Initial findings published by researchers from Tel-Aviv University and RADLogics, show promise through the application of Deep Learning CT Image Analysis to assist doctors in detecting the deadly Coronoavirus (COVID-19).
One of the challenges of rapidly developing a DL algorithm is the phases through which the algorithm must progress: Data-collection, Training, and Testing. Each of the phases of a DL algorithm development takes time, and with the rapid spread of the Coronavirus, it is essential to reduce algorithm development time to allow for the quick application of DL to the immediate problem at hand.
Rapid Application of Deep Learning
This is one of the key aspects highlighted in the researcher’s initial work, whereby the rapid development of AI-based tools to address immediate challenges can be addressed by adapting existing AI models. By combining existing AI-Tools with initial clinical understandings, the researchers were able to achieve the classification of Coronavirus vs Non-coronavirus cases of 0.996 AUC. This level of effectiveness in the detection represents a significant benefit in assisting doctors.
The research can be found here: http://bit.ly/38PJxpG
[share title=”Share this Post” facebook=”true” twitter=”true” linkedin=”true” email=”true” google_plus=”true” pinterest=”true”]