Assessing the effect of scanning parameter on the size and density of pulmonary nodules: a phantom study

Donghua Meng, Zhen Wang, Changsen Bai, Zhaoxiang Ye, Zhipeng GaoBMC Medical Imaging 2024 · DOI

TL;DR

This study investigates how CT scanning parameters (tube voltage and current) affect the accuracy of pulmonary nodule size and density measurements using a phantom. The findings suggest specific parameter combinations are optimal for accurately assessing different nodule types (solid vs. ground-glass), which is crucial for effective lung cancer screening and treatment monitoring.

What Problem Does This Paper Address?

Lung cancer remains a significant health concern, and CT scans are a primary tool for screening and monitoring. However, the accuracy of CT-based measurements of pulmonary nodules, particularly their size and density, can be influenced by various scanning parameters. Existing clinical criteria for assessing treatment response, like RECIST, rely heavily on nodule size, but the optimal CT settings for obtaining these measurements are not always clearly defined or consistently applied across different nodule types. This variability can lead to discrepancies in diagnosis and treatment evaluation. This paper aims to quantify the impact of specific CT scanning parameters (tube voltage and tube current) on the accurate measurement of both solid and ground-glass nodules (GGNs) using a realistic phantom.

Methods

The researchers utilized the LUNGMAN N1, a multipurpose anthropomorphic chest phantom, to simulate various pulmonary nodules. These simulated nodules had known CT values (densities) and sizes, representing both solid tumors (-100 HU) and ground-glass nodules (-630 HU and -800 HU), with diameters of 8, 10, and 12 mm.

The study systematically varied CT scanning parameters, specifically tube voltage (80, 100, and 120 kVp) and tube current (100, 140, and 200 mA), resulting in nine distinct scanning combinations. Image acquisition was performed using a GE Discovery 750HD CT scanner.

Image analysis was conducted by two experienced physicists using 3D Slicer software. They performed segmentation of the nodules and measured their longest diameters, adhering to RECIST criteria. Nodule density was assessed by calculating the mean Hounsfield Unit (HU) within a defined region of interest. Image quality was evaluated using Signal-to-Noise Ratio (SNR) and Contrast-to-Noise Ratio (CNR). Consistency between the two physicists' measurements was assessed using the Interclass Correlation Coefficient (ICC).

Datasets

The study used the LUNGMAN N1 phantom, which contains simulated nodules of varying sizes (8, 10, 12 mm) and densities (-100 HU, -630 HU, -800 HU). This phantom serves as a controlled environment for evaluating CT scanning parameters, effectively acting as a custom benchmark for this specific study. No external datasets were used for training or evaluation.

Key Insights and Ideas

  • Nodule Type Matters: The optimal scanning parameters for accurate size measurement differ significantly between solid nodules and ground-glass nodules. This highlights the need for tailored scanning protocols rather than a one-size-fits-all approach.
  • Density vs. Size Optimization: Parameters that optimize nodule size measurement are not necessarily the same as those that optimize density measurement. This suggests a trade-off or a need to prioritize one aspect over the other depending on the clinical goal.
  • Low kVp for Solid, Higher kVp for GGN Size: Lower tube voltages (80 kVp) appear to be better for accurately measuring the size of solid nodules, while higher voltages (100 kVp) are preferred for GGN size.
  • Density Accuracy at Specific kVp: GGN densities are best represented at 80 kVp, and solid nodule densities at 100 kVp, indicating specific energy levels are crucial for accurate attenuation measurements.
  • High Consistency: The study demonstrated excellent consistency in measurements between two experienced physicists, validating the reliability of their analysis methods.

Key Results

  • Solid Nodule Size: The combination of 80 kVp and 140 mA was found to be preferred for measuring the size of solid nodules, yielding measurements closest to the true size.
  • GGN Nodule Size: The combination of 100 kVp and 100 mA was preferred for measuring the size of ground-glass nodules.
  • Nodule Density: CT values of GGNs were closest to the ground truth when measured at 80 kVp, while solid nodule CT values were closest to the ground truth at 100 kVp.
  • Image Quality: SNR increased with both tube current and tube voltage, while CNR decreased with increasing tube voltage.
  • Consistency: Interclass Correlation Coefficients (ICCs) for both longest diameter and mean HU measurements were consistently high (ICC > 0.94 for diameter, ICC > 0.99 for HU), indicating excellent inter-observer reliability.

Limitations and Future Work

The authors acknowledge several limitations:

  • Phantom-Based Study: The study was conducted using a phantom, and further validation is needed with clinical data to confirm these findings in real-world patient populations. The phantom's body model was based on a 70 kg adult male, so its applicability to other body sizes and types requires further investigation.
  • Single CT Scanner: Only one CT scanner model was used. Future research should explore whether these findings are generalizable to other CT scanner models and their associated reconstruction technologies.
  • Limited Nodule Diameters: The simulated nodule diameters (8, 10, 12 mm) represent a subset of nodules encountered clinically. More diverse nodule sizes and shapes should be investigated.
  • Limited Energy Range: The study explored three energy levels. Investigating more energy levels, particularly using dual-energy CT techniques, could further optimize image quality and assessment accuracy.

Future work should focus on validating these findings in clinical settings, exploring the impact of advanced reconstruction techniques (like DLIR), and investigating dual-energy CT for improved nodule assessment.

Summary updated

BibTeX
@article{meng2024assessing,
  author = {Donghua Meng and Zhen Wang and Changsen Bai and Zhaoxiang Ye and Zhipeng Gao},
  title = {{Assessing the effect of scanning parameter on the size and density of pulmonary nodules: a phantom study}},
  journal = {BMC Medical Imaging},
  year = {2024},
  doi = {10.1186/s12880-023-01190-4},
}
Figure