Electron tomography¶
A transmission-electron-microscopy technique that reconstructs three-dimensional specimen structure from a series of projections collected over incremental tilt angles.
Core Idea¶
Electron tomography reconstructs a three-dimensional specimen estimate from an aligned series of transmission-electron-microscope projections acquired over incremental tilt angles. Electron tomography reconstructs three-dimensional structure from transmission-electron-microscope projections collected across specimen tilts. Alignment and a reconstruction algorithm combine the views, while limited tilt creates missing-wedge anisotropy and electron dose can damage samples. BF-TEM and ADF-STEM have different projection/contrast assumptions. Reconstruction, denoising, segmentation, and visualization are distinct inference stages; reported resolution must name specimen and validation regime. This entry is conceptual, not an operating protocol.
Scope of Application¶
The concept applies in cell biology and related work when its constitutive roles and limits are explicit. Use it with specimen, mode, tilt coverage, dose, alignment, image model, reconstruction, artifact, and resolution evidence explicit; distinguish single images and mere renderings. This entry is descriptive and nonprocedural.
- Cell biology. Images subcellular organization.
- Structural biology. Studies macromolecular assemblies.
- Materials science. Reconstructs nanoscale morphology.
- Nanotechnology. Examines particles and defects.
- Method development. Improves reconstruction/validation.
Clarity¶
State specimen, imaging mode, angle range/increment, dose, alignment, projection model, reconstruction, resolution metric, and missing-wedge effects. The closest near miss sets the boundary: Serial-section 3D microscopy is the closest miss: it reconstructs sections rather than angular electron projections.
Manages Complexity¶
ET compresses many destructive or dose-limited projections into a navigable volume, gaining depth while introducing inverse-problem uncertainty and anisotropic evidence. Electron tomography reconstructs a volume from a tilt series, not from one perspective image. Each projection samples the specimen along a different orientation, and an algorithm combines those projections under assumptions about image formation, alignment, thickness, and stability. Limited tilt range leaves a “missing wedge” of information that creates anisotropic resolution and artifacts; radiation damage constrains the number and dose of views, especially for biological samples. Bright-field TEM and ADF-STEM have different projection requirements and contrast behavior. Reconstruction, denoising, segmentation, and visualization are separate inference stages, and a visually crisp surface can overstate raw resolution. Typical biological resolution and atomic-resolution demonstrations refer to different instruments, specimens, dose budgets, and validation regimes. The technique remains descriptive here and supplies no laboratory operating procedure. The central angular coverage–radiation damage tradeoff is this: More views improve inversion but add dose. A second visual sharpness–evidential resolution tension matters because Processing can make boundaries look more certain.
Abstract Reasoning¶
Use three linked moves: prepare a suitable 3D target conceptually; collect projections across declared tilts; align views and check projection assumptions. As a collapse test, identity collapses when multiple tilt projections or computational 3D inversion is absent. A fourth check is to reconstruct with a named algorithm.
Knowledge Transfer¶
Tomographic inversion transfers to other modalities, but electron tomography requires electron projections and their specimen/dose physics. The entry is nonprocedural. No canonical parent prime is currently asserted; broader structural comparisons remain related-prime analogies until separately adjudicated in the DAG. ET is an electron-based subtype.
Neighborhood in Abstraction Space¶
Electron tomography sits in a moderately populated region (48th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.
Family — Domain-Specific Measurement Parameters (36 abstractions)
Nearest neighbors
- Remote Visual Inspection — 0.88
- Electron backscatter diffraction — 0.87
- Analytical technique — 0.86
- Imaging Method — 0.86
- Face validity — 0.85
Computed from structural-signature embeddings · 2026-10-08