Radiomics of CT Features May Be Nonreproducible and Redundant: Influence of CT Acquisition Parameters.

Purpose To identify the reproducible and nonredundant radiomics features (RFs) for computed tomography (CT). Materials and Methods Two phantoms were used to test RF reproducibility by using test-retest analysis, by changing the CT acquisition parameters (hereafter, intra-CT analysis), and by comparing five different scanners with the same CT parameters (hereafter, inter-CT analysis). Reproducible RFs were selected by using the concordance correlation coefficient (as a measure of the agreement between variables) and the coefficient of variation (defined as the ratio of the standard deviation to the mean). Redundant features were grouped by using hierarchical cluster analysis. Results A total of 177 RFs including intensity, shape, and texture features were evaluated. The test-retest analysis showed that 91% (161 of 177) of the RFs were reproducible according to concordance correlation coefficient. Reproducibility of intra-CT RFs, based on coefficient of variation, ranged from 89.3% (151 of 177) to 43.1% (76 of 177) where the pitch factor and the reconstruction kernel were modified, respectively. Reproducibility of inter-CT RFs, based on coefficient of variation, also showed large material differences, from 85.3% (151 of 177; wood) to only 15.8% (28 of 177; polyurethane). Ten clusters were identified after the hierarchical cluster analysis and one RF per cluster was chosen as representative. Conclusion Many RFs were redundant and nonreproducible. If all the CT parameters are fixed except field of view, tube voltage, and milliamperage, then the information provided by the analyzed RFs can be summarized in only 10 RFs (each representing a cluster) because of redundancy.

Predictors of severity in ischemic colitis: Usefulness of early ultrasonography.

PURPOSE: The main objective was to evaluate the usefulness of early ultrasonography for determining prognosis of patients with ischemic colitis. MATERIALS AND METHODS: We reviewed the histories of patients with diagnosis of ischemic colitis over a period of 11 years. Two hundred twenty nine patients with ischemic colitis were included. Demographic, clinical, laboratory and sonographic findings were retrospectively analyzed. Patients were divided into two groups according to the outcome: mild (those who improved with conservative treatment) or severe (those who died or required surgery), and the findings of each group were compared. In addition, we developed a predictive model of multivariate logistic regression analysis, and then it was validated in a different population of 58 consecutive patients with ischemic colitis. RESULTS: The mild ischemic colitis group consisted of 184 patients (age 74,7±8,8years), and the severe group of 45 patients (age 78,6±7,7years). More pancolitis (p=0.005), altered pericolic fat (p=0.032) and free fluid (p=0.013) was observed in the severe ischemic colitis group compared with the mild group. Severe ischemic group had lower wall thickness (p=0.020) and higher resistive index than mild group (p=0.025). Multivariate analysis identified as protective factors to severe ischemic colitis: pain (p=0.026), diarrhea (p=0.034), rectal bleeding (p=0.000), and hypertension (p=0.001). Altered pericolic fat (p=0.008) and pancolitis (p=0.017) were sonographic findings significantly related to severe ischemic colitis. Our model correctly classified 88.4% of the study patients and 89.6% of the population of 58 patients used in validation. CONCLUSIONS: Sonographic features with clinical findings can predict the outcome of ischemic colitis. Our predictive model could differentiate properly between patients with mild or severe ischemic colitis.

Analysing the integration of MR images acquired in a non-radiotherapy treatment position into the radiotherapy workflow using deformable and rigid registration.

AIM: Our aim was to analyse the feasibility of integrating an MRI acquired in a non-radiotherapy set-up into the prostate cancer radiotherapy workflow. MATERIAL AND METHODS: The MRIs of 15 prostate cancer patients, acquired with a flat table-top (MRI-flat), and with a curved tabletop (MRI-curve) were analysed. MRIs were rigidly (RIR) and non-rigidly registered (DIR) with CT images. The prostate and rectum were contoured in each image set and translated to the CT, and IMRT plans were computed taking into account structural changes after RIR and DIR (P-flatRIR, P-curveRIR and P-curveDIR). In addition the P-curveDIR was overlaid with RIR structures. Spatial overlap parameters and dose-volume histogram metrics were generated. RESULTS: No significant differences were observed among P-flatRIR and P-curveRIR or P-flatRIR and P-curveDIR. Median gamma-values: P-curveRIR, 95.3%; P-curveDIR, 96%, translated-P-curveDIR, 95%. DVH metrics for translated-P-curveDIR were: Dmin, 64.5Gy; Dmean, 70.06Gy; V95%, 100%. No statistically significant differences were found in the dosimetric MRI-flatDIR values. CONCLUSIONS: The dosimetric reproduction of treatment position image following image registration of non-radiotherapy set-up images suggests that it is feasible to integrate these images into the radiotherapy workflow.