Faculty Bibliography

OBJECTIVE. The purpose of this study was to assess perspectives and information relating to CT radiation risk on Twitter, a popular microblogging social network. MATERIALS AND METHODS. Publicly available posts on Twitter ("tweets") containing both the words "CT" and "radiation" were identified from the 1st week of each month in 2013. Type of user posting and source of linked articles were recorded. Two reviewers assessed the content of tweets and links regarding CT's benefit-to-risk ratio (favorable, unfavorable, etc.). RESULTS. Six hundred twenty-one relevant tweets were tweeted by 557 unique users, of whom 90 (16%) were physicians (17 of these were radiologists), 30 (5%) were medical practices or hospitals, 34 (6%) were patients, 8 (1%) were physicists or technologists, and 395 (71%) were other types of users. Two hundred twenty-seven tweets included user commentary regarding CT's benefit-to-risk ratio, of which 134 (59%) were unfavorable or concerned, 65 (29%) were neutral, 22 (10%) were informative regarding CT dose reduction strategies, and only 6 (3%) were favorable. Four hundred seventy-two tweets (76%) included links to a total of 99 unique articles, of which 25 (25%) were unfavorable or concerned, 10 (10%) were favorable, 25 (25%) were neutral, and 39 (39%) were informative regarding CT dose reduction. Article types were non-peer-reviewed medical sources (n = 50), lay press (n = 15), peer-reviewed medical journals (n = 13), blogs (n = 12), advertisements (n = 5), and informational websites (n = 4). CONCLUSION. The large majority of content on Twitter was either unfavorable or concerned regarding CT radiation risk. Most articles were not peer-reviewed and were posted by nonphysicians; posts by physicians were largely by nonradiologists. More active engagement on Twitter by radiologists and physicists and increased dissemination of peer-reviewed articles may achieve a more balanced representation and alleviate concerns regarding CT radiation risk on social networks.

OBJECTIVE: To evaluate the agreement in volumes of prostate tumors determined on multiparametric MRI (mpMRI) and histologic assessment, using detailed software-assisted co-registration. MATERIALS AND METHODS: 37 patients who underwent 3T mpMRI (T2WI, DWI/ADC, DCE) were included. A radiologist traced the borders of suspicious lesions on T2WI and ADC and assigned a suspicion score (SS) from 2-5; a uro-pathologist traced borders of tumors on histopathologic photographs. Software was used to co-register MRI and 3D digital reconstructions of RP specimens and compute imaging and histopathologic volumes. Agreement in volumes between MRI and histology was assessed using Bland-Altman plots and stratified by tumor characteristics. RESULTS: Among 50 tumors, mean difference and 95% limits of agreement on MRI relative to histology were -32% (-128% to +65%) on T2WI and -47% (-143% to +49%) on ADC. For all tumor subsets, volume under-estimation was more marked on ADC maps (mean difference ranging from -57% to -16%) than T2WI (mean difference ranging from -45% to +2%). 95% limits of agreement were wide for all comparisons, with lower 95% limit ranging between -77% and -143% across assessments. Volume under-estimation was more marked for tumors with Gleason score >/=7 or MRI SS 4 or 5. CONCLUSION: Volume estimates of PCa using MRI tended to substantially under-estimate histopathologic volumes, with wide variability in extent of under-estimation across cases. These findings have implications for efforts to use MRI to guide risk assessment.