Room: Exhibit Hall | Forum 1
Purpose: Diagnostic reference levels have proven effective in protocol optimization and dose reduction for a cohort of patients, while further optimization for individual patients requires the knowledge of patient-specific organ dose and image quality. This study summarizes and analyzes radiation output of CT scans of over a thousand adult patients, and estimates their specific organ and effective doses.
Methods: The anonymized CT DICOM files of a total of 1200 patients who received a brain scan, chest scan, abdomen-pelvis (AP) scan, or chest-abdomen-pelvis (CAP) scan were randomly chosen (300 for each body region) and collected under HIPAA compliance. DoseWatcháµ€á´¹ records were collected and summarized for CTDIvol, DLP, and SSDE, and used as inputs for Monte Carlo based organ and effective dose estimator VirtualDoseCTáµ€á´¹. Patients were automatically matched to the most similar computational human phantom of the estimator based on sex and effective diameter. Doses were estimated with scan range defined by anatomical landmarks as in clinical scans. Ratios of patient SSDE to phantom SSDE were applied to estimated results to obtain patient-specific dose estimates.
Results: For brain scans, the 75th percentiles were 60 mGy (CTDIvol) and 1078 mGy*cm (DLP), with the top five doses in mGy as 59.7 (lens of eye), 49.2 (brain), 45.6 (salivary glands), 40.7 (extrathoracic region), and 18.2 (oral mucosa). For chest scans, the 75th percentiles were 12 mGy (CTDIvol), 14 mGy (SSDE), and 474 mGy*cm (DLP), with the top five doses in mGy as 10.6 (spleen), 10.1 (adrenals and thyroid), 9.4 (liver), 9.2 (thymus), and 9.1 (lungs and kidneys). AP and CAP scan results are a work in progress.
Conclusion: CT radiation output were summarized and patient-specific organ doses were estimated for over 1000 adult oncology patients.
Funding Support, Disclosures, and Conflict of Interest: Dr. X. George Xu is the CEO of VirtualPhantoms, Inc.