Another, very important, part of this
project is developing a method of calibrating the meter to different
strip profiles and different lots of the same strip.
For a bare
minimum, we need 2 known glucose control solutions – so we can
determine the slope and intercept of the strips response (linear
regression – fairly simple algebra)
. We need a bare minimum of 2 points to define a straight line (y=mx+b). I initially believed that
this would not be an issue since their must be standard glucose
control solutions on the market for calibrating glucose meters. As it
turns out this is not true, something that
M. Bindhammer figured out a few weeks ago developing
his
Open Source Arduino Blood Glucose Meter Shield (another 2016 Hackaday Prize entry that is going into the finals).
After a bit of research (journal articles, patents, msds sheets – you can check out my reference post to access them), I found that these control solutions are made up of the same basic components:
- Water (major component)
- Viscosity enhancer (thickener)
- Preservative (antimicrobial)
- D-Glucose (Dextrose)
- Color (to visualize filling, does not appear to have any other function)
Patents and Pharmaceutical compounding trade journals provided lists of both viscosity enhancers and antimicrobials.
Viscosity enhacers (not an exhaustive list) :
Natural Hydrocolloids
-
- corn starch
- potato starch
- arrow root
- acacia
- guar gum
- xantham
- amylose
-
Other Enhancers
- hydroxyethyl-cellulose
- hydroxypropyl methyl cellulose
- PEG (poly ethylene glycol)
- Polyvinylpyrrolidone
- Glycerine
Antimicrobials (not an exhaustive list):
- Sodium Benzoate
- Benzoic Acid
- Benzyl Alcohol
- Biguanides
- Chlorocresol
- Chloroxyylenol
- Phenolic compounds
The antimicrobials seem a bit tricky since each one is most active within specific pH ranges. If we are mimicking human blood, pH range from 7.35-7.45 (slightly basic), this limits our choice of antimicrobials.
Need to determine the pH of glucose solutions with viscosity enhancers (look at pKa values and also check with pH meter). Look at pH of commercial solutions.
The makeup of a few commercial glucose control solutions:
Suresign Resure Glucose Control Solution
- Glucose
Normal Control Solution:
- 0.12(w/w)% D-glucose
- 99.88(w/w)% Non-reactive ingredients
McKesson TRUEcontrol® Control Solution
- Water
- D-Glucose
- Viscosity enhancing agents
- Inorganic Salts
- Amaranth
- Preservatives
One Touch Ultra Control Solution
- water 98%
- polyvinylpyrrolidone 8%
- amaranth <1%
- Benzoic acid Na salt <1%
- EDTA <1%
- D-Glucose <1%
CareSens Glucose Control Solution
- Water
- D-Glucose
- Preservative
- FD&C red dye
- Viscosity adjusting agent
These are a few of the only manufacturers that have actually listed ingredients and you can see that only a few of them actually provide any idea of concentrations. Of note, while all control solutions I've looked at state that they are only reliable on a single brand of strip - the formulas for making these solutions are essentially identical in how they are made and the base ingredients used. (Good news for me!!!)
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
Oops, my error. I was actually thinking linear equation (y=mx+b) not linear regression. I'll add this correction to my log.
Are you sure? yes | no
Hi Tom,
If you have only a pair of coordinates, you can't use linear regression. For linear regression you need at least 3. The more the better. I would also not call linear regression 'fairly simple algebra' (https://en.wikipedia.org/wiki/Linear_regression) :)
PS: If you need ONETOUCH Ultra strip connectors, I would be happy to send you a bunch.
Are you sure? yes | no