Organ localization: toward prospective patient-specific organ dosimetry in computed tomography

With increased focus on radiation dose from medical imaging, prospective radiation dose estimates are becoming increasingly desired. Using available populations of adult and pediatric patient phantoms, radiation dose calculations can be catalogued and prospectively applied to individual patients that best match certain anatomical characteristics. In doing so, the knowledge of organ size and location is a required element. Here, the authors develop a predictive model of organ locations and volumes based on an analysis of adult and pediatric computed tomography (CT) data.

Fifty eight adult and 69 pediatric CT datasets were segmented and utilized in the study. The maximum and minimum points of the organs were recorded with respect to the axial distance from the tip of the sacrum. The axial width, midpoint, and volume of each organ were calculated. Linear correlations between these three organ parameters and patient age, BMI, weight, and height were determined.

No statistically significant correlations were found in adult patients between the axial width, midpoint, and volume of the organs versus the patient age or BMI. Slight, positive linear trends were found for organ midpoint versus patient weight (max r(2) = 0.382, mean r(2) = 0.236). Similar trends were found for organ midpoint versus height (max r(2) = 0.439, mean r(2) = 0.200) and for organ volume versus height (max r(2) = 0.410, mean r(2) = 0.153). Gaussian fits performed on probability density functions of the adult organs resulted in r(2)-values ranging from 0.96 to 0.996. The pediatric patients showed much stronger correlations overall. Strong correlations were observed between organ axial midpoint versus age, height, and weight (max r(2) = 0.842, mean r(2) = 0.790; max r(2) = 0.949, mean r(2) = 0.894; and max r(2) = 0.870, mean r(2) = 0.847, respectively). Moderate linear correlations were also observed for organ axial width versus height (max r(2) = 0.772, mean r(2) = 0.562) and for organ volume versus height (max r(2) = 0.781, mean r(2) = 0.601).

Adult patients exhibited small variations in organ volume and location with respect to height and weight, but no meaningful correlation existed between these parameters and age or BMI. Once adulthood is reached, organ morphology and positioning seem to remain static. However, clear trends are evident between pediatric organ locations versus age, height, and weight. Such information can be incorporated into a matching methodology that may provide the highest probability of representing the anatomy of a patient undergoing a clinical exam to prospectively estimate the radiation dose.