[Show abstract][Hide abstract] ABSTRACT: Purpose:
We propose a systematic approach to correlate MRI and digital histopathology in prostate.
T2-weighted (T2W) MRI and diffusion-weighted imaging (DWI) are acquired, and a patient-specific mold (PSM) is designed from the MRI. Following prostatectomy, a whole mount tissue specimen is placed in the PSM and sectioned, ensuring that tissue blocks roughly correspond to MRI slices. Rigid body and thin plate spline deformable registration attempt to correct deformation during image acquisition and tissue preparation and achieve a more complete one-to-one correspondence between MRIs and tissue sections. Each tissue section is stained with hematoxylin and eosin and segmented by adopting a machine learning approach. Utilizing this tissue segmentation and image registration, the density of cellular and tissue components (lumen, nucleus, epithelium, and stroma) is estimated per MR voxel, generating density maps for the whole prostate.
This study was approved by the local IRB, and informed consent was obtained from all patients. Registration of tissue specimens and MRIs was aided by the PSM and subsequent image registration. Tissue segmentation was performed using a machine learning approach, achieving [Formula: see text]0.98 AUCs for lumen, nucleus, epithelium, and stroma. Examining the density map of tissue components, significant differences were observed between cancer, benign peripheral zone, and benign prostatic hyperplasia (p value [Formula: see text]5e[Formula: see text]2). Similarly, the signal intensity of the corresponding areas in both T2W MRI and DWI was significantly different (p value [Formula: see text]1e[Formula: see text]10).
The proposed approach is able to correlate MRI and digital histopathology of the prostate and is promising as a potential tool to facilitate a more cellular and zonal tissue-based analysis of prostate MRI, based upon a correlative histopathology perspective.
International Journal of Computer Assisted Radiology and Surgery 09/2015; DOI:10.1007/s11548-015-1287-x · 1.71 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Approximately 15% of patients who undergo radical prostatectomy (RP) for prostate cancer develop local recurrence, which is heralded by a rise in serum prostate-specific antigen (PSA) levels. Early detection and treatment of recurrence improves the outcome of salvage treatment. We investigated the ability of multiparametric magnetic resonance imaging (mpMRI)-transrectal ultrasound (TRUS) fusion-guided biopsy (FGB) combined with "cognitive biopsy" to confirm local recurrence of prostate cancer after RP.
In this retrospective study conducted between January 2010 and December 2014, patients with rising PSA levels after RP who had no known evidence of distant metastases underwent mpMRI including T2-weighted (T2W) imaging, diffusion-weighted imaging, dynamic contrast-enhanced (DCE) MRI at 3 Tesla, and subsequent MRI-ultrasound fusion biopsy with cognitive assistance. The detection rate of locally recurrent disease was determined.
A total of 10 patients (mean age = 67y, mean PSA level = 3.44ng/ml) met the inclusion criteria. Of the 10 patients, all had positive findings suspicious for local recurrence on mpMRI per entrance criterion. The most important features on mpMRI were early enhancement on DCE MR images and hypointensity on T2W images. The average lesion diameter on mpMRI was 1.12cm (range: 0.40-2.20cm). All suspicious lesions (16/16, 100%) were positive on T2W MR images, 14 (89%) showed positive features on apparent diffusion coefficient maps of diffusion-weighted images, and 16 (100%) were positive on DCE MR images. MRI-TRUS FGBs were positive in 10/16 lesions (62.5%) and 8/10 (80%) patients.
MRI-TRUS FGB with cognitive assistance is able to detect and diagnose locally recurrent lesions after RP, even at low PSA levels. This may facilitate early detection of recurrent disease and improve salvage treatment outcomes.
Published by Elsevier Inc.
[Show abstract][Hide abstract] ABSTRACT: (18)F-FDG PET/CT is used to characterize many malignancies, but is not recommended for localized prostate cancer. This study explores the value of multi-parametric MRI (mpMRI) in characterizing incidental prostate (18)F-FDG uptake.
Thirty-one patients who underwent (18)F-FDG PET/CT for reasons unrelated to prostate cancer and prostate mpMRI were eligible for this retrospective study. The mpMRI included T2-weighted (T2W), dynamic contrast enhancement (DCE), apparent diffusion coefficient (ADC), and MR spectroscopy (MRS) sequences. Fourteen patients were excluded (n = 8 insufficient histopathology, n = 6 radical prostatectomy before PET), and final analysis included 17 patients. A nuclear medicine physician, blinded to clinicopathologic findings, identified suspicious areas and maximum standardized uptake values (SUVmax) on (18)F-FDG PET/CT. Sector-based imaging findings were correlated with annotated histopathology from whole-mount or MRI/transrectal ultrasound fusion biopsy samples. Positive predictive values (PPVs) were estimated using generalized estimating equations with logit link. Results were evaluated with Kruskal-Wallis and Dunn's multiple comparisons tests.
The PPV of (18)F-FDG PET alone in detecting prostate cancer was 0.65. Combining (18)F-FDG PET as a base parameter with mpMRI (T2W, DCE, ADC, and MRS) increased the PPV to 0.82, 0.83, 0.83, and 0.94, respectively. All benign lesions had SUVmax < 6. Malignant lesions had higher SUVmax values that correlated with Gleason scores. There was a significant difference in SUVmax per prostate between the Gleason ≥ 4 + 5 and benign categories (p = 0.03).
Focal incidental prostate (18)F-FDG uptake has low clinical utility alone, but regions of uptake may harbor high-grade prostate cancer, especially if SUVmax > 6. Using mpMRI to further evaluate incidental (18)F-FDG uptake aids the diagnosis of prostate cancer.
[Show abstract][Hide abstract] ABSTRACT: To quantify changes in tumor microvascular (< 1 mm) perfusion relative to commonly used angiographic endpoints.
Rabbit Vx2 liver tumors were embolized with 100-300-μm LC Bead particles to endpoints of substasis or complete stasis (controls were not embolized). Microvascular perfusion was evaluated by delivering two different fluorophore-conjugated perfusion markers (ie, lectins) through the catheter before embolization and 5 min after reaching the desired angiographic endpoint. Tumor microvasculature was labeled with an anti-CD31 antibody and analyzed with fluorescence microscopy for perfusion marker overlap/mismatch. Data were analyzed by analysis of variance and post hoc test (n = 3-5 per group; 18 total).
Mean microvascular density was 70 vessels/mm(2) ± 17 (standard error of the mean), and 81% ± 1 of microvasculature (ie, CD31(+) structures) was functionally perfused within viable Vx2 tumor regions. Embolization to the extent of substasis eliminated perfusion in 37% ± 9 of perfused microvessels (P > .05 vs baseline), whereas embolization to the extent of angiographic stasis eliminated perfusion in 56% ± 8 of perfused microvessels. Persistent microvascular perfusion following embolization was predominantly found in the tumor periphery, adjacent to normal tissue. Newly perfused microvasculature was not detected at complete stasis but was observed following embolization to complete angiographic stasis.
Nearly half of tumor microvasculature remained patent despite embolization to complete angiographic stasis. The observed preservation of tumor microvasculature perfusion with angiographic endpoints of substasis and stasis may have implications for tumor response to embolotherapy.
Published by Elsevier Inc.
Journal of vascular and interventional radiology: JVIR 08/2015; DOI:10.1016/j.jvir.2015.06.036 · 2.41 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The imaging features of unresectable hepatic malignancies in patients who underwent radiofrequency ablation (RFA) in combination with lyso-thermosensitive liposomal doxorubicin (LTLD) were determined.
A phase I dose escalation study combining RFA with LTLD was performed with peri- and post- procedural CT and MRI. Imaging features were analyzed and measured in terms of ablative zone size and surrounding penumbra size. The dynamic imaging appearance was described qualitatively immediately following the procedure and at 1-month follow-up. The control group receiving liver RFA without LTLD was compared to the study group in terms of imaging features and post-ablative zone size dynamics at follow-up.
Post-treatment scans of hepatic lesions treated with RFA and LTLD have distinctive imaging characteristics when compared to those treated with RFA alone. The addition of LTLD resulted in a regular or smooth enhancing rim on T1W MRI which often correlated with increased attenuation on CT. The LTLD-treated ablation zones were stable or enlarged at follow-up four weeks later in 69 % of study subjects as opposed to conventional RFA where the ablation zone underwent involution compared to imaging acquired immediately after the procedure.
The imaging features following RFA with LTLD were different from those after standard RFA and can mimic residual or recurrent tumor. Knowledge of the subtle findings between the two groups can help avoid misinterpretation and proper identification of treatment failure in this setting. Increased size of the LTLD-treated ablation zone after RFA suggests the ongoing drug-induced biological effects.
CardioVascular and Interventional Radiology 07/2015; DOI:10.1007/s00270-015-1186-0 · 2.07 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Therapeutic embolization of blood vessels is a minimally invasive, catheter-based procedure performed with solid or liquid emboli to treat bleeding, vascular malformations, and vascular tumors. Hepatocellular carcinoma (HCC) affects about half a million people per year. When unresectable, HCC is treated with embolization and local drug therapy by transarterial chemoembolization (TACE). For TACE, drug eluting beads (DC Bead(®)) may be used to occlude or reduce arterial blood supply and deliver chemotherapeutics locally to the tumor. Although this treatment has been shown to be safe and to improve patient survival, the procedure lacks imaging feedback regarding the location of embolic agent and drug coverage. To address this shortcoming, herein we report the synthesis and characterization of image-able drug eluting beads (iBeads) from the commercial DC Bead(®) product. Two different radiopaque beads were synthesized. In one approach, embolic beads were conjugated with 2,3,5-triiodobenzyl alcohol in the presence of 1,1'-carbonyldiimidazol to give iBead I. iBead II was synthesized with a similar approach but instead using a trimethylenediamine spacer and 2,3,5-triiodobenzoic acid. Doxorubicin was loaded into the iBeads II using a previously reported method. Size and shape of iBeads were evaluated using an upright microscope and their conspicuity assessed using a clinical CT and micro-CT. Bland and Dox-loaded iBeads II visualized with both clinical CT and microCT. Under microCT, individual bland and Dox loaded beads had a mean attenuation of 7904 ± 804 and 11,873.96 ± 706.12 HU, respectively. These iBeads have the potential to enhance image-guided TACE procedures by providing localization of embolic-particle and drug.
Journal of Materials Science Materials in Medicine 06/2015; 26(6):5530. DOI:10.1007/s10856-015-5530-3 · 2.59 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Local binary pattern (LBP) is a simple gray scale descriptor to characterize the local distribution of the gray levels in an image. Multi-resolution LBP and/or combinations of the LBPs have shown to be effective in texture image analysis. However, it is unclear what resolutions or combinations to choose for texture analysis. Examining all the possible cases is impractical and intractable due to the exponential growth in a feature space. This limits the accuracy and time- and space-efficiency of LBP. Here, we propose a data mining approach for LBP, which efficiently explores a high-dimensional feature space and finds a relatively smaller number of discriminative features. The features can be any combinations of LBPs. These may not be achievable with conventional approaches. Hence, our approach not only fully utilizes the capability of LBP but also maintains the low computational complexity. We incorporated three different descriptors (LBP, local contrast measure, and local directional derivative measure) with three spatial resolutions and evaluated our approach using two comprehensive texture databases. The results demonstrated the effectiveness and robustness of our approach to different experimental designs and texture images.
Expert Systems with Applications 06/2015; 42(9):4529-4539. DOI:10.1016/j.eswa.2015.01.055 · 2.24 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The posterior subcapsular region of the prostate is often undersampled by transrectal ultrasound (TRUS)-guided biopsy. The close proximity of these lesions to the posterior capsular wall of the prostate makes them difficult to localize while increasing the need for early detection because of their increased risk for extracapsular extension. We retrospectively evaluated the multiparametric MRI (mpMRI) features of subcapsular prostate cancers to make radiologists more aware of this condition.
Between January 2010 and July 2014, all patients referred for 3T mpMRI and subsequent MR-US Fusion-guided biopsy (FgBx) and systematic 12-core sextant biopsy (SBx) under an IRB approved protocol, were reviewed, and imaging confirmed subcapsular prostate cancers were identified. Subcapsular lesions were defined as thin lesions that were just inside the prostate capsule. Matching patient demographics and clinical findings including age, PSA, PSA density, whole prostate volume, history of prostate cancer, Gleason score, and clinical management were tabulated.
Of 992 eligible patients, 33 patients had subcapsular lesions in the prostate detected by mpMRI. Mean age, PSA, and prostate volume in this group were 63 years (range: 52-76 years), 8.4 ng/mL (range: 1.22-65.20), and 53 mL (range: 12-125 mL), respectively. The combination biopsy (SBx + FgBx) confirmed prostate cancer in 24 of 33 patients (72.7%) and in 9 patients the biopsy was negative. Of the 24 cancers, 19 were confirmed on both FgBx and conventional biopsy; however, 5 cancers were only detected on FgBx. In 4 of the 19 patients in which both biopsy methods were positive, the FgBx yielded a higher Gleason score.
Subcapsular lesions on mpMRI are relatively infrequent but are usually malignant. Although the majority are confirmed on conventional 12-core biopsies, about 20% of these lesions require FgBx for diagnosis, and FgBx more accurately grades the lesions in another 20%. Thus, FgBx is of considerable benefit in confirming the diagnosis of subcapsular prostate cancer despite their proximity to the prostatic capsule.
[Show abstract][Hide abstract] ABSTRACT: To correlate the highest percentage core involvement (HPCI) and corresponding tumor length (CTL), on systematic 12-core biopsy (SBx) and targeted magnetic resonance imaging/transrectal ultrasound (MRI/TRUS) fusion biopsy (TBx), with total MRI prostate cancer (PCa) tumor volume (TV).
Fifty patients meeting criteria for active surveillance (AS) based on outside SBx, who underwent 3.0T multiparametric prostate MRI (MP-MRI), followed by SBx and TBx during the same session at our institution were examined. PCa TV's were calculated using MP-MRI and then correlated using bivariate analysis with the HPCI and CTL, for SBx and TBx.
For TBx, HPCI and CTL showed a positive correlation (R2 = 0.31, p<0.0001 and R2 = 0.37, p<0.0001 respectively) with total MRI PCa TV, whereas for SBx these parameters showed a poor correlation (R2 = 0.00006, p=0.96 and R2 = 0.0004, p=0.89 respectively). For detection of patients with clinically significant MRI derived tumor burden greater than 500 mm3, SBx was 25% sensitive, 90.9% specific (falsely elevated due to missed tumors and extremely low sensitivity) and 54% accurate in comparison to TBx, which was 53.6% sensitive, 86.4% specific and 68% accurate.
HPCI and CTL on TBx positively correlates with total MRI PCa TV, whereas there was no correlation seen with SBx. TBx is superior to SBx for detecting tumor burden greater than 500 mm3. When using biopsy positive MRI derived TV's, TBx better reflects overall disease burden, improving risk stratification amongst candidates for active surveillance.
Journal of endourology / Endourological Society 04/2015; DOI:10.1089/end.2015.0027 · 1.71 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Purpose:
The authors propose a computer-aided diagnosis (CAD) system for prostate cancer to aid in improving the accuracy, reproducibility, and standardization of multiparametric magnetic resonance imaging (MRI).
The proposed system utilizes two MRI sequences [T2-weighted MRI and high-b-value (b = 2000 s/mm2) diffusion-weighted imaging (DWI)] and texture features based on local binary patterns. A three-stage feature selection method is employed to provide the most discriminative features. The authors included a total of 244 patients. Training the CAD system on 108 patients (78 MR-positive prostate cancers and 105 benign MR-positive lesions), two validation studies were retrospectively performed on 136 patients (68 MR-positive prostate cancers, 111 benign MR-positive lesions, and 117 MR-negative benign lesions).
In distinguishing cancer from MR-positive benign lesions, an area under receiver operating characteristic curve (AUC) of 0.83 [95% confidence interval (CI): 0.76–0.89] was achieved. For cancer vs MR-positive or MR-negative benign lesions, the authors obtained an AUC of 0.89 AUC (95% CI: 0.84–0.93). The performance of the CAD system was not dependent on the specific regions of the prostate, e.g., a peripheral zone or transition zone. Moreover, the CAD system outperformed other combinations of MRI sequences: T2W MRI, high-b-value DWI, and the standard apparent diffusion coefficient (ADC) map of DWI.
The novel CAD system is able to detect the discriminative texture features for cancer detection and localization and is a promising tool for improving the quality and efficiency of prostate cancer diagnosis.
Medical Physics 04/2015; 42(5):2368-2378. DOI:10.1118/1.4918318 · 2.64 Impact Factor