A methodology to estimate the patient diameter and thickness from thoracic and abdominal projection radiographs of adult patients

Patient dose management systems can be part of a department's quality management tools to estimate items such as the radiation burden in specific groups or list dose outliers for further follow up. Patient size information could improve both aspects, but is not generally available. A new metric is proposed to estimate patient size for thorax and abdominal projection radiography from parameters available in thedicomheader and therefore accessible by dose management software. The tested hypothesis was that an attenuation metric, related to the ratio of detector air-kerma to incident air-kerma, inversely correlates with patient size. Such a metric was defined and then worked out for thorax and abdomen projection radiography. Input material consisted of the thorax or abdominal radiographs of 137 cases, completed with a recent CT scan as ground truth size. From the CT, the water equivalent diameter (WED) and water equivalent thickness (WET) were calculated. The correlation between the attenuation metric and the patient size was established separately for thorax and abdomen. Validation of the attenuation metric predicting the patient size was performed using extra sets of examinations on three more radiographic x-ray devices, with available CT scan. The attenuation metric had a good correlation (R²) of 0.91 and 0.84 with the WED for thorax and abdomen respectively. The corresponding values for the WET were 0.89 and 0.78. Validation of the methodology on the devices with standardized exposure index in thedicomheaders showed that the WED could be estimated within ±15% and the WET within ±30% for thorax and abdomen examinations. The ground truth and estimated size were found statistically equivalent. An attenuation metric based ondicomtags allows to estimate the patient size in projection radiography. This could now be implemented in patient dose management systems.

Comparing different methods for estimating radiation dose to the conceptus

PURPOSE

To compare different methods available in the literature for estimating radiation dose to the conceptus (D_conceptus) against a patient-specific Monte Carlo (MC) simulation and a commercial software package (CSP).

METHOD

Eight voxel models from abdominopelvic CT exams of pregnant patients were generated. D_conceptus was calculated with an MC framework including patient-specific longitudinal tube current modulation (TCM). For the same patients, dose to the uterus, D_uterus, was calculated as an alternative for D_conceptus, with a CSP that uses a standard-size, non-pregnant phantom and a generic TCM curve. The percentage error between D_uterus and D_conceptus was studied. Dose to the conceptus and percent error with respect to D_conceptus was also estimated for three methods in the literature.

RESULTS

The percentage error ranged from -15.9% to 40.0% when comparing MC to CSP. When comparing the TCM profiles with the generic TCM profile from the CSP, differences were observed due to patient habitus and conceptus position. For the other methods, the percentage error ranged from -30.1% to 13.5% but applicability was limited.

CONCLUSIONS

Estimating an accurate D_conceptus requires a patient-specific approach that the CSP investigated cannot provide. Available methods in the literature can provide a better estimation if applicable to patient-specific cases.

KEY POINTS

• A patient's internal anatomy affects the dose to the conceptus. • Conceptus position has an influence on its dose estimation. • Patient anatomy and specific TCM must be considered for accurate conceptus dosimetry. • D _uterus to a standard-size phantom should not be used as D _conceptus .