New CT Architecture Uses Foveated Imaging and Diffusion Models for High-Resolution Reconstruction
Researchers have developed a deep learning model that synthesizes intermediate CT slices to halve the effective through-plane spacing in head CT scans, reducing anisotropy and noise in a single inference pass. Head CT imaging typically achieves sub-millimeter in-plane resolution but uses 2–5 mm spacing between axial slices, which degrades 3D reconstructions and volumetric measurements such as hematoma estimation. The system outperforms classical interpolation and pretrained video frame interpolation methods, offering potential improvements for clinical imaging workflows.
A preprint posted to arXiv presents a deep learning system designed to address a longstanding limitation of head computed tomography: while in-plane resolution is sub-millimeter, through-plane slice spacing typically ranges from 2 to 5 mm, creating anisotropy that hampers multiplanar reconstructions, volumetric measurements, and downstream algorithms assuming near-isotropic voxels. The model synthesizes intermediate axial slices from pairs of neighboring slices, effectively halving through-plane spacing while simultaneously producing denoised outputs — two benefits from a single inference pass. The researchers systematically evaluated several loss functions, including MSE, L1, SSIM, MS-SSIM, and hybrid combinations, finding that MS-SSIM combined with L1 offered the strongest balanced performance profile. All converged models outperformed classical interpolation baselines and pretrained video frame interpolation methods (RIFE and FILM) on structural metrics across a held-out test set, with results reported using patient-level bootstrap confidence intervals and paired statistical tests. The authors also noted training instability with SSIM-family losses, identifying partial remedies but acknowledging residual divergence at smaller batch sizes. External validation was demonstrated on a single out-of-distribution head CT series from Hospital Universitario Virgen del Rocío, where the model exhibited the implicit-denoising behavior predicted by the authors' theoretical analysis, though this constitutes only a single case.
What's missing
The study's external validation relies on a single out-of-distribution case, which is insufficient to establish generalizability across diverse scanner types, patient populations, or clinical pathologies. The paper does not report downstream clinical validation — such as whether improved slice interpolation measurably improves diagnostic accuracy or clinical outcomes. The work has not yet undergone peer review.
What different sources said
- arXiv physicsCenter
DD-INR: Dynamics-Driven Implicit Neural Representation for Accelerated Whole-Brain Functional MRI Reconstruction
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.