Room: Exhibit Hall | Forum 8
Purpose: In digital radiography, the exposure dose is highly associated with body parameters, such as body size and thickness; however, the manual measurement of body size and thickness is prone to error and subjective. Therefore, the purposes of this study were to measure body thickness using a non-contact infrared sensor, and to investigate the relationship between different body parameters and exposure factor (mAs) in chest radiography.
Methods: This study was approved by a local institutional review board. 929 consecutive male subjects (Age=20â€“85 y/o, BMI=13.9â€“40.9) who underwent regular chest screening examinations were enrolled. After the chest thickness (T) was measured with a home-made infrared sensor device, and body height (H) and weight (W) were recorded, posterior-to-anterior chest X-ray images were acquired from a digital radiography system (CXDI-401C CANON; FPD-TFT; KXO-32S TOSHIBA) using 105 kVp and 250 mA (AEC controls the time factor to keep similar exposure index, an index of image quality). Three curve fitting models, including exponential, polynomial and power, were performed to understand the relationships between five body parameters (T, T*W, BMI, BMI*T, and W/H) and exposure factor (mAs). Finally, the goodness of fit (R-square) was compared between them.
Results: The results showed that in three fitting models, the R-square was higher (> 0.8) for W/H than other body parameters in chest radiography. The highest R-square (0.8757) was in the polynomial fitting of W/H to mAs, whereas the lowest R-square (0.3585) was in the exponential fitting of T to mAs. These results indicated that W/H better represented the size and thickness of patients than other body parameters for chest radiography.
Conclusion: We conclude that under similar image quality, W/H is a suitable body parameter to predict mAs in chest radiography, and the relationship between W/H and mAs might be helpful to minimize radiation dose to patients clinically.
Funding Support, Disclosures, and Conflict of Interest: This study was supported by grant 107-26 from Kaohsiung Armed Forces General Hospital and by grant MOST107-2314-B-037-050-MY2 from Ministry of Science and Technology of Taiwan.