Automatic Mapping of CT Scan Locations on Computational Human Phantoms for Organ Dose Estimation

Abstract

To develop an algorithm to automatically map CT scan locations of patients onto computational human phantoms to provide with patient-specific organ doses. We developed an algorithm that compares a two-dimensional skeletal mask generated from patient CTs with that of a whole body computational human phantom. The algorithm selected the scan locations showing the highest Dice Similarity Coefficient (DSC) calculated between the skeletal masks of a patient and a phantom. To test the performance of the algorithm, we randomly selected five sets of neck, chest, and abdominal CT images from the National Institutes of Health Clinical Center. We first automatically mapped scan locations of the CT images on a computational human phantom using our algorithm. We had several radiologists to manually map the same CT images on the phantom and compared the results with the automated mapping. Finally, organ doses for automated and manual mapping locations were calculated by an in-house CT dose calculator and compared to each other. The visual comparison showed excellent agreement between manual and automatic mapping locations for neck, chest, and abdomen-pelvis CTs. The difference in mapping locations averaged over the start and end in the five patients was less than 1 cm for all neck, chest, and AP scans: 0.9, 0.7, and 0.9 cm for neck, chest, and AP scans, respectively. Five cases out of ten in the neck scans show zero difference between the average manual and automatic mappings. Average of absolute dose differences between manual and automatic mappings was 2.3, 2.7, and 4.0% for neck, chest, and AP scans, respectively. The automatic mapping algorithm provided accurate scan locations and organ doses compared to manual mapping. The algorithm will be useful in cases requiring patient-specific organ dose for a large number of patients such as patient dose monitoring, clinical trials, and epidemiologic studies.

Construction of patient‐specific computational models for organ dose estimation in radiological imaging

“…The ImPACT CT patient dosimetry calculator estimates organ‐level absorbed doses and effective dose based on spreadsheet tools and adult stylized phantoms . Although this approach is practical, it is impaired by the dosimetric uncertainties resulting from the large differences between the anatomy of patients and simulated computational phantoms (categorized by gender and age) in addition to the inherent uncertainties associated with protocol‐based mapping of the scan location on the computational models . Since Monte Carlo calculations using patient‐specific models are commonly considered as reference for organ dose estimation from diagnostic imaging procedures, the implementation of an easy to use and reliable framework enabling to estimate patient‐specific organ dose for individual patients in clinical setting is highly desirable.…”

A Review of Organ Dose Calculation Tools for Patients Undergoing Computed Tomography Scans

“…However, it is reported that general scan protocols are not always strictly followed [46] even in a clinical trial where protocols are carefully designed and provided to participating hospitals. To avoid dose errors caused by a disagreement between a patient and phantom, automatic mapping methods have been investigated [52] that compare patient CT images with predefined anatomical landmarks in computational human phantoms to map the locations of scan start and end of a patient on those of phantoms. Machine learning could be used for mapping organs from CT images to computational human phantoms [53].…”

Automatic organ completion with image stitching for personalized radiation dosimetry in CT examinations

“…However, large differences between the anatomy of patients and computational phantoms may cause nonnegligible uncertainties in radiation dosimetry. 26 Delineation of the target volume and organs at risk from patient-specific images are commonly used in radiation treatment planning, 27 where CT numbers are converted into material properties 12 and voxel-based radiation dosimetry can be obtained through Monte Carlo simulations. A voxel dose map combined with automatic organ segmentation can achieve high accuracy in dosimetry with clinically acceptable speed.…”

“…Recently, habitus‐dependent phantom series 22–24 and deformable computational phantoms 25 have been developed to represent individual patients for radiation dose estimation. However, large differences between the anatomy of patients and computational phantoms may cause non‐negligible uncertainties in radiation dosimetry 26 …”