- The AI software can predict the chemicals released by cold plasma used in medical treatments.
- Researchers at Princeton and GWU work together on this important project.
- Machine learning and physics work together to improve how we use and understand cold atmospheric plasma.
A special computer program that can think on its own has been made to help improve medical care using a special type of electrified gas called cold atmospheric plasma. The computer program, made by scientists at the Princeton Collaborative Low-Temperature Plasma Research Facility (PCRF), uses both machine learning and physics to guess what chemicals are in the emissions from CAP. This discovery has big effects on treating cancer, helping tissues grow, and cleaning surfaces.
Using AI to estimate air pollution from cars
The software, called a physics-informed neural network (PINN), was able to predict the different chemicals released by CAP jets. It did this by looking at information from real experiments and using the basic ideas of physics. This new way of using AI, called machine learning, helps the system get better at making predictions by using the information it’s given.
Cold atmospheric plasma: A versatile tool
Cold plasma has been used in different medical ways, like killing cancer cells, helping wounds heal, and getting rid of bacteria on food. But scientists don’t completely understand how these effects work.
Yevgeny Raitses, a research physicist in the U. S, says The AI software from PPPL helps us understand CAP jets better. This new knowledge could help make CAP technology work better in medical treatments.
Working together with others to achieve a common goal
The project was a team effort between researchers at PPPL and George Washington University, working together at the Princeton Collaborative Low-Temperature Plasma Research Facility. PPPL is very good at studying plasma and now they are also using AI in areas like medicine and making things.
Sophia Gershman, who is a top research engineer at PPPL, highlighted how hard it is to figure out the chemicals in CAP jets because we have to think about how they interact very quickly on a very small scale. Machine learning can help with complicated problems by making very accurate calculations that were before almost impossible.
Making information and teaching
The project started with a small set of data collected using a method called Fourier-transform infrared absorption spectroscopy. This first set of information was used to create a bigger set of data. Using the idea of natural selection, a computer program was used to teach the brain-like computer system. The AI system got better at its job by choosing better data and making more accurate predictions over time.
Precise math for CAP Jets
The team made a software that can accurately figure out the amounts of chemicals, gas temperature, and electron temperature in cold plasma jets. This accomplishment is very impressive because CAP jets can have really hot electrons while keeping other particles at almost room temperature, which is good for medical treatments.
Customized plasma therapy coming soon
Michael Keidar, who is a teacher of engineering at GWU, talked about how they want to use real-time calculations to make medical procedures better. Keidar is making a new medical tool that can change to fit each person’s needs. By keeping an eye on how patients react and using advanced technology, the device can change the plasma settings to make it work better.
This study looked at what the CAP jet is made of and how it changes over time, but it only looked at one specific spot in space. In the future, more research should look at different points along the jet’s output stream. Also, it is important to include the surfaces treated by the plasma in the analysis to understand how it affects the chemical composition at the treatment site.
This important study was paid for by the U. S The Department of Energy and the Princeton Collaborative Research Facility are working together to find better ways to treat illnesses using cold atmospheric plasma. By using AI, doctors can make personalized plasma treatments better, which gives hope for improved healthcare.
Warning: This is not a recommendation for trading.