Faculty Bibliography

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.

OBJECTIVE: To assess recent national specialty trends in the use of diagnostic ultrasound (US) services in the Emergency Department (ED) setting. METHODS: We searched aggregated 1998-2012 Medicare Part B Master Files for ED diagnostic US studies, excluding cardiac and ophthalmic examinations. Studies were classified by body part and interpreting specialty. Subsequent analysis was performed for higher-volume services rendered by emergency physicians for which discrete codes were present longitudinally for complete vs limited examinations. National trends were analyzed. RESULTS: From 1998 to 2012, paid ED US studies interpreted by radiologists, emergency physicians, and all other physicians increased by 332% (from 221 712 to 735 858 examinations), 4454% (from 561 to 24 992), and 251% (from 26 961 to 67 787), respectively. The fraction of ED US examinations interpreted remained around 90% for radiologists, increased from 0.2% to 3% for emergency physicians, and decreased from 11% to 8% for other physicians. The fraction of complete abdominal and complete retroperitoneal studies interpreted by emergency physicians remained less than 1% from 1998 through 2012. However, emergency physicians experienced disproportionate growth in limited examinations (from 1% to 9% for abdominal studies and from <1% to 20% for retroperitoneal studies). Likewise, the fraction of (typically targeted) chest studies interpreted by emergency physicians increased from less than 1% to 63%. CONCLUSION: From 1998 to 2012, there was substantial growth in ED US studies for Medicare beneficiaries interpreted by radiologists and emergency physicians alike. For more commonly performed services distinguishable as complete vs limited in nature, growth in services by emergency physicians was most dramatic for less complex services.

PURPOSE: The aim of this report is to describe our early experience with radiologist-led direct public education about imaging-based cancer screening examinations, with a focus on the level of satisfaction and the educational impact reported by session participants. METHODS: Subspecialty radiologists provided 1-hour educational sessions to small lay public audiences covering breast, prostate, and lung cancer screening, focusing on key radiologic screening tests. Subsequently, session participants completed a survey pertaining to their levels of satisfaction and the perceived impact of the sessions on knowledge of the diseases and relevant screening tests; results from all topic sessions were pooled for analysis. RESULTS: One hundred nineteen participants attended a total of 144 sessions. The large majority of respondents agreed or strongly agreed that the sessions increased their understanding (95.4%), they gained information not obtained from clinic visit (88.1%), seeing radiology images helped their understanding (92.6%), they felt more involved in their own care (85.8%), they felt more prepared to undergo screening tests (87.4%), they were satisfied with the sessions' format (93.3%) and content (96.2%), their main questions were answered (88.3%), they would be interested in similar future radiologist-led sessions (88.5%), and they would recommend the sessions to friends (93.9%). The provided information (43.1%) and images (34.7%) were identified as the most helpful aspects of the sessions. CONCLUSION: Radiologist-led direct public education sessions about imaging-based cancer screening examinations yield high levels of audience satisfaction and add value to the public's understanding of the role of radiology and radiologists. Our experience may provide an encouraging model for other radiology practices in pursuing similar endeavors.

AIM: To compare the size and shape of the prostate between in-vivo and fresh ex-vivo magnetic resonance imaging (MRI), in order to quantify alterations in the prostate resulting from surgical resection. MATERIAL AND METHOD: Ten patients who had undergone 3 T prostate MRI using a phased-array coil and who were scheduled for prostatectomy were included in this prospective study. The ex-vivo specimen underwent MRI prior to formalin fixation or any other histopathological processing. Prostate volume in vivo and ex vivo was assessed using planimetry. Prostate shape was assessed by calculating ratios between the diameters of the prostate in all three dimensions. RESULTS: Mean prostate volume was significantly smaller ex vivo than in vivo (39.7 +/- 18.6 versus 50.8 +/- 26.8 cm3; p = 0.008), with an average change in volume of -19.5%. The right-to-left (RL)/anteroposterior (AP) ratio of the prostate, representing the shape of the prostate within its axial plane, was significantly larger ex vivo than in vivo (1.33 +/- 0.14 versus 1.21 +/- 0.12; p = 0.015), with an average percent change in RL/AP ratio of the prostate of +12.2%. There was no significant difference between in-vivo and ex-vivo acquisitions in terms of craniocaudal (CC)/AP (p = 0.963, median change = -2.1%) or RL/CC (p = 0.265, median change = +1.3%) ratios. CONCLUSION: The observed volume and shape change following resection has not previously been assessed by comparison of in-vivo and fresh ex-vivo MRI and likely represents loss of vascularity and of connective tissue attachments in the ex-vivo state. These findings have implications for co-registration platforms under development to facilitate improved understanding of the accuracy of MRI in spatial localization of prostate tumours.

INTRODUCTION: To compare qualitative and quantitative imaging features from conventional and diffusion-weighted (DW) magnetic resonance imaging (MRI) in detection of metastatic pelvic lymph nodes in bladder cancer patients undergoing cystectomy. MATERIALS AND METHODS: Thirty-six patients who had undergone cystectomy for bladder cancer with preoperative MRI with DWI sequence prior to surgery were included. Imaging features on conventional and DW-MRI were compared with histopathology at cystectomy. RESULTS: Nodal features associated with metastatic lymphadenopathy were short axis (AUC = 0.85, p < 0.001; when SA > 5 mm: sensitivity = 88%, specificity = 75%), long axis (AUC = 0.80, p < 0.001; when LA > 6 mm: sensitivity = 88%, specificity = 71%), apparent diffusion coefficient (ADC) on DWI, normalized to muscle (AUC = 0.66, p = 0.113; when nADC < 1.35: sensitivity = 75%, specificity = 68%), and absence of fatty hilum on conventional imaging (AUC = 0.73, p = 0.012; when fatty hilum absent, sensitivity = 75%, specificity = 71%). ADC without normalization was not associated with metastasis (p = 0.303). CONCLUSIONS: Imaging findings from conventional MRI and DWI achieved reasonable accuracy for detecting metastatic lymph nodes in bladder cancer, although sensitivity was higher than specificity. A short axis greater than 5 mm on conventional MRI had the highest accuracy of any individual finding. When using DWI, normalization of ADC values to muscle ADC may improve diagnostic performance.

Introduction and Objectives: With improving accuracy of prostate cancer detection and localization using multi-parametric MRI (mpMRI), there is an increasing interest in mpMRI guidance of diagnosis and surveillance biopsy. Widespread application of such concept must address the challenging issues, including the ability to perform MRI-guided biopsy in realtime with adequate accuracy, and in the simple urology office environment. We propose and assess a new approach a new approach to directly coregister 2D standard TRUS to MRI. Materials and Methods: The developed concept is to use a raw 2D Ultrasound (US) (B&K 8848 device) prostate image and register it to the corresponding MRI slice. Pre-acquired MRI data represents the whole gland in 3D as an ordered collection of 2D MRI slices of known thickness. We have developed software for USMRI coregistration based on image intensity, texture, and boundaries. The power parameter P is directly proportional to algorithm speed and accuracy. The result is source US overlaying target MRI for visualization. The system was prospectively tested on 8 data sets corresponding to US images matching with prostate MRI. These data come from patients who underwent MRI prior to biopsy. Coregistration results were evaluated as success/failure by an expert urologist who interactively adjusted the transparency of US-MRI overlays and recorded the alignment of anatomical landmarks in both modalities, especially the veru montanum. Results: The system was able to find the matching slice of T2WI for all single 2D standard US slice. In all cases the location of the veru montanum confirmed the coregistration accuracy in axial plan. The median rank of the T2WI slice was the fifth one among median number of 10 T2WI slices. We tested the power parameter P=1; P=5 and P=20. Twenty three coregistrations over 23 were correct. There was a significant positive correlation between prostate volume and time of computation for each P value. The image similarity reached a 0.93 mean Dice index u!