An approach towards non-invasive blood group detection

Thread Starter

nhel_101

Joined Apr 5, 2020
2
I am making a project about Opt101
A simple experiment that measures the voltage level when I put my finger between the led and the photodiode. I attached a diagram for you to figure out I sm sayin
And can someone give me the right code for these. Thank you in advanceScreenshot_20200405-154700~2.png
 

WBahn

Joined Mar 31, 2012
26,398
Is this homework (or any kind of academic assignment)?

If so, what are the actual requirements of the assignment?

You will also find that a much more specific thread title will garner much more interest; variations on "help me" make lousy titles and most members will simply ignore the thread. Come up with a better one and I can change it for you.
 

Thread Starter

nhel_101

Joined Apr 5, 2020
2
AN APPROACH TOWARDS NON INVASIVE BLOOD GROUP DETECTION

Hi everyone can someone help me to do this project. I want learn how they did the circuit and the code for this project. Thank you
 

Attachments

ci139

Joined Jul 11, 2016
1,696
it's a weird case you have https://acadpubl.eu/hub/2018-119-15/1/55.pdf
both those doc.-'s reference list is somewhat "weird" - no much advancement since 2018 ? -- further more one would expect some calibration for given (likely precisely chosen pair of iR emitter/detector ← there are variations inside the patch . . . for both . . . )

all this really much does not sound correct ??

+ even if the hardware side is ok . . . one would expect a statistical miracle formula for even moderately reliable results . . .
"An exception confirms the rule" -- it might be such works - as easy as presented - which requires sounding out such bands that are unique - the simple scattering of light does not let you tell - what they propose by a simple light power measurement . . . or at least not in any way i can imagine
 
Last edited:

WBahn

Joined Mar 31, 2012
26,398
It looks like a pure junk paper and I can't find any rankings for the journal (but I might just not be looking in the right place), so that doesn't bode well. The paper is very poorly written, makes pretty grandiose claims that it doesn't back up, and provides no meaningful test results. It looks like the "data" they are basing their results on is just one student for each blood group and the only statistical analysis they did was to average some readings (how many? 2? 2000?) for each student.

As for how they did it, they read the voltage at one of the inputs of the Arduino and then used some undefined ranges to pick a blood group and displayed that group along with the voltage reading.

They give NO information about how they configured the OPT101. Instead, as with every thing else, they just copied some information from the marketing blurbs for each of their components.

They appeared to assume that the voltage output from the detector is completely independent of the LED used or how much current it is being driven with or what its spectral distribution is. At the end they throw in the remark that factors like skin color and finger thickness should be considered prior to testing, but say nothing about HOW those factors should be considered, almost certainly meaning that they made no effort to actually see if they had any effect at all and are just throwing out what comes to mind to make it sound like they actually did something.
 

WBahn

Joined Mar 31, 2012
26,398
it's a weird case you have https://acadpubl.eu/hub/2018-119-15/1/55.pdf
both those doc.-'s reference list is somewhat "weird" - no much advancement since 2018 ? -- further more one would expect some calibration for given (likely precisely chosen pair of iR emitter/detector ← there are variations inside the patch . . . for both . . . )

all this really much does not sound correct ??
Looks like another junk paper. There's also so much verbiage that is common that it is a clear ripoff, despite no reference being made by the later one to the earlier one.

Also notice how the results from the two papers are not even in agreement regarding the ordering of the groups. The early paper has the order (low voltage to high voltage) of {O,A,B,AB} while the later paper has the order {A,O,B,AB}.

1586085581797.png1586085608908.png
 

LesJones

Joined Jan 8, 2017
3,654
There was also no suggested explanation of why different blood groups would be more or less transparent to white light. If it measured the absorption of specific wavelengths of light (In the same way as an IR or UV spectrophotometer does.) then I think it MAY be possible to make it work.

Les.
 

jpanhalt

Joined Jan 18, 2008
11,088
The basic blood group antigen is called H. That antigen is not modified in people with Type O blood. Type B have galactose added. Type A have N-acetylgalactose amine added. And Type AB people have H-antigens that can be modified with either Type-A or Type B epitopes. Here's a cute interactive site: https://www.chem.wisc.edu/deptfiles/genchem/netorial/modules/biomolecules/modules/carbs/carb8a.htm

One would not expect those changes to be detected with crude visible or IR spectroscopy. Even with whole blood and narrow bandwidth spectroscopy, there will probably be so much background interference that blood groups cannot be identified.

Agree with others in doubting anything clinically useful will come out of this. However, the TS may learn some basic electronics, programming, and spectroscopy skills from this project. That is unlikely to happen if he is given detailed procedures to that end.
 

WBahn

Joined Mar 31, 2012
26,398
It's a project that could make for an interesting undergraduate project for a small interdisciplinary team. Perhaps an EE, a math major, and someone from biosciences. Take good data with good analysis and even if the answer is that it's not likely to be useful, that's a perfectly fine result.
 

ci139

Joined Jul 11, 2016
1,696
there must be a comprehensive initial study - for the blood spectral response - that i have not spotted yet . . . if such exists - it must be found first - so you'd knew what you are aiming for

also a statistics that allow you to conclude anything by proposed method of measurement ... combined with the detailed description of the measurement it-self

the claims they make without the ↑previous↑ - bare no significance alone

as for your request - i have no time nor much knowledge about anatomy nor much academic access rights to dig it all out

basically the finger here acts as a complex light filter . . . that may or may not have a signifficant "statistical pass window" for the data of your interest ← simple as that
 
Last edited:
Top