Faculty Bibliography

Introduction and Objectives: Focal therapy (FT) is an emerging approach for treatment of localized prostate cancer. Multi-parametric (mp) MRI demonstrated capability to monitor the effect of FT procedures. At time of followup, the ablated zone (AZ) clearly undergoes local shrinkage. Accurate definition of AZ at follow-up will improve FT evaluation and oncologic safety. We analyzed the volume and shape changes of the gland between pre and post FT MRI by developing a 3D coregistration method to compensate for deformation of the gland in response to FT. Materials and Methods: We studied 10 patients who underwent FT (interstitial laser ablation and photodynamic therapy) within IRB approved trials. All patients underwent preoperative, early control and late postoperative 3T MRI which included T2, T1, diffusion and perfusion weighted sequences. We have developed image registration software to analyze, transfer and model shape changes using a deformable and a rigid body transformation. Alignment between pre- and post-op images of AZ was assessed using the overlap index or Dice index (Di) and the maximum boundary distance, or Hausdorff distance (HD). Correction for deformation was measured using the HD normalized by the volume to transform in mm/ cc and automated feature of the software. Results: There was a significant volume decrease D of the gland that averaged 6.49 cc (p=0.017) between preoperative and postoperative gland. D was directly correlated (=0.738, p=0.014) with the ablated volume (7.88, p=0.04). We successfully co registered pre operative to post operative MRI in each cases. There was a significant increase D computed with deformable versus rigid transform. Deformable model achieved a significantly more accurate match of pre- vs. post-FT AZ with deformable (Di=0.88, HD=1.98 mm) vs. rigid transformation (Di=0.95 HD=3.83mm). The deformable approach also yielded a higher (p=0.019) correction of deformation (0.72mm/cc) compared to the rigid model (0.15mm/cc). Conclusion: We described a novel !

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.

Evidence suggests that normal pressure hydrocephalus (NPH) is underdiagnosed in day to day radiologic practice, and differentiating NPH from cerebral atrophy due to other neurodegenerative diseases and normal aging remains a challenge. To better characterize NPH, we test the hypothesis that a prediction model based on automated MRI brain tissue segmentation can help differentiate shunt-responsive NPH patients from cerebral atrophy due to Alzheimer disease (AD) and normal aging. Brain segmentation into gray and white matter (GM, WM), and intracranial cerebrospinal fluid was derived from pre-shunt T1-weighted MRI of 15 shunt-responsive NPH patients (9 men, 72.6 +/- 8.0 years-old), 17 AD patients (10 men, 72.1 +/- 11.0 years-old) chosen as a representative of cerebral atrophy in this age group; and 18 matched healthy elderly controls (HC, 7 men, 69.7 +/- 7.0 years old). A multinomial prediction model was generated based on brain tissue volume distributions. GM decrease of 33 % relative to HC characterized AD (P < 0.005). High preoperative ventricular and near normal GM volumes characterized NPH. A multinomial regression model based on gender, GM and ventricular volume had 96.3 % accuracy differentiating NPH from AD and HC. In conclusion, automated MRI brain tissue segmentation differentiates shunt-responsive NPH with high accuracy from atrophy due to AD and normal aging. This method may improve diagnosis of NPH and improve our ability to distinguish normal from pathologic aging.

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!

Septal nuclei, located in basal forebrain, are strongly connected with hippocampi and important in learning and memory, but have received limited research attention in human MRI studies. While probabilistic maps for estimating septal volume on MRI are now available, they have not been independently validated against manual tracing of MRI, typically considered the gold standard for delineating brain structures. We developed a protocol for manual tracing of the human septal region on MRI based on examination of neuroanatomical specimens. We applied this tracing protocol to T1 MRI scans (n=86) from subjects with temporal epilepsy and healthy controls to measure septal volume. To assess the inter-rater reliability of the protocol, a second tracer used the same protocol on 20 scans that were randomly selected from the 72 healthy controls. In addition to measuring septal volume, maximum septal thickness between the ventricles was measured and recorded. The same scans (n=86) were also analysed using septal probabilistic maps and Dartel toolbox in SPM. Results show that our manual tracing algorithm is reliable, and that septal volume measurements obtained via manual and automated methods correlate significantly with each other (p<.001). Both manual and automated methods detected significantly enlarged septal nuclei in patients with temporal lobe epilepsy in accord with a proposed compensatory neuroplastic process related to the strong connections between septal nuclei and hippocampi. Septal thickness, which was simple to measure with excellent inter-rater reliability, correlated well with both manual and automated septal volume, suggesting it could serve as an easy-to-measure surrogate for septal volume in future studies. Our results call attention to the important though understudied human septal region, confirm its enlargement in temporal lobe epilepsy, and provide a reliable new manual delineation protocol that will facilitate continued study of this critical region.

PURPOSE: To evaluate the effect of different methods to convert magnetic resonance (MR) signal intensity (SI) to gadolinium concentration ([Gd]) on estimation and reproducibility of model-free and modeled hepatic perfusion parameters measured with dynamic contrast-enhanced (DCE)-MRI. MATERIALS AND METHODS: In this Institutional Review Board (IRB)-approved prospective study, 23 DCE-MRI examinations of the liver were performed on 17 patients. SI was converted to [Gd] using linearity vs. nonlinearity assumptions (using spoiled gradient recalled echo [SPGR] signal equations). The [Gd] vs. time curves were analyzed using model-free parameters and a dual-input single compartment model. Perfusion parameters obtained with the two conversion methods were compared using paired Wilcoxon test. Test-retest and interobserver reproducibility of perfusion parameters were assessed in six patients. RESULTS: There were significant differences between the two conversion methods for the following parameters: AUC60 (area under the curve at 60 s, P < 0.001), peak gadolinium concentration (Cpeak, P < 0.001), upslope (P < 0.001), Fp (portal flow, P = 0.04), total hepatic flow (Ft, P = 0.007), and MTT (mean transit time, P < 0.001). Our preliminary results showed acceptable to good reproducibility for all model-free parameters for both methods (mean coefficient of variation [CV] range, 11.87-23.7%), except for upslope (CV = 37%). Among modeled parameters, DV (distribution volume) had CV <22% with both methods, PV and MTT showed CV <21% and <29% using SPGR equations, respectively. Other modeled parameters had CV >30% with both methods. CONCLUSION: Linearity assumption is acceptable for quantification of model-free hepatic perfusion parameters while the use of SPGR equations and T1 mapping may be recommended for the quantification of modeled hepatic perfusion parameters. J. Magn. Reson. Imaging 2014;40:90-98 (c) 2013 Wiley Periodicals, Inc.

OBJECTIVE. The purpose of this study was to evaluate split renal function, estimate single-kidney renal function, and identify cause of obstruction in patients with ureteropelvic junction (UPJ) obstruction by using contrast-enhanced dynamic MR renography (MRR). MATERIALS AND METHODS. Seventeen patients with UPJ obstruction underwent MRR and diuresis nuclear renography. Nuclear renography assessment of split renal function and mechanical versus functional obstruction served as the reference standard. The Baumann-Rudin model for determining glomerular filtration rate (GFR) was applied to generate single-kidney renal function (SK-GFRMRR) from MRR cortical and medullary enhancement curves. MRR split renal function of the right kidney (SK-GFRMRR of the right kidney normalized to the sum of SK-GFRMRR of both kidneys) was compared with nuclear renography. The MRR estimate of total GFR (eGFRMRR) was compared with that derived from Modification of Diet in Renal Disease (MDRD) formula (eGFRMDRD). Renal pelvic rate of signal intensity change (PUR) was compared between functionally and mechanically obstructed kidneys. RESULTS. There was excellent correlation between MRR and nuclear renography measure of split renal function ratio (r = 0.87, p < 0.01), with mean difference of less than 10%. There was moderate correlation (r = 0.60, p = 0.01) between eGFRMRR and eGFRMDRD. eGFRMRR underestimated eGFRMDRD, with mean difference of 13.3 mL/min/1.73 m(2). PUR in mechanically obstructed units was significantly lower (0.39 +/- 0.26 vs 2.0 +/- 1.38 min(-1); p < 0.01) compared with functionally obstructed units. PUR discriminated mechanical from functional obstruction with accuracy of 89%. CONCLUSION. In patients with UPJ obstruction, MRR can measure split renal function, estimate eGFRMDRD with moderate correlation, and accurately discriminate mechanical from functional obstruction, thus potentially providing a "one-stop shop" examination.

Established as a method to study anatomic changes, such as renal tumors or atherosclerotic vascular disease, magnetic resonance imaging (MRI) to interrogate renal function has only recently begun to come of age. In this review, we briefly introduce some of the most important MRI techniques for renal functional imaging, and then review current findings on their use for diagnosis and monitoring of major kidney diseases. Specific applications include renovascular disease, diabetic nephropathy, renal transplants, renal masses, acute kidney injury, and pediatric anomalies. With this review, we hope to encourage more collaboration between nephrologists and radiologists to accelerate the development and application of modern MRI tools in nephrology clinics.

Blood oxygen level-dependent (BOLD) MRI data of kidney, while indicative of tissue oxygenation level (Po2), is in fact influenced by multiple confounding factors, such as R2, perfusion, oxygen permeability, and hematocrit. We aim to explore the feasibility of extracting tissue Po2 from renal BOLD data. A method of two steps was proposed: first, a Monte Carlo simulation to estimate blood oxygen saturation (SHb) from BOLD signals, and second, an oxygen transit model to convert SHb to tissue Po2. The proposed method was calibrated and validated with 20 pigs (12 before and after furosemide injection) in which BOLD-derived tissue Po2 was compared with microprobe-measured values. The method was then applied to nine healthy human subjects (age: 25.7 +/- 3.0 yr) in whom BOLD was performed before and after furosemide. For the 12 pigs before furosemide injection, the proposed model estimated renal tissue Po2 with errors of 2.3 +/- 5.2 mmHg (5.8 +/- 13.4%) in cortex and -0.1 +/- 4.5 mmHg (1.7 +/- 18.1%) in medulla, compared with microprobe measurements. After injection of furosemide, the estimation errors were 6.9 +/- 3.9 mmHg (14.2 +/- 8.4%) for cortex and 2.6 +/- 4.0 mmHg (7.7 +/- 11.5%) for medulla. In the human subjects, BOLD-derived medullary Po2 increased from 16.0 +/- 4.9 mmHg (SHb: 31 +/- 11%) at baseline to 26.2 +/- 3.1 mmHg (SHb: 53 +/- 6%) at 5 min after furosemide injection, while cortical Po2 did not change significantly at approximately 58 mmHg (SHb: 92 +/- 1%). Our proposed method, validated with a porcine model, appears promising for estimating tissue Po2 from renal BOLD MRI data in human subjects.

The precision, accuracy, and efficiency of a novel semi-automated renal segmentation technique for volumetric interpolated breath-hold sequence (VIBE) MRI sequences was analyzed using 7 clinical datasets (14 kidneys). Two observers performed whole-kidney segmentation using EdgeWave segmentation software based on constrained morphological growth. Ground truths were prepared by manual tracing of kidney on each of approximately 40 slices. Using the software, the average inter-observer disagreement was 2.7%+/- 2.1% for whole kidney volume, 2.1%+/- 1.8% for cortex, and 4.1%+/- 3.2% for medulla. In comparison to the ground truth kidney volume, the error was 2.8%+/- 2.5% for whole kidney volume, 3.1%+/- 1.7% for cortex, and 3.6%+/-.3.1% for medulla. It took an average of 4:14 +/- 1:42 minutes of operator time, plus 2:56 +/- 1:23 minutes of computer time to perform segmentation of one whole kidney, cortex, and medulla. Segmentation speed, inter-observer agreement and accuracy were several times better than those of our existing graph-cuts segmentation technique requiring approximately 20 minutes per case and with 7-10% error. With the expedited image processing, high inter-observer agreement, and volumes closely matching true values, kidney volumetry becomes a reality in many clinical applications.