis of diabetes. Blood Test Preparation: How one should prepare for the blood test is also very important. The person should have normal diet 3-4 days prior to the test. One shouldnot try to reduce sugar intake purposely just to show less blood sugar values. By this, one is not only fooling the doctor but oneself also. Actually one is delaying the diagnosis which may prove very costly later. One should to diagnose the disease so as to take correct measures to control it at the earliest time. One should remember that by checking sugar levels doctors are not making one a patient but trying to avoid serious complications related to high sugars. A) At the time of blood test the person should be fasting overnight. The person can have water. B) The next morning fasting blood glucose sample is collected. C) Two hours after 75g of glucose load or lunch second blood glucose sample is collected. Another way is, to give post lunch plasma sample on one day and to give fasting glucose sample next morning. A known diabetic patient is supposed to follow the normal daily routine on the day of blood test also. That means the timings of breakfast, medicines and the lunch should be followed on this day as usual. The doctor can do any modification in the drug and the doses if the patient follows a regular schedule on the blood test day. Diabetes management for each person is different. Do not compare one s doses and drugs with other patients. Management of Diabetes is not only sugar control but much more beyond that. Controlling Diabetes: Basic idea of control is that the person should get relief from the symptoms also; blood biochemistry must be in an acceptable range (Table 5.3) so as to avoid acute and longterm complications due to UNCONTROLLED DISEASE. It is important to know the fact that though the blood biochemistry prescribes normal range of values, they may be little more 137

5 or less without causing any alarm signals. Hence one should be aware of the concept of Acceptable Controls. Table 5.3 Blood Biochemistry Parameter Good control Acceptable control Fasting plasma glucose mg/dl 140 mg/dl Post prandial plasma glucose mg/dl 180 mg/dl Glycated haemoglobin <7 % 7 to 8 % Total cholesterol 180 mg/dl 200 mg/dl HDL cholesterol 40 mg/dl 35 mg/dl Triglycerides 150 mg/dl < 180 mg/dl All these values have to be correlated with other factors like associated disease, patient s age etc. The management will depend upon overall health of the patient and certainly not on blood reports! Regular medical examination includes Monthly weight record Blood pressure check up Blood glucose testing Complete lipid profile Urine and blood tests for assessing kidney functions Tests for evaluation of heart disease Examination of eyes Evaluation of diabetic foot changes Any sexual problems Other specialized tests can be suggested depending upon results of tests mentioned above and the clinical examination by the specialist. Diabetes Care Schedule: Every 3-6 Months: Regular visits to your doctor, it includes 138

6 Test for plasma glucose levels Glycated hemoglobin test (such as HbA1c) Examination of injection site. Feet: examined without shoes and socks Every Year: Cholesterol: fasting profile including HDL & LDL cholesterol and triglycerides. Kidneys: microalbumin measured Eyes: examined through dilated pupils Every 2-3 Years: HDL/cholesterol: if the last reading was normal Number of visits to the doctor depends on the control of blood sugar About Diabetes Database: The classification of Diabetes (Super Class) includes 15 subclasses namely, Doctor Details, Patient Details, Medical Category, Diet, Specialization, Complication Of Diabetes, Medical Condition, Diabetes Surveillance Record I, Diabetes Surveillance Record II, Diabetes Hospital Details, Diet Timing, Diet Type, Insulin Does If Appropriate, Blood Glucose, and Affecting Disease and also includes 67 Objects with 257 Instances. Class : Diabetes Sub class: Doctor Details Object : GP Name : Dr.K.Mohan Address Practice nurse Surgery Address : Vaikolkara Street, Thanjavur : Kavitha : Nursing home Phone No :

7 Sub class: Patient Details Object : PNo : 101 Pname : Uma Devi Year of Identification : 2009 Address : Rajiv Street Phone No : Suggested Drugs : Danonil,Glimel Sub class : Medical Category Object : Type : I Year Of Identification Of Type I : 2009 Year Diagnosed : 2009 Symptoms at diagnosis Symptoms of Type I Weight Height : Burning foot : Burning Finger : 85 kg : 155 cm Year tablet started : 2009 Year insulin started : 2011 Family history : Yes Sub class: Diet Object : Tablets : Daonil, Glimel Drug Manufacturer Physical activity Insulin and injecting Hypoglycaemia Hyperglycaemia Blood testing Foot care : Ranbaxy : Exercise : 2010 p.insulin : 2010 Jan : 2009 Nov : Every month : Once in 3 month 140

8 Eye checks Smoking Alcohol : Once in 3 month : No : No Sub class: Specialization Object : Pharmacist : M.Deepa Dietitian Podiatrist : K.Kumaran : P.Sathish Phone No : Sub class : Complication of diabetes Object : Diabetic eye disease : Refractive error Diabetic kidney disease : Alburnia + Diabetic foot disease : Tingling sensation Sub class: Medical Condition Object : High blood pressure : 140/100 High cholesterol : 220 Angina Transient ischaemic attack Stroke Peripheral vascular disease : Ischaemia : Yes : No : No Sub class: Diabetes Surveillance Record I Object : Fasting blood glucose : 8.9 HbA1c% : 10.8 Smoking : Nil Blood pressure : 140/100 Cholesterol :

9 Sub class: Diabetes Surveillance Record II Object : Weight : 85 kg Body Mass : 20.5 Eye Grade : 0.4 Foot Grade : High Risk Sub class: Diabetes Hospital Details Object : Hospital Name : P.K.R.Hospital Hospital Address : Thanjavur Phone No : Sub class: Diet Timing Object : Time Difference : After ½ Hour Sub class: Diet Type Object : Type Name : Type I Sub class: Insulin Does If Appropriate Object : Morning : A.M. Afternoon Evening : A.M. : A.M. Sub class: Blood Glucose Object : Breakfast : A.M. Lunch Dinner : P.M. : A.M. 142

10 Sub class: Affecting Disease Object : Heart Attack : No Transient ischaemic attack : Yes 5.2 Diabetes Ontology Design Individuals : Individuals, represent objects in the domain. This means that two different names could actually refer to the same individual. For example, 101, 102 and DR.K.MOHAN, KAVITHA HOSPITAL might all refer to the same individual. In OntoSri, it must be explicitly stated that individuals are the same as each other, or different to each other otherwise they might be the same as each other, or they might be different to each other. Figure 5.1 shows a representation of some individuals as diamonds in some domain. KAVITHA HOSPITAL DR.K.MOHAN Figure 5.1 Representation of Individuals 143

11 5.2.2 Properties: Properties are binary relations on individuals - i.e. properties link two individuals together. For example, the property IS PATIENT might link the individual 101 to the individual kavitha hospital. Properties can be limited to having a single value i.e. to being functional. They can also be either transitive or symmetric. Figure 5.2 shows a representation of some properties linking some individuals together. IS PATIENT HOSPITAL KAVITHA 101 HAS SIBLING 102 Figure 5.2 Representation of Properties Classes: OntoSri classes are interpreted as sets that contain individuals (Figure 5.3). They are described using formal (mathematical) descriptions that state precisely the requirements for membership of the class. For example, the class contains all the individuals that are represented as circles or ovals, rather like sets in Venn diagrams. 144

12 DR.MOHAN IS PATIENT DR.VINO 102 DR.BALA 101 DOCTOR PATIENT IS PATIENT KAVITHA KUMARAN HOSIPTAL Figure 5.3 Representation of Classes (Containing Individuals) Diabetes Properties: OntoSri Properties represent relationships (Figure 5.4). There are two main types of properties, Object properties and Data type properties. Object properties are relationships between two individuals Object properties link an individual to an individual. Annotation properties can be used to add information (metadata data about data) to classes, individuals and object/data type properties. 145

13 PATIENT HAS AGE Figure 5.4 The Different Types Of Diabetes Properties Symmetric Properties: If a property P is symmetric, and the property relates individual a to individual b then individual b is also related to individual a via property P. Figure 5.5 shows an example of a symmetric property. If the individual 101 is related to the individual 102 via the HASPATIENT property, then one can infer that 102 must also be related to 101 via the HASPATIENT property. In other words, if 101 have a sibling that is 102, then 102 must have a sibling that is 101. HAS SIBLING HAS SIBLING Figure 5.5 An Example of Symmetric Properties 146

14 5.2.6 Antisymmetric Properties: If a property P is antisymmetric, and the property relates individual a to individual b then individual b cannot be related to individual a via property P. Figure 5.6 shows an example of a antisymmetric property. If the individual 101 is related to the individual Dr.Mohan via the ISPATIENTOF property, then it can be inferred that Dr.Mohan is not related to 101 via the ISPATIENTOF property. It is, however, reasonable to state that Dr.Mohan could be related to another individual Kavitha Nursing Home via the is DoctorOf property. In other words, if 101 is a PatientOf Dr.Mohan, then Dr.Mohan cannot be a PatientOf 101, but Dr.Mohan can be a DoctorOf Kavitha Nursing Home. IS PATIENT OF 101 DR.MOHAN IS PATIENT OF Figure 5.6: An Example of Antisymmetric Properties Reflexive properties: A property P is said to be reflexive when the property must relate individual a to itself. In Figure 5.7 one can see an example of this: using the property knows, an individual Dr.Mohan must have a relationship to itself using the property knows. In other words, Dr.Mohan must know herself. However, in addition, it is possible for Dr.Mohan to know other people; therefore the individual Dr.Mohan can have a relationship with 101 along the property knows. 147

15 KNOWS Is Doctor Of DR.MOHAN 101 Figure 5.7An Example of Reflexive Properties Irreflexive properties: If a property P is irreflexive, it can be described as a property that relates an individual a to individual b, where individual a and individual b are not the same. An example of this would be the property DOCTOROF: an individual Dr.Mohan can be related to individual 101 along the property DOCTOROF, but Dr.Mohan cannot be DOCTOROF herself Figure 5.8. IS PATIENT OF DR.MOHAN 101 IS PATIENT OF Figure 5.8 An Example of Irreflexive Properties 148

16 5.2.9 Transitive Properties: If a property P is transitive, and the property relates individual a to individual b and also individual b to individual c, then one can infer that individual a is related to individual c via property P. Figure 5.9 shows an example of a transitive property. If the individual 101 is related to the individual Type, and Type IS TYPE II PATIENT OF Kavitha Hospital, then one can infer that 101 IS TYPE II PATIENT OF Kavitha Hospital. IS TYPE II PATIENT OF I KAVITHA HOSPITAL IS TYPE II PATIENT OF Type 101 IS TYPE II PATIENT OF Figure 5.9 An Example of Transitive Properties Table 5.4 shows five Object Properties namely, Transitive Property, Symmetric Property, Antisymmetric Property, Reflexive Property and Irreflexive Property. The respective relationships are, IS TYPE II PATIENT OF, HAS SIBLING, IS PATIENT OF, IS NOT PATIENT OF, IS DOCTOR OF and KNOWS 149

17 Table 5.4 Concept Level Relationship Object Properties Relationship Description IS TYPE II PATIENT OF X <IS_TYPE II_PATIENT_OF> Y. E.g. 101 < IS_TYPE II_PATIENT_OF> TYPE. Y<IS_TYPE II_PATIENT_OF> Z. E.g. TYPE < IS_TYPE IS TYPE II PATIENT OF Transitive Property II_PATIENT _OF> KAVITHA HOSPITAL. X <IS_TYPE II_PATIENT_OF> IS TYPE II PATIENT OF Z. E.g. 101 < IS_TYPE II_PATIENT _OF> KAVITHA HOSPITAL. Symmetric Property HAS SIBLING X <HAS_SIBLING> Y. E.g. 101 <HAS_SIBLING> 102 Antisymmetric Property IS PATIENT OF X <IS_PATIENT_OF> Y. E.g. 101 <IS_PATIENT_OF> DR.MOHAN. Y < IS_NOT_ PATIENT_OF > IS NOT PATIENT OF X. E.g. DR.MOHAN < IS_NOT_PATIENT_OF > 101. Reflexive Property IS DOCTOR OF X <IS_DOCTOR_OF> Y. E.g. DR.MOHAN <IS_DOCTOR_OF> 101. X <KNOWS> X. E.g. KNOWS DR.MOHAN <KNOWS> DR.MOHAN. IS PATIENT OF X <IS_ PATIENT_OF> Y. E.g. DR.MOHAN <IS_ PATIENT_OF> 101. Irreflexive Property IS NOT PATIENT OF X <IS_NOT_ PATIENT_OF> Y. E.g. DR.MOHAN <IS_NOT_ PATIENT _OF> DR.MOHAN. 5.3 Description of Use Cases of OntoSri for the Case Study OntoSri provides developers, researchers, educators, and learners the ability to retrieve information about Diabetes systems meeting specific criteria, to get detailed information about a particular system of interest, to add a system to the database, and to provide feedback about OntoSri or any Diabetes system currently classified in OntoSri. 150

18 Figure 5.10 is a screenshot of the homepage that displays a visualization of the Diabetes ontology and a toolbar that allows access to the various OntoSri operations. Figure 5.10 OntoSri Homepage. 151

19 Figure 5.11 OntoSri Create Class. Name: Create Class Description: Create a new class for Ontology. Actors: Ontologiest Flow: 1. Ontologiest should click the tab Create Class 2. Enter class name Patient Detial in text tab. 3. Click Create Class button. 4. If the class name already exists, then enter a new name. 5. Click button Create Class to create new class. 6. Repeat step 2 for more classes. 7. Press right to remove one or more classes. Post conditions: The new class will be created as Patient Details. 152

20 Figure 5.12 OntoSri Adding Objects. Name: Create Object properties Description: Create the Object properties of the classes. Actors: Ontologiest Flow: 1. Ontologiest should click the tab Adding Object 2. Choose class Complication of Diabetes from list box tab. 3. Give object name Diabetic eye diseases, Kidney diseases, Foot diseases and data type General. 4. Click Add button. 5. If object name already exists, then enter a new object name. else create an object on that name. 6. Repeat step 2 for more objects. 7. Press right to remove one or more objects. Post conditions: The new object properties will be created Complication of Diabetes. 153

21 Figure 5.13 OntoSri Relationship. Name: Create Relationships Description: Create the relationships between classes. Actors: Ontologiest Flow: 1. Choose class and object of the class1 tab as Patient Details 2. Next, choose class and object of the class2 tab as Doctor Details 3. Click button Create Relation. 4. Repeat step1 to create more relations. Post conditions: Relationships created between the classes Patient Details and Doctor Details. 154

22 Figure 5.14 OntoSri Available Classes. Name: Available classes Description: Shows the classes which are available in the particular ontology. Actors: Ontologiest Flow: 1. Choose Super class (i.e. Diab) 2. Display subclasses of the chosen super class. Post conditions: Will show the subclasses as Patient Details of the chosen super class. 155

23 Figure 5.15 OntoSri Create Instance. Name: Create Instances Description: Create the Instances of objects for the classes. Actors: Ontologiest Flow: 1. Ontologiest should click the tab Create New Instance as Patient Details 2. Choose the class from list box tab. 3. Give the instance values of the text box tab, and click Complete button as PNo, PName, Address, Year of Registration, Phone No, and Prescribed Drugs. 4. Repeat step 1 for more objects. 5. Press right to remove one or more objects. Post conditions: The new instance values of the objects will be created for classes as Patient Details. 156

24 Figure 5.16 OntoSri OntoGraph. Name: OntoGraph Description: Show the particular ontology in graphical view. Actors: Ontologiest Flow: 1. Click the tab OntoGraph 2. Display the title OntoSri Classes. 3. Click OntoSri Classes to visualize Super classes as OntoSri Class. 4. Select any one of the super classes. Shows the subclasses of the Selected super class. 5. Go to initial stage, just click again. Post conditions: Visualizes ontology in graphical view in the form of tree as given Sixteen Nodes in figures. 157

25 Figure 5.17 OntoSri OntoGraph- Find Name: Find Description: To find the particular values of instance Actors: Ontologiest Flow: 1. Click the tab Find 2. Choose the class and object property as Patient Details, PNo and Give the value of the text box. Then Click the button Search, 3.1. If the value is true, then shows the data structure table of the class. 3.2 Otherwise the message shows no record(s) found 4. Click button Show All to view all the values of the particular class, Post conditions: Search results of some or all values of the class of any one of the ontologies. 158

26 Figure 5.18 OntoSri OntoGraph- Reasoner Name: Reasoner Description: To find values of all ontology Actors: Ontologiest Flow: 1. Click the tab Reasoner 2. Display the text box tab. Enter the string or part of the instance value. 3. Then Click button Search, 3.1. If the value is true, then shows the data(s) and respective classes of all ontology. 3.2 Otherwise the message shows no record(s) found Post conditions: Search results of some or all values of the class of any one of the ontologies. 159