An interactive map of ~5,000 papers in AI-driven computed tomography reconstruction and enhancement.
This is an interactive tool for exploring the research landscape of AI methods applied to CT image reconstruction — sparse-view, limited-angle, low-dose denoising, super-resolution, and the deep learning foundations behind them.
There are three views and one tool:
Papers are positioned by what they're about, not by who cites whom. Each paper's title and abstract were embedded into 384-dimensional vectors using a sentence transformer (all-MiniLM-L6-v2), projected to 2D with UMAP, and clustered with HDBSCAN. Papers that discuss similar topics land near each other — even if they never cite one another.
The same papers, but now positioned by citation relationships. Edges show who references whom. Click any node to see its references (blue) and papers that cite it (orange).
Publication growth over time for each topic and sub-cluster. A stacked area chart shows which areas are expanding, and a sortable table ranks sub-topics by growth rate, recency, and average citations. Hover any row to overlay its trend line on the chart.
Paste your paper's title and abstract and the tool will embed it in your browser and show you where it falls on the map — which cluster it belongs to and which existing papers are most similar. No data is sent to any server; the embedding model runs locally via WebAssembly.
- Go to the live map
- Zoom, pan, and click papers to explore
- Use the sidebar to filter by topic or toggle dot sizing between global and in-field citations
- Switch between Semantic Map, Citation Network, and Research Trends tabs
- Click Place Your Paper to see where your own work fits
The map covers the intersection of medical physics and deep learning as applied to CT image formation:
- CT Reconstruction — sparse-view, limited-angle, few-view, iterative methods, FDK, algebraic techniques
- CT Enhancement — low-dose denoising, super-resolution, artifact reduction, sinogram inpainting
- AI Foundations — the architectures these methods build on (U-Net, GANs, diffusion models, transformers, NeRF, compressed sensing)
Papers were sourced from OpenAlex using bidirectional citation traversal from 43 seed papers, plus keyword-targeted discovery searches. Coverage of the core field is estimated at ~90% with all the important papers definitely included.
If you use this tool or the underlying dataset in your work, please cite:
Wiegmann, F.L. & Ford, N.L. (2026). CT Reconstruction Literature Map. University of British Columbia. https://ubc-ford-lab.github.io/ct_literature_maps/
Falk L. Wiegmann & Nancy L. Ford
University of British Columbia
2026