Exploring Correlation Between CT Number and ADC Value for Understanding Radiation Response.

Abstract

PURPOSE/OBJECTIVE(S)\nRadiation induced changes in CT number (CTN) have been observed during/after radiation therapy (RT) in various tumors and healthy tissues and have been shown to correlate with radiation response. As CTN is related to the combined radiation attenuations of the intracellular volume (ICV) and the extracellular volume (ECV), the observed CTN changes may be contributed by the radiation induced variations in ICV and ECV, e.g., the ratio of ICV and ECV, in the tumor/tissue. The apparent diffusion coefficient (ADC), derived from diffusion-weighted MRI (DWI) is known to represent cellularity, i.e., the IVC/ECV ratio. The ADC or its change during/after RT has been shown to predict radiation response. CT and ADC images are commonly used in RT, offering necessary and complimentary data, one for anatomic and another for functional information. Based on these observations, we hypothesize that CTN correlates with ADC and this correlation can help understand radiation response in RT. The purpose of this study is to verify this hypothesis by analyzing selected CT and ADC data collected in our clinic.\n\n\nMATERIALS/METHODS\nSpleen and kidney were selected as the target organs for this study as they have no fat tissue that can interfere with and add complexity to the analysis. Data for a total of 20 and 15 patients with abdominal/pelvic cancers were used for the spleen and kidney analyses, respectively. The image data included the CT acquired during the standard CT simulation using the same CT scanner at 120 kVp and the ADC maps derived from the DWI acquired immediately before the first fraction using a 1.5T MR-Linac. The time gap between the CT and ADC acquisitions were approximately 3 weeks. Spleen and kidneys (left and right) were contoured on both the ADC and the CT images. The mean values of the ADC and CTN from the voxels within the contours were analyzed using a linear regression method.\n\n\nRESULTS\nThe CTNs over the patient cohort have a median of 44.7 HU (range 31.5 ∼ 56.9 HU) for spleen, and 29.7 HU (range 23.0 ∼ 38.2 HU) for kidney. Both spleen and kidney data show a strong correlation between CTN and ADC (r\u202f=\u202f-0.80 for spleen and -0.83 for kidney, P < 0.001 for both). The linear regression lines were ADC\u202f=\u202f-0.0212 x CTN\u202f+\u202f1.86 for the spleen and ADC\u202f=\u202f-0.0220 x CTN\u202f+\u202f2.64 for the kidney. The difference between the y-intercepts may be interpreted as the additional water flow related to the renal filtration. With such an adjustment the data for the two organs can be fitted into one regression line same as that for the spleen.\n\n\nCONCLUSION\nThere is a strong correlation between CTN and ADC in kidneys and spleen, suggesting that CTN is related to the ICV/ECV ratio (measured by ADC) and the radiation-induced changes in CTN during RT is related to the ADC changes. Further studies are needed to confirm the CTN and ADC correlation and to use the correlation for understanding radiation response.