Modelling Design
To first get a feel on what the modeling software should be like, we had to ask our mentors and the wet lab on how they would rather have it since they will be using it afterwards.
Then we got started on working on the diagram of our model, that way we will have a plan and an algorithm already set in place before we start coding the whole simulation. The diagram was really helpful and it made the coding easier and faster since we didn’t have to reiterate on the code multiple times.
From the diagram we only had to extract the information related to the interplay of the parameter on CAR-T cells and Tumor cells. It was easier for us to work on only one case at first, modeling one CAR-T cell, and make sure it is working just fine. And since we were working alongside the results of the “CARTmath—A Mathematical Model of CAR-T Immunotherapy in Preclinical Studies of Hematological Cancers” [1] research paper, we matched our parameters value to their parameters to make sure the code is working properly. Since we got the same results as them then it was okay for us to continue to our next task and write the code for the second CAR-T cell.
Now in order for us to model the effect of injecting two CAR-T cells at a time we had to modify the Tumor equation so it will implement the feature we’re aiming for. So going back to the equation used in the research paper and looking at the modified diagram we had, the change we had to do wasn’t that drastic and it was a straightforward way to implement it.
Fortunately for us after going through all these steps in MATLAB and getting the desired output and making sure it was working. We decided since we already have the algorithm and the way to code it, to migrate to Python since it will be easier for us to build a user interface using python, as well as we wanted to have the simulation up and running on our Wiki and Python gave us the option to do that.
After migrating to Python, we wanted to implement another feature for the modeling and that is changing one parameter’s value and make change between a defined range that will be imputed by the user, and then the Python script will generate a graph showing the change of tumor cells number following the change of the parameter’s value in the predefined change.
We thought about this feature since it will help the wet lab in seeing how a slight change in the parameter can have an effect of the tumor cells number, from this they will be able to pinpoint the important parameters or the parameters that have the most effect of the tumor cells number, and do numerous studies on them so they can engineer a better more effective CAR-T cell.
Hardware Design
To get the Thermal method up and working we had a lot of research to do on various topics.
So we had the following questions asked:
Where to get the laser ?
Will the laser end up burning the test strip and kill the Gold Nanoparticles if it got really hot ?
Will the angle of the laser change the results ?
Will having Gold Nanoparticles attached to other cells be a problem for the result accuracy ?
Does the device need to be made of a certain material or color for maximum efficiency ?
And to answer all these questions we had to go back and talk to Dr. Gurnam and Dr. Zazubovits. Since getting your hands on a class 4 laser for such a project is a bit hard and expensive, we asked Dr. Zazubovits if we can use his laboratory for our experiments since he has the laser we are looking for, and we got access to a workstation at his lab, which answered our question on where to get the laser.
To see if the laser will end up burning the test strip and damaging the samples we have, we ended up 3D printing a small square box where we made some measurements. The box will hold the piece of paper in question, and we will be able to slide the test strip on the x-axis away from the laser at any time and then save the distance. This should give us the right distance between the laser and the paper, where we will have the output power we want and no damage will be done to the paper.
Noting that doing this, the laser is shooting horizontally and the paper is in a vertical position. When the whole device is finished we will have the laser set to an angle and the paper will be laying horizontally in the device. That change of angle for the laser won’t change much of it’s accuracy but it may introduce some minor errors that can be neglected.
Now since the whole experiment consists of us hitting the Gold Nanoparticles with a certain light frequency to excite them, will having the Gold Nanoparticles attached to other cells cause a problem or change the frequency of light they should be hit with? And the answer that we got is no, it will not be affecting us in any way as long as the other cells are not photoreceptor cells.
After making sure all these questions were answered, it was time to start 3D modeling the end device that will hold everything in place.
And we decided to 3D print the device in black color, since black color will limit the amount of light reflected. White color will give the light the ability to bounce back to other surfaces and that could cause reading errors from the IR sensor if the laser ends up being reflected on it. So black was the best and our safest option to go with.
Software Design
Now after we got the device made and all the sensors and microcontroller in place. It’s time to start working on an electronic diagram for the circuit. This step was pretty straightforward and easy to do since we already know what we need from our circuit, and there exist some libraries on github that help us code the IR Sensor functionality.
The IR Sensor used is : MLX90614ESF-BCA-000-SP.
The Microcontroller is : ESP32 with Wifi module.
We added another IR sensor for the circuit only for the purpose of turning it on/off with a remote controller, since we don’t want to constantly be reading the heat emitted, and getting close to a laser working area each time to turn off the circuit manually is dangerous.
References
- Barros, L. R. C., Paixão, E. A., Valli, A. M. P., Naozuka, G. T., Fassoni, A. C., & Almeida, R. C. (2021, June 11). CARTmath-a mathematical model of CAR-T immunotherapy in preclinical studies of hematological cancers. Cancers.
- Qin, Z., Chan, W. C. W., Boulware, D. R., Akkin, T., Butler, E. K., & Bischof, J. C. (2012, April 27). Significantly improved analytical sensitivity of lateral flow immunoassays by using thermal contrast. Angewandte Chemie (International ed. in English).