SkyScan 1276 - Best practices

Imaging preparation

• Minimize motion artifacts: Use appropriate anesthesia to reduce breathing and movement. Minimize animal motion by gently taping the legs and securing the body. This reduces motion blur and improves image sharpness.

• Support thermoregulation: Maintain the scanner gantry warm during in vivo scans Place animals on a heated bed or use warming pads during imaging to prevent hypothermia, which can alter physiology and image quality.

• Standardize animal positioning: Securely position animals in the same anatomical orientation using positioning beds, molds and tape. Consistent posture is essential for reproducibility in longitudinal or comparative scans.

• Remove metal artifacts: Ensure the animal is free from any metallic tags or implants in the region of interest, as these can introduce significant artifacts.

Acquisition settings

• Use standardized scan parameters: Fix voltage, current, filters, integration time, and exposure settings across scans to enable reliable comparisons. Avoid using automatic exposure settings when quantification is needed.

• Use flat field correction to improve image uniformity: Apply flat field correction during or prior to reconstruction to compensate for spatial sensitivity variations across the detector and uneven X-ray illumination. This reduces ring artifacts and gradient shading across the image, especially important for high-precision or quantitative applications.

• Optimize resolution versus dose: Choose a resolution that balances image quality with radiation exposure. High-resolution scans (≤10 µm voxel size) significantly increase radiation dose and may not be necessary for all applications.

• Minimize radiation dose in longitudinal studies: Use low-dose protocols when tracking changes over time to avoid affecting biological processes or animal health.

• Use respiratory gating for thoracic imaging: For lungs or heart imaging, respiratory/cardiac gating improves image clarity by synchronizing acquisition with the breathing cycle.

Calibration and controls

• Perform scanner calibration regularly: Use phantoms to calibrate for Hounsfield units, density, or BMD depending on application. Recalibrate if scan quality degrades or after hardware changes.

• Include internal or external phantoms: When quantifying density or structure (bone), use calibration phantoms in the field of view or perform concurrent scans (with same settings) for normalization.

• Use age- and sex-matched controls: MicroCT outcomes can vary significantly with age, sex, and body size. Matching across groups is essential for valid comparisons.

Data processing and quantification

• Apply consistent reconstruction parameters: Use the same reconstruction algorithm, voxel size, and filters to avoid introducing variability during image processing.

• Use a consistent image segmentation and analysis pipeline across individuals: Standardize the entire workflow, from image segmentation to quantitative analysis, using fixed parameters and methods across all animals. This includes thresholding, filtering, ROI/VOI placement, and feature extraction. Consistency minimizes variability, improves reproducibility, and ensures valid comparisons between individuals, groups, or time points.

• Segment using standardized thresholds: When segmenting structures, use predefined and validated Hounsfield Unit thresholds to ensure consistency across scans and subjects.

• Use 3D ROI placement for repeatability: Define volumes of interest (VOIs) using anatomical landmarks in 3D software. Save templates to reuse across subjects and time points.

• Specify quantitative endpoints: Clearly define whether using volume, density, trabecular thickness, porosity, and always report units and calculation methods.

• Avoid partial volume effects: Especially in small structures, ensure resolution is sufficient to avoid under- or over-estimation due to voxel averaging.

General recommendations

• Pilot scan before full study: Run a small test cohort to define scan parameters, and analysis workflow.

• Schedule imaging consistently: Perform scans at the same time of day to control for circadian effects on physiology and biodistribution.

• Document all scan parameters and animal metadata: Include anesthesia protocol, positioning details, scan settings, and animal weight/condition in each imaging session log to enable accurate interpretation and troubleshooting.

Ex vivo sample hydration and stabilisation

• Immobilize the sample inside the tube

  • Wrap the sample in a moist gauze, sponge, or soft foam soaked with 70% ethanol or PBS.

  • Alternatively, embed delicate tissues in 1–2% agarose gel (excellent support and hydration) or low-melting-point paraffin (scanning after dehydration).

  • This combination of hydration and mechanical stabilization prevents drift during scanning.

  • Use only low-density, X-ray transparent materials (polystyrene foam, Kapton) to avoid imaging artifacts.

• Use custom or sealed containers

  • Place the sample in sealed tubes or containers (microcentrifuge tubes, Kapton tubing…) or when using Falcon tubes, add a custom insert (3D-printed or foam) to cradle the sample snugly.

  • Limit liquid volume: Use only enough to cover the sample, and fill air gaps to minimize turbulence.

  • Ensure a tight fit to eliminate sample floating or shifting during rotation.

• Seal and maintain a moist environment

  • Seal containers with parafilm or caps to prevent evaporation.

  • If scanning without full sealing, place a moist cotton or sponge pad nearby inside the chamber to maintain humidity.

  • Minimize scan time for sensitive tissues to reduce dehydration risk.

Image quality

In microCT, achieving high image quality involves balancing multiple factors that influence spatial resolution, signal-to-noise ratio, and artifact suppression. Below is an integrated overview of how each parameter contributes to overall image quality.

• Pixel/Voxel size

  • Smaller voxel size improves spatial resolution, allowing finer detail to be captured.

  • Small pixel size reduces the number of X-ray photons per voxel, potentially lowering SNR unless compensated by longer exposure or averaging.

  • Optimize by choosing the smallest voxel size that still provides sufficient SNR for your sample.

• Rotation step (angular increment)

  • Smaller angular steps yield more projection images per rotation.

  • Low rotation step improves image consistency and reconstruction quality, reduces noise, and helps minimize artifacts.

  • Very large steps can cause undersampling, degrading both resolution and SNR.

  • Use finer steps (<0.5°) for high-detail samples; larger steps for faster scans where ultra-fine detail isn’t needed.

• Frame averaging

  • Averaging multiple frames per projection reduces random noise and enhances SNR.

  • Particularly valuable when scanning at lower voltages or with small voxel sizes.

  • Typically, 3–5 frames per projection (ex vivo) offer a good SNR improvement without excessive scan time.

• Filter selection

  • Filters (Al, Cu) remove low-energy X-rays that contribute to noise and beam hardening.

  • This hardens the X-ray beam, improving penetration and contrast uniformity, indirectly enhancing image quality and SNR.

  • Use filters to improve image consistency, especially for dense samples like bone or metal.

• Tube voltage (kV)

  • Higher voltage increases X-ray energy, improving penetration and SNR, especially in dense materials.

  • Lower voltage enhances contrast in low-density samples but may reduce penetration and increase noise.

  • Match voltage to sample type: low kV for soft tissue, high kV with filters for dense samples.

• Source-to-sample distance (geometric magnification)

  • Bringing the X-ray source closer increases magnification, reducing effective voxel size and improving spatial resolution.

  • Increased magnification may amplify motion artifacts and reduce SNR if photon flux is not adjusted.

  • Use for high-resolution scans of small features, ensuring other parameters (exposure, averaging) support good SNR.

• Non-binning versus inning (detector settings)

  • Non-binning retains full detector resolution, capturing maximum spatial detail.

  • It collects fewer photons per pixel, resulting in lower SNR.

  • Binning (2×2) combines adjacent pixels, improving SNR at the cost of spatial resolution.

  • Use non-binning when high detail is essential and SNR can be managed through other settings.

Imaging parameters affecting microCT quality