TL;DR
This study investigated the variability of gray scale values across different cone-beam computed tomography (CBCT) scanners and a multislice CT (MSCT) scanner, crucial for accurate bone density assessment in dentistry. While significant correlations were found, absolute gray values differed, highlighting the need for standardization, though CBCT still offers reliable data for clinical applications.
What Problem Does This Paper Address?
In dentistry, cone-beam computed tomography (CBCT) is increasingly used for detailed 3D imaging, particularly for planning dental implants. A key aspect of this planning involves assessing bone quality, which is often quantified using gray scale values (voxel values). However, a significant challenge is the lack of standardization in these gray values across different CBCT machines and even compared to traditional multislice computed tomography (MSCT) used in medicine. This non-standardization raises questions about the reliability and comparability of bone density assessments made with CBCT, potentially impacting treatment predictability. This paper aims to quantify and compare these gray scale variations to understand their implications for clinical use.
Methods
The researchers employed an in-vitro approach using a specially constructed acrylic resin phantom designed to mimic the positioning of a human head. This phantom contained standardized orifices filled with iodine solutions of varying concentrations (5%, 10%, 15%, 20%) and air, along with an aluminum stick, to represent different tissue densities and structures.
Four tomographic images were acquired from each of three different CBCT scanners (i-CAT Next Generation, Orthophos XG 3D, and PreXion 3D CBCT) and one MSCT scanner (SIEMENS® Dual Source CT Scanner). The acquisition protocols for each scanner were detailed.
The obtained DICOM images were then analyzed using two distinct software programs: ONIS (a general DICOM viewer) and DentalSlice® (a specialized dental imaging software). ONIS was used for both CBCT and MSCT images, while DentalSlice® was used exclusively for CBCT images. Gray values were measured in specific regions of interest within the phantom, including the iodine solutions, aluminum stick, and the external part of the central cylinder (representing water). The mean of multiple measurements within these regions was considered the gray value. Statistical analysis, including Pearson correlation and Friedman tests, was performed to compare the gray values obtained from different scanners and software.
Datasets
Not discussed in the paper. The study utilized a custom-built acrylic resin phantom with iodine solutions and an aluminum stick as its "dataset" for image acquisition.
Key Insights and Ideas
- Gray Scale Variability is Real: The study confirms that absolute gray scale values (voxel values) are not standardized across different CBCT scanners, nor are they directly equivalent to MSCT values. This is a fundamental challenge for direct quantitative comparisons.
- Software Matters: The choice of analysis software (ONIS vs. DentalSlice®) can influence the measured gray values, with DentalSlice® showing similar values between i-CAT and Sirona, but lower values for Prexion compared to i-CAT.
- i-CAT and MSCT Show Similar Behavior: The i-CAT CBCT scanner demonstrated gray values that were more similar to the MSCT scanner than to the other CBCT scanners (Sirona and Prexion).
- Sirona and Prexion are Similar to Each Other: These two CBCT scanners exhibited comparable gray value behaviors, distinct from i-CAT and MSCT.
- High Correlation, Different Absolute Values: Despite the differences in absolute gray values, there was a very high correlation between the values obtained from MSCT and all CBCT scanners, and also among the CBCT scanners themselves. This suggests that while the absolute numbers differ, the relative relationships and trends are consistent.
- CBCT Remains Clinically Relevant: Even with the observed variability, the study concludes that CBCT still provides reliable gray values for assessing bone quality, which is crucial for improving surgical planning and treatment predictability in dentistry.
Key Results
- High Correlation: Pearson correlation coefficients between scanners were consistently high, often above 0.97, and statistically significant (p<0.000). For example, the correlation between MSCT and i-CAT was 0.987, and between Sirona and Prexion was 0.964.
- Absolute Value Differences: Table 2 shows significant differences in mean gray values. MSCT had the highest mean (1103.271), followed by i-CAT (1089.975), then Sirona (703.596) and Prexion (571.262). The DentalSlice® analysis showed i-CAT and Sirona with similar values, while Prexion had lower values.
- Friedman Test Findings: The Friedman test revealed no statistically significant difference between MSCT and i-CAT, nor between Prexion and Sirona (using ONIS software). However, significant differences were found between MSCT and the other CBCT scanners, and between i-CAT and Prexion/Sirona. In DentalSlice® analysis, Prexion showed lower values than the other two CT scanners.
Limitations and Future Work
Limitations: * In Vitro Study: The study was conducted in vitro using an acrylic phantom. While designed to mimic human head positioning, it cannot fully replicate the complex biological environment and variations of in vivo bone. The authors acknowledge this as a potential divergence from clinical behavior. * Limited Software Comparison: While two software programs were used, the study focused on specific regions of interest and did not explore the full range of software capabilities or potential biases. * Scan Settings Not Exhaustively Explored: Although acquisition protocols were used, the paper notes that the influence of scan settings (like FOV) on gray values has not been widely explored and can vary significantly.
Future Work: * Standardization Efforts: The findings underscore the need for further research into standardizing gray scale values or developing robust calibration methods for CBCT in dentistry. * Clinical Validation: Future studies should focus on correlating these in vitro findings with in vivo clinical outcomes to fully validate the reliability of CBCT for bone density assessment in real-world implant cases. * Broader Software and Setting Analysis: Investigating a wider array of software and exploring the impact of various scan parameters (FOV, kVp, mA) on gray scale values would provide a more comprehensive understanding.
Summary updated