Evaluating bone trabeculae using CBCT: A comparative study with Micro-CT

Inside the human body, the jaw is not a solid block of stone but a complex, spongy lattice of bone. This internal framework, known as trabecular bone, consists of tiny struts and plates that provide strength while keeping the structure light. For a dentist planning to place an implant, understanding the thickness and spacing of these microscopic struts is vital. If the bone is too thin or the gaps between the struts are too wide, the implant may fail to hold. For decades, doctors have relied on two-dimensional X-rays to guess at this internal structure, but those flat images often blur the details, hiding the true nature of the bone beneath the surface. To see the bone clearly, scientists have long used a powerful tool called micro-computed tomography, or micro-CT, which acts like a high-resolution microscope for bone, revealing details down to the width of a human hair. However, this gold-standard machine is slow, expensive, and too large to fit in a dental office, limiting its use to research labs.

In the real world of dental clinics, doctors use a different machine called cone-beam computed tomography, or CBCT. It is faster, smaller, and delivers a lower dose of radiation, making it the standard for planning implant surgery. But a lingering question remains: can this clinical machine see the tiny details of the bone's internal scaffolding as well as the research-grade micro-CT? The answer depends heavily on a setting called voxel size, which determines the resolution of the digital image. A smaller voxel size creates a sharper picture but requires more radiation and time, while a larger size is faster and safer but might blur the fine lines. Researchers at Babol University of Medical Sciences set out to find the balance, asking whether the lower-resolution images from a standard dental scanner could still provide a reliable map of the bone's microstructure.

To answer this, the team did not scan living patients, but rather six blocks of sheep jawbone, carefully prepared and preserved. They placed these blocks into a CBCT scanner and took images using three different levels of detail: a fine resolution of 0.1 millimeters, a medium resolution of 0.2 millimeters, and a coarser resolution of 0.3 millimeters. They then took the exact same bone blocks and scanned them with a micro-CT machine, which served as the perfect reference point. Using specialized software, they measured four specific features of the bone: how much of the space was filled with actual bone, the average thickness of the tiny struts, the average distance between them, and the number of struts per unit of space.

The results revealed a clear pattern of distortion when comparing the clinical scanner to the research standard. The CBCT images consistently made the bone struts appear thicker and the gaps between them wider than they actually were. Conversely, the machine counted fewer struts than were truly present. This happened because the larger digital blocks used to build the image could not capture the finest edges of the bone, causing the boundaries to blur and merge. The researchers found that these differences were not just minor glitches; they were statistically significant across all the measurements. Even when they used the finest setting available on the CBCT machine, the images still showed the bone structure differently than the micro-CT did.

The study also looked at how changing the resolution affected the measurements within the CBCT scans themselves. When the team used the coarsest setting of 0.3 millimeters, the bone struts appeared significantly thicker than when they used the finest setting of 0.1 millimeters. This suggests that the choice of resolution directly alters the doctor's perception of the bone's quality. In the upper jaw, the spacing between the struts looked much wider at the 0.3-millimeter setting compared to the 0.1-millimeter setting. In the lower jaw, the thickness of the struts showed a similar trend, appearing much larger in the lower-resolution images. The researchers noted that because the average thickness of these bone struts falls between 0.1 and 0.3 millimeters, the coarser settings struggle to represent them accurately, often inflating their size.

Ultimately, the study concludes that while CBCT is an excellent tool for general planning, it does not show the bone's microscopic reality with perfect fidelity. The differences in how the bone appears between the two machines are large enough for the human eye to notice, which means a doctor looking at a CBCT scan might perceive the bone to be denser or more robust than it truly is. The authors suggest that for a balance between image quality and patient safety, a medium resolution of 0.2 millimeters might offer the best compromise, providing a clear enough view to assess the bone without exposing the patient to unnecessary radiation. The findings remind us that every image is a representation, not a perfect mirror, and that understanding the limitations of our tools is just as important as the images they produce.