Requirement Blood glucose level (BGL) for normal 1.2 BGL

Requirement Analysis

Initial Situation and Objectives

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

 

This section of document explains the problem statement of
project and some of the globally accepted terms. It also covers the current
research done on symptoms of the problem statement.

Insulin pump therapy.

Insulin pumps are designed to work in a smart manner based
on logic controls, the prime function of insulin/glucagon pump is to inject
precise dose of insulin as per body requirement. Insulin pump mainly consists
of followings 1

1.   
The reservoir for each insulin and Glucagon.

2.   
From the pump’s reservoir, insulin/glucagon is
infused into the human body through an infusion set.

3.   
The infusion set is implanted on the human body
and is infused through a tiny flexible tube called a cannula that sits just
underneath the skin in dermis layer. 1

4.   
The infusion set is then connected to the
reservoirs through a small flexible tube which is detachable on requirement.

5.   
Buttons are to configure the device and navigation
through the menu.

6.   
A graphical user interface to get alarms, set
program, get recorded history and critical logs

Functional requirements

1.   
Simulation of diabetic human body glucose level

Simulate the scenarios of
normal, hyperglycemia and hypoglycemia conditions. 2

1.1        
Blood glucose level (BGL) for normal

1.2        
BGL with fasting i.e. before meal (80 to 130
ml/dL)

1.3        
BGL without fasting i.e. after meal (less than
180 ml/dL)

1.4        
Hyperglycemia BGL ? 200ml/dL, Hypoglycemia BGL ? 70 ml/dL.

2.   
Controller

it
performs all decision-making computation for dosing of insulin/glucagon into
human body, report generation, historic log records and system functionalities
as per following. 1

2.1        
Calculate the dosing based on mathematical
model and inject accordingly.

2.2        
Maintain logs / history for GUI and analysis

2.3        
Raise alarms when -insufficient reservoir
level, hyper/hypo glycemia, no
acknowledgment from patient (raise SOS alarm to caregiver), alarm for exceeding
maximum dosing limit per day.

2.4        
Save configurations settings for patient
profile.

2.5        
Act proactively with highest priority in
critical conditions of hyper/hypo glycemia.

2.6        
Functionality to work in auto manual mode

3.   
Continues glucose monitoring system (CGMS)

3.1        
It monitors the BGL every 3 minutes for simulated
body with effect of injected insulin/glucagon

4.   
Graphical user interface.

There
shall be 3 graphical user interfaces – Doctor, patient and caregiver

Common functionality for all users.

4.1        
Mode selection based on configuration

4.2        
Plot graphs and update every 3 minutes or event based.

4.3        
Graphical data read access, alarm monitoring/acknowledging.

4.4        
Display level of reservoirs, number of injected units,
date and time.

Additional functionality view based.

4.5        
GUI for
doctor – Configuration
window with user authentication, patient profiling, dosage settings, maximum
dosing limits, auto/manual mode settings.

4.6        
GUI
for patient –
Emergency action button (SOS).

SAFETY REQUIREMENT

BGL set limits 3

1.1        
BGL with fasting i.e. before meal (80 to 130
ml/dL) à Normal

1.2        
BGL without fasting i.e. after meal (150?BGL?180
ml/dL) à Action required

1.3        
Hyperglycemia BGL ? 200ml/dL à Unsafe à insulin injected

1.4        
Hypoglycemia BGL ? 70 ml/dL à Criticalà Glucagon injected

1.5        
Software Interlock à single reservoir selection
for pumping at a time.

1.6             
Raise alarm based on conditions defined in
functional requirements.

Requirement Negotiation

We have referred SWEBOOK V3.0 to finalize our
requirements, considering first draft of requirement, the estimated cost for
project was approximate 14
months, on negotiating and revising the second version reduced the cost
of project to approximate by 3.5
months.

Mathematical model

The simulation of blood glucose
level (BGL) was little complicated because BGL not only depends upon just food
intake, it depends upon very complex dependency of different hormones which are
neuropeptides released mainly from the pancreas, intestine, liver and brain 4. They are also
affected by other factors as well. Few of them are like stress level, timely
intake of diet and life style. We have referred many mathematical models such
as neural network (NN), compartment model (CM), hybrid of CM & NN model 5, rudimentary model
of glucose to Stress 6
 but it was very difficult to implement
such complex models in such a short time. We have chosen model 7, it is
relevantly simple model to simulate BGL time dependent behavior. This model
uses the first order differential equation based on initial value of BGL,
variable time, rate constant carbohydrate digestion and rate constant insulin
release. This model was formulated based on the practical data sample collected
from a person.

The
kinetic of reaction for carbohydrate digestion over time is given by following
equation.

                                                               (1)

and
rate of change of BGL over time is given by following equations

               (2)

Whereas:

C(t)                  : Concentration of
Carbohydrate over time

G(t)                  : BGL over time.

C
(0) & G (0)   : Initial condition of
carbohydrates and BGL respectively at t = 0

K1
                   : Rate constant
carbohydrate digestion considered as glycemic index (GI)

K2                    : Rate constant insulin
release

t                       : Time

 

Solving
the first order differential equation (1) & (2) give us following equation.

 à      à                                                    (3)

                                                          (4)

Whereas,
for some constant d,

Equation
(4) gives us the time behavior of glucose over time ‘t’. Assumptions:

A
(0) =C (0): 121.7; G (0): 90, k1: 0.0453, k2: 0.0224

On
Simulation we got the following output as graphical in comparison to the model
data.

                                 

Figure
1: Mathematical model reference data 7                              Figure
2: Simulated model data

 

************************ Elis
editing***********

Human Machine Interface (HMI)

We
have used java FX and scan builder tool to design the HMI for the simulation.
There are 3 views designed, Patient view, Care giver view and Doctor view.

Patient view:

In
simulation mode the simulated data can be viewed in patient view, it provides
the information about the level of reservoirs, graphical representation of BGL
over time, historical data, send emergency signal button (SOS) and feature to
switch between auto/manual mode based on configuration.

Care giver view

This
is similar view as of patient view but with no SOS button functionality, it can
switch to auto manual mode without restrictions but with authentication

Doctor view:

This
is the view designed for doctor, here the configuration for dosing and patient
information can be stored. Only doctor has the rights to configure the device
functionality including giving rights for auto/manual mode. This protected
screen can be only accessed by user authentication.

 

             

Figure
3: Patient                     Figure 4:
Care giver                            Figure
5: Doctor

 

 

 

Task

Responsibility

Documentation

 

Requirement Analysis

Baldev Raj Barrsiwal

Mathematical Model

Baldev
Raj Barrsiwal, Elis, Raul

Human Machine
Interface Designing

Baldev
Raj Barrsiwal, Elis, Raul

Project Scheduling

 

Cost Estimation

 

Engineering Process

 

Implementation

 

Hazard Analysis

 

Test Cases

 

Testing

 

 

 

 

 

 

List of Literature

1        Insulin
Pump

2        Continues
glucose monitoring

3        American Diabetes Association

4        Pancreatic regulation of glucose
homeostasis

5       A Real Time Simulation Model of
Glucose-Insulin Metabolism for Type 1 Diabetes Patients S. G. Mougiakakou, K.
Prountzou, K. S. Nikita, Proceedings of the 2005 IEEE, Engineering in Medicine
and Biology 27th Annual Conference, Shanghai, China, September 1-4, 2005

6       Rudimentary Model of Glucose Response to
Stress, Nasha Rios-Guzman, University of South Florida, Undergraduate Journal
of Mathematical Modeling: One + Two, 2017

7       Blood Glucose Levels, Carlos Estela,
University of South Florida, Undergraduate Journal of Mathematical Modeling:
One + Two, 2011, Volume  3, article 12