A Fuzzy-Ontology Based Diabetes Monitoring System Using IoT
291
membership fuzzy variables. The fuzzy concepts define concepts and relations
to describe uncertain and vague knowledge. The main difference between fuzzy
and classic concepts is that in classic concepts the membership degree of each
property is equal to 1 or 0 while in fuzzy concepts it is equal to a certain degree
belongs the interval [0, 1]. The classes extended are the following:
– Fuzzy patient class: represents all the information required to supervise the
condition of the diabetic elderly. It describes 5 fuzzy variables which are
height, weight, gender, disease history and age. The fuzzy variable weight
has the fuzzy sets “Light”, “Normal”, and “Heavy” and “Obese”. The fuzzy
variable gender has the fuzzy set “Male” and “Female”. The fuzzy variable
age has these sets, “Young”, “Adult” and “Old”.
– Fuzzy MedicalProperty class: describes and manages the medical observations. It has the following Fuzzy sub-classes: blood pressure, blood glucose,
BMI, heart rate, temperature. These measurements are used as the input
variables to identify the health condition of the patient which is the output
variable. Each fuzzy variable has several fuzzy terms. For example, the blood
sugar glucose variable has the fuzzy sets: (very-low 0–90, low 71–130, medium
125–154, high 142–180, very-high 165–250).
– Fuzzy Health condition class defines the patient’s health condition calculated
based on medical data collected. The health condition is the output variable
determined based on fuzzy input variables defining the medical measurements
and the fuzzy rules. This variable has fuzzy sets “Healthy”, “Moderate” and
“Serious”. The system acts automatically based on the patient’s health condition: If it is healthy, the system indicates to the patient to maintain his
lifestyle. If it is moderate, the system recommends the appropriate drugs,
foods and physical exercises required for the patient to establish his normal health condition and notify the corresponding caregiver to do the regular
health services. If it is serious, the system generates alarms to call the medical
staff and recommends different foods and drugs.
– Fuzzy Food class: defines the food eaten by a diabetic patient. Foods are distributed in meals. According to the nutritionists, a diabetic patient should
maintain a healthy diet that allows him to maintain a normal blood glucose level. The meal eaten is considered healthy or UnHealthy based on the
percentage of carbohydrate PC, protein PP, and fat consumed PF, BMI, the
difference between the calories consumed by the patient and the planned total
calories required for patient’s body defined by nutritionists DCP. The total
calories needed to maintain or lose weight is calculated based on the basal
metabolic rate BMR and the activity level. The BMR is calculated based on
patient’ age, gender, weight and height using Mifflin St Jeor formula [4]. The
nutritionists recommend that the planned total calories should be divided
into the five meals: Breakfast, breakfast, snack 1, lunch, snack 2, and dinner
with the respective percentage: 25%, 12.5%, 25%, 12.5%, and 25%. For each
meal, the number of calories should be distributed in three nutrients, which
are carbohydrates, fat, and protein with the respective percentage 50%, 30%,
and 20%. Therefore, four fuzzy variables are described by this class which are
291
membership fuzzy variables. The fuzzy concepts define concepts and relations
to describe uncertain and vague knowledge. The main difference between fuzzy
and classic concepts is that in classic concepts the membership degree of each
property is equal to 1 or 0 while in fuzzy concepts it is equal to a certain degree
belongs the interval [0, 1]. The classes extended are the following:
– Fuzzy patient class: represents all the information required to supervise the
condition of the diabetic elderly. It describes 5 fuzzy variables which are
height, weight, gender, disease history and age. The fuzzy variable weight
has the fuzzy sets “Light”, “Normal”, and “Heavy” and “Obese”. The fuzzy
variable gender has the fuzzy set “Male” and “Female”. The fuzzy variable
age has these sets, “Young”, “Adult” and “Old”.
– Fuzzy MedicalProperty class: describes and manages the medical observations. It has the following Fuzzy sub-classes: blood pressure, blood glucose,
BMI, heart rate, temperature. These measurements are used as the input
variables to identify the health condition of the patient which is the output
variable. Each fuzzy variable has several fuzzy terms. For example, the blood
sugar glucose variable has the fuzzy sets: (very-low 0–90, low 71–130, medium
125–154, high 142–180, very-high 165–250).
– Fuzzy Health condition class defines the patient’s health condition calculated
based on medical data collected. The health condition is the output variable
determined based on fuzzy input variables defining the medical measurements
and the fuzzy rules. This variable has fuzzy sets “Healthy”, “Moderate” and
“Serious”. The system acts automatically based on the patient’s health condition: If it is healthy, the system indicates to the patient to maintain his
lifestyle. If it is moderate, the system recommends the appropriate drugs,
foods and physical exercises required for the patient to establish his normal health condition and notify the corresponding caregiver to do the regular
health services. If it is serious, the system generates alarms to call the medical
staff and recommends different foods and drugs.
– Fuzzy Food class: defines the food eaten by a diabetic patient. Foods are distributed in meals. According to the nutritionists, a diabetic patient should
maintain a healthy diet that allows him to maintain a normal blood glucose level. The meal eaten is considered healthy or UnHealthy based on the
percentage of carbohydrate PC, protein PP, and fat consumed PF, BMI, the
difference between the calories consumed by the patient and the planned total
calories required for patient’s body defined by nutritionists DCP. The total
calories needed to maintain or lose weight is calculated based on the basal
metabolic rate BMR and the activity level. The BMR is calculated based on
patient’ age, gender, weight and height using Mifflin St Jeor formula [4]. The
nutritionists recommend that the planned total calories should be divided
into the five meals: Breakfast, breakfast, snack 1, lunch, snack 2, and dinner
with the respective percentage: 25%, 12.5%, 25%, 12.5%, and 25%. For each
meal, the number of calories should be distributed in three nutrients, which
are carbohydrates, fat, and protein with the respective percentage 50%, 30%,
and 20%. Therefore, four fuzzy variables are described by this class which are