Digital Pregnancy/Drug/Diagnostic test

Status
Not open for further replies.

Scott Bruton

New Member
Hi All

I work for a medical diagnostic company which develops drug/malaria/HIV/etc diagnostic tests. I have been tasked with the role of learning and developing a digital display to mitigate human error when reading the results.
I was guided to look at the ClearBlue pregnancy test (with LCD display), however, their PCB design and schematics are obviously Protected and cannot be found online. However, this is basically the system that I with to create, but for my company purposes.

This is the clear blue device:

This is how it the PCB looks inside:


Could anyone please help me out, and point me in the direction of a URL or tutorial (or give me some general advice) on how to create such a system so that I can begin the learning process. Any help would be appreciated.

Kind Regards
Scott
 
Whats your outlook, do you want to design & build a prototype from scratch or are you project managing this.
A significant amount of electronics and programming knowledge is required to make something like this fit for production.
If your just looking to make a rough prototype then consider something like an arduino and a generic Lcd.
I cant help you with the chemic side of things, or whatever process is used to detect 'a bun the oven'.
 
That device is based on a custom designed integrated circuit under the black epoxy blob. this is called chip-on-board manufacturing, and is common in simple, high-volume, low-cost devices. I'm not trying to be discouraging, but this is not a simple design or manufacturing system, and is way beyond what can be taught on a web page.

Re-reading you question, are you asking about only the display?

ak
 
Call Kyocera Displays (Plymouth, MI) or Pioneer or Sharp or one of the hundreds of Chinese shops that make custom LCD displays.

They will make a custom unit for you.
The CLear blue easy is a simple display with one field for each word with the word PREGNANT always on - so only 4 possible states ( 2 bits) required to interface with the display driver and 5 segments to control.
0) "Not"
1) " 1-2"
2) "2-3"
3) "3+"
4) Pregnant

Further evidence this is a custom 2-bit display
- notice that the letters are crystal clear/razor sharp (no pixelation)
- the 2 is not recycled for the "1-2" vs the "2-3" image. Those are in unique fields to minimize the umber of states that must be controlled.

if you want to kill some time, here is a video...
Note the price of this rather large LCD with 212 sections vs the CBEZ with only 5 segments - the big display was under $3 in a 1000 piece run. The millions made by CBEZ and they are less than 1/8 the size, the price is likely under $0.25 each.
also Phoenix displays is a good option.
 
Last edited:
Thank you all for your quick and efficient responses.

I feel I should be a bit more clear on the project.
The display is for a virus (HIV/Malaria/Flu) test kit, which we already produce inhouse and distribute.
We have seen a issue with human error, when people read the 1 or 2 lines (similar to a pregnancy test) that indicates positive or negative for the virus.
We wish to attach a small display that mearly displays Positive or Negative (Maybe just a + or - sign) to mitigate this human error.

I felt that I should look into pregnancy tests because devices such as Clear Blue has successfully done this, however, obviously for a different purpose.

With that being clarified, I would like to answer all of your questions individually:

dr pepper:
- I will be building this from scratch, at first as a prototype, but then into mass production.
- Luckily I do have quite alot of programming and electrical knowledge. I have a diploma in software development, a honours in mechatronics and have built multiple Arduino based production machines, which has given me have many many hours of programming experience. However, my electronics knowledge has been more focused on certain projects (Very DIY, make motors move from here to there sort of stuff), and I have never built a circuit similar to this one.
- I'm assuming that I would have to measure a certain element within the blood (or urine) using an analog sensor and feed that into a small micro controller and display it onto a small screen.

atferrari
- Im unsure what you mean by implement the test, so I will answer a summary of what I need to do.
- The test stick needs to be subject to blood or urine, and the material will absorb this input, sensor will read the concerntation of an element, and the display must output the results.
- We already produce the test kits which we distribute.

AnalogKid
- The test kits we sell do not have any circuitry. Therefore, I would need to prototype, or outsource a company to develop, this circuity to read the test and output it to the display.

gophert
- Thank you for your suggestions, I will most certainly look into them.
- I really appreciate the video and help.

I hope I have answered all your questions successfully.

The feeling I get, is that this is quite a bit project, that requires integrate knowledge about micro-electronics etc. If this job is too big a learning curve and will take me too long to master, I would rather have it outsourced to have the product up and running as fast as possible.
If any of you have any suggestions of companies that would take on this project for mass production ( such as gophert has kindly given) please could you let me know. All help will be appreciated.

Kind Regards
Scott Bruton
 
Last edited:
Assuming it's a test strip similar to blood glucose strips, do you have any details on the peripheral (socket) that the strip plugs into? As you (your company) already makes a meter then you must have details of how the test strip is readable. Can you supply more information about the actual test strips and interface?

Mike.
 
Pommie

Thanks for the response Mike.

I officially start at this new company tomorrow.
I believe in prepping for things, therefore, this is initial research and work so that I can hit the ground running.

The company makes test kits that look similar to the following:


Therefore, its more of an absorbant strip, that gives either 1 or 2 red lines on the output strip.
No digital display at the moment.

Scott
 
I have little knowledge of virology, however I understand a virus is a micro organism and I suspect that testing is done a bit like testing water samples for cooling towers where you take a sample and incubate it over a period of time, then use a dye to show up colony forming units.
Not sure how you would do this over a period of minutes, maybe you'd use some kind of enzyme or something.
Sounds like maybe a corona virus machine, that would be very popular right now if you got it to work reliably.
For the likes of us on this forum a micropower system with an Lcd is fairly straightforward, for a prototype I'd probably use a standard 65 x 128 I2c lcd panel, in fact to keep costs down this might be ok in the final unit unless your making 10's of 1000's.
I'd recommend you start by looking into an 8 pin microcontroller, like a Attiny 85, possibly for the final version a padauk (spelling), as they can be had for pence.
 
If all you need is a plus or minus sign, you can get that from one seven segment display digit or LEDs.
The electronics shouldn’t be too complicated. Minus should come on when test is finished. Extra segments come on to make it a plus if need be.
 
Antibody

A "plus sign":from a 7-segment digit? How does that work?
 
For a simple "yes"/"no" indication wouldn't two differently coloured LEDs do the job?
 

Until her contract ran out in January, my daughter was working at the University Of Twente, as part of a multi-national and multi-university team trying to develop 'virus detection techniques via nano-technology' - but I've no idea how they were approaching it?, but there were physicists and biologists as well as chemists (and god knows who else).
 

The whole project probably made sense until they hired an accountant - then the contracts were no longer renewed.
 
The whole project probably made sense until they hired an accountant - then the contracts were no longer renewed.

Not at all, you obviously have no idea how Dutch Universities (and MANY places of work there) operate - commonly you only get a maximum two year contract, at the end of that time you can't get a new contract for 6 months, then you can start a new two year contract. So it's quite common, two years at work, six months off - and repeat - certainly her job paid more than enough to fund six month work breaks.

I presume it's something to do with contract rights, if you're there longer then two years or so it's possibly considered you're permanent?.

As it happened her contract was extended by three months, I don't if that was because of the virus work, or her teaching duties?.
 
The tests you provided seem to be purely mechanical, which means the struggle would be in creating/obtaining a sensor, which you would likely need a different one for each test you wish to conduct. My understanding of pregnancy tests at least is that it is based on the levels of a certain protein, and you are likely pregnant if your levels are above a certain amount. I would recommend either a reusable sensor (like diabetic tests) or a modular design so the main unit can be reused to reduce waste.

Your sensor will likely be analog, so you could just put one if/else statement in an arduino and output to LEDs (green - positive, red - negative)? That way you wouldn't have to deal with the code for a display. Then 3D print a housing with either a port for a urine-soaked sample or the sensor itself (I don't know what the sensor would look like, but I'm imagining something shaped like a pH probe), and just have two LEDs sticking out. Easy, cost effective, and small. Perfect for product scaling.
 
Last edited:
Klein1415

Thank you for your feedback.
I appreciate the advice.
The only problem using colour LEDs to indicate positive or negative is that you need to cater for everyone, including those with severe colour blindness. This could cause large human error within a certain group of individual.

all
So, I have been thinking and researching, and there is a potential way of sensing the concentration of a certain substance within an area is by shining an LED of a certain light wavelength, onto the substance and then reading the wavelength of the light emitted by the substance.

Does anyone of you know of a micro lightwave sensor that can read with in the nm range of light?
 

just use a reverse-biased LED of the same wavelength (cathode connected to Positive voltage,) and add a resistor of 10k to 100k ohms between anode of LED and ground. Measure voltage from ground to LED anode monitor absorption delta at various concentrations. Watch for ambient light bleed for noise.

Also, signal saturation can be a problem (if too much current flows, the IxR voltage will be near or exceed your positive voltage source and your signal will be cutoff. Use a lower value resistor.

if you are not seeing much change when sample is present or blank is present, you'll need to use a bigger resistor to get more dynamic range (but not too big or you'll saturate as above paragraph describes).

make sure to use a high impedence ADC to maintain linear response. Works quite well (do not use it with white LEDs with yellow phosphor (or any phosphor). Also, the sensing LED must be of equal or longer wavelength than the emitting LED. This setup is capable of extremely fast response times as you are making a photodiode sensor (faster than phototransistor sensor).
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…