Webinar summary
The webinar, the first in the series Mastering CT: Advanced Techniques in Practice, provided a deep dive into challenges and practical solutions for CT users dealing with dense or large samples that attenuate X-rays heavily. The session began with a refresher on the basic principles of X-ray CT: capturing multiple 2D projections as an object rotates, then reconstructing them into a 3D digital twin using computer algorithms. However, Ted moved beyond the basics, emphasizing that while CT seems straightforward in theory, real-world applications—especially with high-attenuation samples—can be far from simple.
High-attenuation samples, such as metals, concrete, bone, and large dense objects, pose three primary problems: low contrast, artifacts (particularly metal and beam hardening), and sample mounting difficulties. Ted explained that contrast in X-ray imaging stems from differential attenuation; when materials absorb X-rays similarly, distinguishing internal structures becomes difficult. This is common in dense metals, where the image may reveal only silhouettes, not internal features. Metal artifacts, like streaks and false structures, arise when dense materials distort X-ray paths or scatter radiation, misleading the detectors. Beam hardening, meanwhile, causes the edges of dense objects to appear unnaturally bright due to lower-energy photons being absorbed more readily, creating a cupping effect that can obscure actual structure. Mounting issues, though perhaps unintuitive, were shown to be just as impactful; improperly oriented samples can exacerbate artifacts or reduce image quality significantly, even if all other scan parameters are ideal.
To combat these issues, Ted discussed a methodical optimization process, using a sintered steel and copper sample as an illustrative example. First, increasing the X-ray tube voltage from 130 to 200 kV improved photon penetration, boosting contrast. However, this alone didn’t eliminate artifacts. Introducing physical filters, such as copper or tin plates, removed low-energy photons from the beam, further improving contrast and reducing beam hardening. Finally, Ted used post-processing corrections—software-based adjustments applied during or after reconstruction—to suppress residual artifacts. The cumulative effect was striking, as shown in a side-by-side comparison: the optimized scan revealed fine internal features clearly, while the initial scan was blurry and riddled with false structures.
Ted emphasized that improving mounting technique is just as critical. Using directional knowledge of artifacts, he showed how simply rotating a sample could avoid having dense regions block key features. An example came from scanning a PCB board: mounted in one orientation, the scan was plagued by streaks and noise; mounted differently, the same board yielded a clean, artifact-reduced image with vastly better visibility of internal solder joints. He likened mounting strategy to a mindset rather than a checklist—users should always consider how to minimize problematic X-ray paths through dense regions.
The session concluded with real-world examples, including scans of a VTOL drone, a prismatic lithium-ion battery, and a car radiator. Each showcased how proper energy selection, filtration, and mounting strategy could yield vastly improved image quality. In particular, drone rotor and circuit board images demonstrated how higher energy and better orientation could resolve fine features otherwise obscured by artifacts or low contrast. Ted reiterated that while it’s helpful to consult attenuation tables and theoretical models, practical trial-and-error—guided by fundamental understanding—is often the most efficient route to a usable scan.
Overall, the webinar successfully bridged theory and practice, showing that mastering high-attenuation imaging isn’t about memorizing fixed procedures but about developing intuition for how CT parameters interact with physical sample properties. Participants left with both conceptual insights and practical strategies to improve their CT scans, regardless of how dense or challenging their samples might be.
