We all know that the COVID-19 pandemic is getting worse, the economy is low, and people died. We may feel unable to do anything but follow the government-issued guideline and pray. But I believe that we, as researchers and developers in IT technologies, probably can make contributions from our aspects to help curb the situation. Below is the project idea. Whoever has the capability and ressource can work with it – we are working for the mankind.
A recent paper in Science [*] claims that substantial undocumented infection facilitates the rapid dissemination of the novel coronavirus. I believe we can build a web platform and use neural networks (NN) to help diagnose much of the undocumented infection. Below is a project idea, and I hope to team up with researchers and developers. In particular, I am looking for someone who has prior experience working on a project related to COVID-19, which would allow us to apply for the fast-grant. I am assistant prof. of Computer Science at IT University of Copenhagen, Denmark. Please let me know if you are interested in the project. Below is the idea:
- We build a website, that is sufficiently appealing so that patients or suspect patients are willing to submit their data
- We include in the database, for example, when people get a fever, their symptoms when they get a positive/negative test results when they get a full recovery.
- We build a neural network to learn from data. The learned model from NN allows us to make a diagnose.
To summarize: The platform has three parts, a website, a database, and a deep learning engine. Users of the platform are (suspected) patients and doctors. The diagnose can:
- give confidence to suspected patients – the main reason they may want to submit data, as they can get a probability of their potential fallouts and compare with other people; and
- complement doctors’ diagnose, helping them to make informed decisions on whether they should test the patients, suggest self-isolation to the suspected patents etc.
The idea of applying deep learning for medical diagnosis is certainly not new; my knowledge on the subject is little. References include:
- [*] Li, Ruiyun, et al. “Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2).” Science (2020).
- Bakator, Mihalj, and Dragica Radosav. “Deep learning and medical diagnosis: A review of literature.” Multimodal Technologies and Interaction 2.3 (2018): 47.
- Liu, Siqi, et al. “Early diagnosis of Alzheimer’s disease with deep learning.” 2014 IEEE 11th international symposium on biomedical imaging (ISBI). IEEE, 2014.