What is it about?
Our study leverages a convolutional neural network to extract imaging features. These features are jointly modelled with conventional biomarkers to predict the survival of patients with brain tuberculosis. Our study highlights the potential of this implementation compared with standard statistical model. By using an interpretability map, we presented some potential findings that may be interesting to investigate in future research.
Featured Image
Photo by Bhautik Patel on Unsplash
Why is it important?
This is amongst the first applications that leverage an end-to-end automatic feature extraction and prediction in tuberculous meningitis. By using a data-driven approach, we are not constrained by inductive bias and assumption, which in turn leads to more exploratorily intriguing findings.
Perspectives
This is my first application in using AI model and my first research in imaging dataset. Overall, TBM is a complicated disease with heterogenous appearance, variable lesion and paramount complications. With a small amount of data, we tried to investigate the potential, which would pave the wave for further research, rather than create a complete toolset of its own.
Trinh Dong
King's College London
Read the Original
This page is a summary of: Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis, PLOS One, May 2025, PLOS,
DOI: 10.1371/journal.pone.0321655.
You can read the full text:
Resources
Contributors
The following have contributed to this page
