Reached the funding target!
Due to everyone’s generous support this project can proceed! I am honored for the privilege of being able to contribute in some way during these trying times.
---------------------------------------------------
Thanks to the support we received from you, we were able to achieve our goal within eight days of the start!
We highly appreciate your support for this project!
Since we have achieved our initial target, we can proceed with the research project.
In addition, we set the second goal at 1 million yen because we still have time to complete the crowdfunding challenge.
This second goal will be used to cover the cost of installing a higher-performance CPU, memory, and GPU on the computer we originally planned to purchase.
By using more sophisticated computers, large-scale simulations are possible to generate more data with higher precision than originally planned.
We would like to ask for your continued support for this project.
Comment from academist staff
Using deep learning on COVID-19 as an example
academist Shimojo
Why is detection of viruses in air important?
Viruses are known as pathogens that cause a variety of infectious diseases, such as influenza, measles, and rubella, as well as COVID-19, which is currently endemic around the world. Viruses have caused the suffering of many people in the past and today. Some viruses, such as COVID-19, are known to be transmitted through air via droplets and aerosols due to people coughing, sneezing, speaking, and singing. The presence of a large number of viruses in the air increases the possibility of infection. So the rapid detection of viruses is important to prevent infection and prevent its spread especially at airports, train stations, sporting events, and other places where many people travel.
Thermal scanners are used to detect viruses. However, because people can shed viruses even when they don’t have fevers, this does not reliably detect infected people. One method for detecting viruses in air directly is to collect samples using an air filter and analyze them by PCR. However, detection takes time. Biosensors have been developed, which can detect viruses in real time and show real promise. However, this will require a large number of sensors to detect viruses over a large area.
Possibility of detecting airborne viruses by LIBS tested with COVID-19
I focused on Laser Induced Breakdown Spectroscopy (LIBS) as a method for rapidly detecting viruses in air. LIBS is a quick method for analyzing the elemental composition of various substances, including airborne substances. Ultra high power terawatt lasers used for LIBS can generate plasma at specific distances (up to a few 100 meters), allowing detection over a wide area.
LIBS first applies a high power laser to the material to break it apart at the atomic level. Fragmented atoms generate plasma during ionization, which emits radiation with spectral lines characteristic of the atoms which have been ionized. The element composition of a substance can be determined from the spectrum obtained by separating the light into wavelength components in a spectrometer.
It has been shown that the spectral patterns obtained by LIBS can detect viruses contained in liquids. However, it is not known whether LIBS can detect viruses in the very small quantity of small droplets that are able to float in the air, either in airborne droplets or in the same fluid.
The purpose of this project is to theoretically verify whether LIBS can detect airborne viruses through numerical simulations using deep learning. As a specific example of theoretical verification, we will consider and analyze the possibility for COVID-19.
Utilizing Deep learning for theoretical verification
Numerical simulations use deep learning to optimize the detection of viruses by LIBS. Deep learning is a widely used technique in situations where analysis of large amounts of complex data is required. Deep learning involves building an artificial neural network on a computer and training the network using a large amount of data to solve a particular problem. Deep learning, for example, is used in a variety of contexts, such as Go and colorizing black and white photos and films.
We will use deep learning in two aspects of this research project.
The first is spectral generation simulation. In addition to viruses, there are a variety of other substances in the air, such as volatile organic compounds. Therefore, when detecting viruses, it is necessary to determine whether the plasma spectrum generated in the actual air is characteristic of the air containing a particular virus.
Therefore, we train an artificial neural network using spectral data generated in the air of various material compositions to determine the spectral characteristic of the air containing COVID-19, and determine whether it can be used for virus detection. In addition, these results determine the conditions under which the plasma is generated so that the optimal spectrum for detection can be obtained.
The second is ultra high power laser propagation simulation. This simulation calculates the plasma generated from various laser pulse shapes and using this data we will train an artificial neural network. This determines the optimum laser pulse shape to produce a plasma with a specific length, distance and temperature required for detection.
Request for Research Funding Support
Funds obtained through crowdfunding are used for simulation. Laser propagation simulations and spectral calculations require large memory and a personal computer with a high-speed CPU. Deep learning computations also require large memory and a large number of GPU cores. The rest of the funding will be used to present the results of the research project at academic conferences and in scientific papers.
My research is currently concentrated more on theoretical physics. In the past I worked on air breakdown using lasers. So I thought that this could be a way to detect the virus. The goal of this project is to theoretically verify the idea.
Although this research is quite challenging, I hope that maybe it can contribute to the current COVID-19 pandemic in some way. One day, we will continue this research with the dream of developing possible devices for virus detection using ultra high power lasers.
＊＊＊
Contact Information
National Institutes for Quantum and Radiological Science and Technology
Innovation Center, Research Promotion Section
E-mail: kifu@qst.go.jp
＊＊＊
Profile
Koga James
Project timeline
| Date | Plans |
|---|---|
| April 2021 | Challenge Crowdfunding |
| July 2021 |
Examination of simulation method
|
| July 2022～ |
Start of simulation
|
| July 2023～ |
Data Analysis, Writing
|
Pledge Rewards
I will summarize the results of this project in a report and send it as a PDF file around December 2021.
Research report
12 supporters are supporting with this reward. （No quantity limit）
Your name will be listed on the research report as a funder.
Acknowledgement in the research report / Research report
16 supporters are supporting with this reward. （No quantity limit）
I invite you to an online science cafe. I would like to discuss with you high power lasers. The event is scheduled to be held around August 2021.
Online Science cafe / Acknowledgement in the research report / Research report
15 supporters are supporting with this reward. （No quantity limit）
I invite you to an online seminar on the possibility for virus detection in air using high power lasers. The event is scheduled to be held around November 2021.
Online seminar / Online Science cafe / Acknowledgement in the research report / Research report
1 supporters are supporting with this reward. （No quantity limit）
Your name will be listed and acknowledged as a funder when I publish the outcome of this project in an academic paper.
Acknowledgement in an academic paper / Online seminar / Online Science cafe / Acknowledgement in the research report / Research report
1 supporters are supporting with this reward. （No quantity limit）
I invite you to a one-on-one, individual discussion on the possibility for virus detection in air using high power lasers. I can also talk about high-intensity and high-power lasers, etc., according to your needs. The event is scheduled to be held around December 2021.
Individual discussion / Acknowledgement in an academic paper / Online seminar / Online Science cafe / Acknowledgement in the research report / Research report
3 supporters are supporting with this reward. （No quantity limit）
SSL encryption communication is used in this Web site, and the informations filled out are safely transmitted.
Research report and others
12
supporters
back
（No quantity limit）
Acknowledgement in the research report and others
16
supporters
back
（No quantity limit）
Online Science cafe and others
15
supporters
back
（No quantity limit）
Online seminar and others
1
supporters
back
（No quantity limit）
Acknowledgement in an academic paper and others
1
supporters
back
（No quantity limit）
Individual discussion and others
3
supporters
back
（No quantity limit）
academist Prize 4th adopted
