Search

OBJECTIVE: The objective of this study is to compare detection rates of extracapsular extension (ECE) of prostate cancer (PCa) using artificial intelligence (AI)-generated cancer maps versus MRI and conventional nomograms. MATERIALS AND METHODS: We retrospectively analysed data from 147 patients who received MRI-targeted biopsy and subsequent radical prostatectomy between September 2016 and May 2022. AI-based software cleared by the United States Food and Drug Administration (Unfold AI, Avenda Health) was used to map 3D cancer probability and estimate ECE risk. Conventional ECE predictors including MRI Likert scores, capsular contact length of MRI-visible lesions, PSMA T stage, Partin tables, and the PRedicting ExtraCapsular Extension nomogram were used for comparison.Postsurgical specimens were processed using whole-mount histopathology sectioning, and a genitourinary pathologist assessed each quadrant for ECE presence. ECE predictors were then evaluated on the patient (Unfold AI versus all comparators) and quadrant level (Unfold AI versus MRI Likert score). Receiver operator characteristic curves were generated and compared using DeLongs test. RESULTS: Unfold AI had a significantly higher area under the curve (AUC = 0.81) than other predictors for patient-level ECE prediction. Unfold AI achieved 68% sensitivity, 78% specificity, 71% positive predictive value, and 75% negative predictive value. At the quadrant level, Unfold AI exceeded the AUC of MRI Likert scores for posterior (0.89 versus 0.82, p = 0.003), anterior (0.84 versus 0.80, p = 0.34), and all quadrants (0.89 versus 0.82, p = 0.002). The false negative rate of Unfold AI was lower than MRI in both the anterior (-60%) and posterior prostate (-40%). CONCLUSIONS: Unfold AI accurately predicted ECE risk, outperforming conventional methodologies. It notably improved ECE prediction over MRI in posterior quadrants, with the potential to inform nerve-spare technique and prevent positive margins. By enhancing PCa staging and risk stratification, AI-based cancer mapping may lead to better oncological and functional outcomes for patients.

PURPOSE: Defining prostate cancer contours is a complex task, undermining the efficacy of interventions such as focal therapy. A multireader multicase study compared physicians' performance using artificial intelligence (AI) vs standard-of-care methods for tumor delineation. MATERIALS AND METHODS: Cases were interpreted by 7 urologists and 3 radiologists from 5 institutions with 2 to 23 years of experience. Each reader evaluated 50 prostatectomy cases retrospectively eligible for focal therapy. Each case included a T2-weighted MRI, contours of the prostate and region(s) of interest suspicious for cancer, and a biopsy report. First, readers defined cancer contours cognitively, manually delineating tumor boundaries to encapsulate all clinically significant disease. Then, after ≥ 4 weeks, readers contoured the same cases using AI software. Using tumor boundaries on whole-mount histopathology slides as ground truth, AI-assisted, cognitively-defined, and hemigland cancer contours were evaluated. Primary outcome measures were the accuracy and negative margin rate of cancer contours. All statistical analyses were performed using generalized estimating equations. RESULTS: The balanced accuracy (mean of voxel-wise sensitivity and specificity) of AI-assisted cancer contours (84.7%) was superior to cognitively-defined (67.2%) and hemigland contours (75.9%; P < .0001). Cognitively-defined cancer contours systematically underestimated cancer extent, with a negative margin rate of 1.6% compared to 72.8% for AI-assisted cancer contours (P < .0001). CONCLUSIONS: AI-assisted cancer contours reduce underestimation of prostate cancer extent, significantly improving contouring accuracy and negative margin rate achieved by physicians. This technology can potentially improve outcomes, as accurate contouring informs patient management strategy and underpins the oncologic efficacy of treatment.

Micro-ultrasound has recently been introduced as a low-cost alternative to multi-parametric MRI for imaging prostate cancer. Early clinical studies have demonstrated promising results; however, robust validation via comparison with whole-mount pathology has yet to be achieved. Due to micro-ultrasound probe design and tissue deformation during scanning, it is difficult to accurately correlate micro-ultrasound imaging planes with ground truth whole-mount pathology slides. In this study, we developed a multi-step methodology to co-register micro-ultrasound and MRI to whole-mount pathology. The three-step process had a registration error of 3.90 ± 0.11 mm and consists of: (1) micro-ultrasound image reconstruction, (2) 3D landmark registration of micro-ultrasound to MRI, and (3) 2D capsule registration of MRI to whole-mount pathology. This process was then used in a preliminary reader study to compare the diagnostic accuracy of micro-ultrasound and MRI in 15 patients who underwent radical prostatectomy for prostate cancer. Micro-ultrasound was found to have equivalent performance to retrospective MRI review for index lesion detection (91.7% vs. 80%), while demonstrating an increased detection of tumor extent (52.5% vs. 36.7%) with similar false positive regions-of-interest (38.3% vs. 40.8%). Prospective MRI review had reduced detection of index lesions (73.3%) and tumor extent (18.9%) but improved false positive regions-of-interest (22.7%) relative to micro-ultrasound and retrospective MRI. Further evaluation is needed with a larger sample size.

BackgroundPartial gland ablation (PGA) is a new option for treatment of prostate cancer (PCa). Cryotherapy, an early method of PGA, has had favorable evaluations, but few studies have employed a strict protocol using biopsy endpoints in men with clinically significant prostate cancer (csPCa).Methods143 men with unilateral csPCa were enrolled in a prospective, observational trial of outpatient PGA-cryotherapy. Treatment was a 2-cycle freeze of the affected prostate part. Participants were evaluated with MRI-guided biopsy (MRGB) at baseline and at 6 months and 18 months after treatment. Absence of csPCa upon MRGB was the primary endpoint; quality-of-life at baseline and at 6 months after treatment was assessed by EPIC-CP questionnaires in the domains of urinary and sexual function.ResultsOf the 143 participants, 136 (95%) completed MRGB at 6 months after treatment. In 103/136 (76%), the biopsy revealed no csPCa. Of the 103, 71 subsequently had an 18-month comprehensive biopsy; of the 71 with 18-month biopsies, 46 (65%) were found to have no csPCa. MRI lesions became undetectable in 96/130 (74%); declines in median serum PSA levels (6.9 to 2.5 ng/mL), PSA density (0.15 to 0.07), and prostate volume (42 to 34cc) were observed (all p

Focal laser ablation is a minimally invasive method of treating cancerous lesions in organs such as prostate, liver and brain. Oncologic control is achieved by inducing hyperthermia throughout the target while minimizing damage to surrounding tissue. Consequently, successful clinical outcomes are contingent upon achieving desired ablation volumes. Magnetic resonance thermometry is frequently used to monitor the formation of the induced thermal damage zone and inform the decision to terminate energy delivery. However, due to the associated cost and complexity there is growing interest in the development of alternative approaches. Here we investigate the utility of real-time interstitial interrogation of laser-tissue interaction as an inexpensive alternative monitoring modality that provides direct assessment of tissue coagulation without the need for organ specific calibration. The optical contrast mechanism was determined using a Monte Carlo model. Subsequently, four interstitial probe designs were manufactured and assessed in a tissue mimicking phantom under simultaneous magnetic resonance imaging. Finally, the optimal probe design was evaluated in ex vivo bovine muscle. It was found to be capable of providing sufficient feedback to achieve pre-defined ablation radii in the range 4-7 mm with a mean absolute error of 0.3 mm. This approach provides an inexpensive monitoring modality that may facilitate widespread adoption of focal laser ablation.

PurposeThe underlying premise of prostate cancer active surveillance (AS) is that cancers likely to metastasize will be recognized and eliminated before cancer-related disease can ensue. Our study was designed to determine the prostate cancer upgrading rate when biopsy guided by magnetic resonance imaging (MRGBx) is used before entry and during AS.Materials and methodsThe cohort included 519 men with low- or intermediate-risk prostate cancer who enrolled in prospective studies (NCT00949819 and NCT00102544) between February 2008 and February 2020. Subjects were preliminarily diagnosed with Gleason Grade Group (GG) 1 cancer; AS began when subsequent MRGBx confirmed GG1 or GG2. Participants underwent confirmatory MRGBx (targeted and systematic) followed by surveillance MRGBx approximately every 12 to 24 months. The primary outcome was tumor upgrading to ≥GG3.ResultsUpgrading to ≥GG3 was found in 92 men after a median followup of 4.8 years (IQR 3.1-6.5) after confirmatory MRGBx. Upgrade-free probability after 5 years was 0.85 (95% CI 0.81-0.88). Cancer detected in a magnetic resonance imaging lesion at confirmatory MRGBx increased risk of subsequent upgrading during AS (HR 2.8; 95% CI 1.3-6.0), as did presence of GG2 (HR 2.9; 95% CI 1.1-8.2) In men who upgraded ≥GG3 during AS, upgrading was detected by targeted cores only in 27%, systematic cores only in 25% and both in 47%. In 63 men undergoing prostatectomy, upgrading from MRGBx was found in only 5 (8%).ConclusionsWhen AS begins and follows with MRGBx (targeted and systematic), upgrading rate (≥GG3) is greater when tumor is initially present within a magnetic resonance imaging lesion or when pathology is GG2 than when these features are absent.

IntroductionA functional tool to optimize patient selection for magnetic resonance imaging (MRI)-guided prostate biopsy (MRGB) is an unmet clinical need. We sought to develop a prostate cancer risk calculator (PCRC-MRI) that combines MRI and clinical characteristics to aid decision-making for MRGB in North American men.MethodsTwo prospective registries containing 2354 consecutive men undergoing MRGB (September 2009 to April 2019) were analyzed. Patients were randomized into five groups, with one group randomly assigned to be the validation cohort against the other four groups as the discovery cohort. The primary outcome was detection of clinically significant prostate cancer (csPCa) defined as Gleason grade group ≥2. Variables included age, ethnicity, digital rectal exam (DRE), prior biopsy, prostate-specific antigen (PSA), prostate volume, PSA density, and MRI score. Odds ratios (OR) were calculated from multivariate logistic regression comparing two models: one with clinical variables only (clinical) against a second combining clinical variables with MRI data (clinical+MRI).ResultscsPCa was present in 942 (40%) of the 2354 men available for study. The positive and negative predictive values for csPCa in the clinical+MRI model were 57% and 89%, respectively. The area under the curve of the clinical+MRI model was superior to the clinical model in discovery (0.843 vs. 0.707, p

PurposeThe appropriate number of systematic biopsy cores to retrieve during magnetic resonance imaging (MRI)-targeted prostate biopsy is not well defined. We aimed to demonstrate a biopsy sampling approach that reduces required core count while maintaining diagnostic performance.Materials and methodsWe collected data from a cohort of 971 men who underwent MRI-ultrasound fusion targeted biopsy for suspected prostate cancer. A regional targeted biopsy (RTB) was evaluated retrospectively; only cores within 2 cm of the margin of a radiologist-defined region of interest were considered part of the RTB. We compared detection rates for clinically significant prostate cancer (csPCa) and cancer upgrading rate on final whole mount pathology after prostatectomy between RTB, combined, MRI-targeted, and systematic biopsy.ResultsA total of 16,459 total cores from 971 men were included in the study data sets, of which 1,535 (9%) contained csPCa. The csPCa detection rates for systematic, MRI-targeted, combined, and RTB were 27.0% (262/971), 38.3% (372/971), 44.8% (435/971), and 44.0% (427/971), respectively. Combined biopsy detected significantly more csPCa than systematic and MRI-targeted biopsy (p

ObjectiveTo demonstrate enabling multi-institutional training without centralizing or sharing the underlying physical data via federated learning (FL).Materials and methodsDeep learning models were trained at each participating institution using local clinical data, and an additional model was trained using FL across all of the institutions.ResultsWe found that the FL model exhibited superior performance and generalizability to the models trained at single institutions, with an overall performance level that was significantly better than that of any of the institutional models alone when evaluated on held-out test sets from each institution and an outside challenge dataset.DiscussionThe power of FL was successfully demonstrated across 3 academic institutions while avoiding the privacy risk associated with the transfer and pooling of patient data.ConclusionFederated learning is an effective methodology that merits further study to enable accelerated development of models across institutions, enabling greater generalizability in clinical use.

PurposeDeveloping large-scale datasets with research-quality annotations is challenging due to the high cost of refining clinically generated markup into high precision annotations. We evaluated the direct use of a large dataset with only clinically generated annotations in development of high-performance segmentation models for small research-quality challenge datasets.Materials and methodsWe used a large retrospective dataset from our institution comprised of 1,620 clinically generated segmentations, and two challenge datasets (PROMISE12: 50 patients, ProstateX-2: 99 patients). We trained a 3D U-Net convolutional neural network (CNN) segmentation model using our entire dataset, and used that model as a template to train models on the challenge datasets. We also trained versions of the template model using ablated proportions of our dataset, and evaluated the relative benefit of those templates for the final models. Finally, we trained a version of the template model using an out-of-domain brain cancer dataset, and evaluated the relevant benefit of that template for the final models. We used five-fold cross-validation (CV) for all training and evaluation across our entire dataset.ResultsOur model achieves state-of-the-art performance on our large dataset (mean overall Dice 0.916, average Hausdorff distance 0.135 across CV folds). Using this model as a pre-trained template for refining on two external datasets significantly enhanced performance (30% and 49% enhancement in Dice scores respectively). Mean overall Dice and mean average Hausdorff distance were 0.912 and 0.15 for the ProstateX-2 dataset, and 0.852 and 0.581 for the PROMISE12 dataset. Using even small quantities of data to train the template enhanced performance, with significant improvements using 5% or more of the data.ConclusionWe trained a state-of-the-art model using unrefined clinical prostate annotations and found that its use as a template model significantly improved performance in other prostate segmentation tasks, even when trained with only 5% of the original dataset.

ObjectiveTo detail the development of an electronic report that graphically conveys all relevant information from targeted prostate biopsy.MethodsThe Urology Integrated Diagnostic Report (Uro-IDR) is based on a published framework (RadPath) which enables the compilation of diagnostic data from urology, radiology, and pathology. Each component of the Uro-IDR is generated by the contributing clinician, is assembled in one document, and provides correlation of the 3 inputs at a glance. Upon completion, the Uro-IDR is automatically linked to the electronic medical record as an interactive file and can also be downloaded for offline sharing as a PDF.ResultsAt our institution, 1638 individual Uro-IDRs were generated between June 2016 and April 2019. There were 5715 views of these documents via the EMR. The average turnaround time for the creation of an individual report decreased from nearly 8 days at the time of its launch to 2 days after 6 months of use. The average time for report generation was 22 seconds for the pathologist and 69 seconds for the radiologist. An instructive video is linked to this article.ConclusionThe Uro-IDR has proven to be a feasible, efficient, clinically useful form to concisely transmit key information about targeted prostate biopsy to both clinicians and patients.

OBJECTIVES: To create reliable predictive metrics of unilateral disease using spatial tracking from a fusion device, thereby improving patient selection for hemi-gland ablation of prostate cancer. PATIENTS AND METHODS: We identified patients who received magnetic resonance imaging (MRI)/ultrasound-guided biopsy and radical prostatectomy at a single institution between 2011 and 2018. In addition to standard clinical features, we extracted quantitative features related to biopsy core and MRI target locations predictive of tumour unilaterality. Classification and Regression Tree (CART) analysis was used to create a decision tree (DT) for identifying cancer laterality. We evaluated concordance of model-determined laterality with final surgical pathology. RESULTS: A total of 173 patients were identified with biopsy coordinates and surgical pathology available. Based on CART analysis, in addition to biopsy- and MRI-confirmed disease unilaterality, patients should be further screened for cancer detected within 7 mm of midline in a 40 mL prostate, which equates to the central third of any-sized prostate by radius. The area under the curve for this DT was 0.82. Standard diagnostics and the DT correctly identified disease laterality in 73% and 80% of patients, respectively (P = 0.13). Of the patients identified as unilateral by standard diagnostics, 47% had undetected contralateral disease or were otherwise incorrectly identified. This error rate was reduced to 17% (P = 0.01) with the DT. CONCLUSION: Using spatial tracking from fusion devices, a DT was more reliable for identifying laterality of prostate cancer compared to standard diagnostics. Patients with cancer detected within the central third of the prostate by radius are poor hemi-gland ablation candidates due to the risk of midline extension of tumour.

BACKGROUND: Hemiablation is a less morbid treatment alternative for appropriately selected patients with unilateral prostate cancer (PCa). However, to the authors knowledge, traditional diagnostic techniques inadequately identify appropriate candidates. In the current study, the authors quantified the accuracy for identifying hemiablation candidates using contemporary diagnostic techniques, including multiparametric magnetic resonance imaging (mpMRI) and MRI-fusion with complete systematic template biopsy. METHODS: A retrospective analysis of patients undergoing MRI and MRI-fusion prostate biopsy, including full systematic template biopsy, prior to radical prostatectomy in a single tertiary academic institution between June 2010 and February 2018 was performed. Hemiablation candidates had unilateral intermediate-risk PCa (Gleason score [GS] of 3+4 or 4+3, clinical T classification ≤T2, and prostate-specific antigen level

ObjectiveTo study the relationship of maximum cancer core length (MCCL), on targeted biopsy (TB) of magnetic resonance imaging (MRI)-visible index lesions, to volume of that tumour found at radical prostatectomy (RP).Patients and methodsIn all, 205 men undergoing fusion biopsy and RP were divided into two groups: 136 in whom the MCCL came from an index MRI-visible lesion (TB) and 69 in whom MCCL came from a non-targeted lesion (non-targeted biopsy [NTB]). MRI was 3-T multi-parametric and biopsy was via MRI-ultrasonography fusion.ResultsIn the TB group, MCCL correlated with volume of clinically significant index tumours (ρ = 0.44-0.60, P 10 mm) and MRI lesion diameter (>20 mm) were independently associated with tumour volume. TB MCCLs >10 mm and Gleason scores >7 were each associated with pathological T3 disease (odds ratios 5.73 and 5.04, respectively), but MRI lesion diameter lesion was not.ConclusionsMCCL on a TB from an MRI-visible lesion is an independent predictor of both cancer volume and pathological stage. This relationship does not exist for MCCL from a NTB core. Quantifying CCL on MRI-TBs may have a value, not previously described, to risk-stratify patients with prostate cancer before treatment.

ObjectiveThis investigation was performed to evaluate the registration accuracy between magnetic resonance imaging (MRI) and pathology using three-dimensional (3-D) printed molds.MethodsTissue-mimicking prostate phantoms were manufactured with embedded fiducials. The fiducials were used to measure and compare target registration error (TRE) between phantoms that were sliced by hand versus phantoms that were sliced within 3-D-printed molds. Subsequently, ten radical prostatectomy specimens were placed inside molds, scanned with MRI, and then sliced. The ex vivo scan was used to assess the true location of whole mount (WM) slides relative to in vivo MRI. The TRE between WM and in vivo MRI was measured using anatomic landmarks.ResultsManually sliced phantoms had a 4.1-mm mean TRE, whereas mold-sliced phantoms had a 1.9-mm mean TRE. Similarly, mold-assisted slicing reduced mean angular misalignment around the left-right (LR) anatomic axis from 10.7° to 4.5°. However, ex vivo MRI revealed that excised prostates were misaligned within molds, including a mean 14° rotation about the LR axis. The mean in-plane TRE was 3.3 mm using molds alone and 2.2 mm after registration was corrected with ex vivo MRI.ConclusionPatient-specific molds improved accuracy relative to manual slicing techniques in a phantom model. However, the registration accuracy of surgically resected specimens was limited by their imperfect fit within molds. This limitation can be overcome with the addition of ex vivo imaging.SignificanceThe accuracy of 3-D-printed molds was characterized, quantifying their utility for facilitating MRI-pathology registration.

PurposeWe assessed focal therapy eligibility in men who underwent multiparametric magnetic resonance imaging and targeted biopsy with correlation to whole mount histology after radical prostatectomy.Materials and methodsSubjects were selected from among the 454 men in whom targeted biopsy proven prostate cancer was derived from regions of interest on multiparametric magnetic resonance imaging from 2010 to 2016. Focal therapy eligibility was limited to a maximum Gleason score of 4 + 3 in regions of interest with or without other foci of low risk prostate cancer (Gleason score 3 + 3 and less than 4 mm). Men who did not meet NCCN® intermediate risk criteria were classified as ineligible for focal therapy. Of the 454 men 64 underwent radical prostatectomy and biopsy findings were compared to final pathology findings.ResultsOf the 454 men with a biopsy proven region of interest 175 (38.5%) were eligible for focal therapy. Fusion biopsy, which combined targeted and template biopsy, had 80.0% sensitivity (12 of 15 cases), 73.5% specificity (36 of 49) and 75.0% accuracy (48 of 64) for focal therapy eligibility. Targeted cores alone yielded 73.3% sensitivity (11 of 15 cases), 47.9% specificity (23 of 48) and 54.7% accuracy (35 of 64). Gleason score and extension across the midline differed in 4 and 9, respectively, of the 13 cases that showed discordant biopsy and whole mount histology.ConclusionsUsing intermediate risk eligibility criteria more than a third of men with a targeted biopsy proven lesion identified on multiparametric magnetic resonance imaging would have been eligible for focal therapy. Eligibility determined by fusion biopsy was concordant with whole mount histology in 75% of cases. Improved selection criteria are needed to reliably determine focal therapy eligibility.

PurposeWe evaluated the accuracy of magnetic resonance imaging in determining the size and shape of localized prostate cancer.Materials and methodsThe subjects were 114 men who underwent multiparametric magnetic resonance imaging before radical prostatectomy with patient specific mold processing of the specimen from 2013 to 2015. T2-weighted images were used to contour the prostate capsule and cancer suspicious regions of interest. The contours were used to design and print 3-dimensional custom molds, which permitted alignment of excised prostates with magnetic resonance imaging scans. Tumors were reconstructed in 3 dimensions from digitized whole mount sections. Tumors were then matched with regions of interest and the relative geometries were compared.ResultsOf the 222 tumors evident on whole mount sections 118 had been identified on magnetic resonance imaging. For the 118 regions of interest mean volume was 0.8 cc and the longest 3-dimensional diameter was 17 mm. However, for matched pathological tumors, of which most were Gleason score 3 + 4 or greater, mean volume was 2.5 cc and the longest 3-dimensional diameter was 28 mm. The median tumor had a 13.5 mm maximal extent beyond the magnetic resonance imaging contour and 80% of cancer volume from matched tumors was outside region of interest boundaries. Size estimation was most accurate in the axial plane and least accurate along the base-apex axis.ConclusionsMagnetic resonance imaging consistently underestimates the size and extent of prostate tumors. Prostate cancer foci had an average diameter 11 mm longer and a volume 3 times greater than T2-weighted magnetic resonance imaging segmentations. These results may have important implications for the assessment and treatment of prostate cancer.

We have developed a system for evaluating magnetic resonance imaging of prostate cancer, using patient-specific 3D printed molds to facilitate MR-histology correlation. Prior to radical prostatectomy a patient receives a multiparametric MRI, which an expert genitourinary radiologist uses to identify and contour regions suspicious for disease. The same MR series is used to generate a prostate contour, which is the basis for design of a patient-specific mold. The 3D printed mold contains a series of evenly spaced parallel slits, each of which corresponds to a known MRI slice. After surgery, the patient's specimen is enclosed within the mold, and all whole-mount levels are obtained simultaneously through use of a multi-bladed slicing device. The levels are then formalin fixed, processed, and delivered to an expert pathologist, who identifies and grades all lesions within the slides. Finally, the lesion contours are loaded into custom software, which elastically warps them to fit the MR prostate contour. The suspicious regions on MR can then be directly compared to lesions on histology. Furthermore, the false-negative and false-positive regions on MR can be retrospectively examined, with the ultimate goal of developing methods for improving the predictive accuracy of MRI. This work presents the details of our analysis method, following a patient from diagnosis through the MR-histology correlation process. For this patient MRI successfully predicted the presence of cancer, but true lesion volume and extent were underestimated. Most cancer-positive regions missed on MR were observed to have patterns of low T2 signal, suggesting that there is potential to improve sensitivity.