Weakly supervised volumetric prostate registration for MRI-TRUS image driven by signed distance map

Image registration is a fundamental step for MRI-TRUS fusion targeted biopsy. Due to the inherent representational differences between these two image modalities, though, intensity-based similarity losses for registration tend to result in poor performance. To mitigate this, comparison of organ segmentations, functioning as a weak proxy measure of image similarity, has been proposed. Segmentations, though, are limited in their information encoding capabilities. Signed distance maps (SDMs), on the other hand, encode these segmentations into a higher dimensional space where shape and boundary information are implicitly captured, and which, in addition, yield high gradients even for slight mismatches, thus preventing vanishing gradients during deep-network training. Based on these advantages, this study proposes a weakly-supervised deep learning volumetric registration approach driven by a mixed loss that operates both on segmentations and their corresponding SDMs, and which is not only robust to outliers, but also encourages optimal global alignment. Our experimental results, performed on a public prostate MRI-TRUS biopsy dataset, demonstrate that our method outperforms other weakly-supervised registration approaches with a dice similarity coefficient (DSC), Hausdorff distance (HD) and mean surface distance (MSD) of 87.3 ± 11.3, 4.56 ± 1.95 mm, and 0.053 ± 0.026 mm, respectively. We also show that the proposed method effectively preserves the prostate gland's internal structure.

Limitations of overlapping cores in systematic and MRI-US fusion biopsy

OBJECTIVES

To evaluate the clinically-significant prostate cancer (csCaP) detection rate of systematic (SBx) vs. targeted biopsy (TBx), after accounting for the overlapping systematic cores within the MRI regions of interest.

MATERIALS AND METHODS

We identified 398 consecutive men who underwent both transperineal systematic and targeted biopsy between January 2015 to January 2019. We reclassified overlapping systematic cores in the MRI regions of interest as target cores. The detection rates of SBx and TBx were compared using McNemar's test.

RESULTS

Detection rate of csCaP (grade group ≥2) was 42% (168/398). Median number of systematic and targeted cores were 23 (IQR 19-29) and 9 (IQR 6-12) respectively. A median of 3 (IQR 2-4) overlapping systematic cores were reclassified as targeted cores. After accounting for overlap, csPC detection rate on SBx decreased from 37% and 21% while the csCaP detection rate of TBx increased from 34% to 39% (both P < 0.001), with TBx having a better detection rate (39% vs. 21%, P < 0.001). A previous negative biopsy was associated with a lower risk of having csCaP on non-targeted SBx (OR 0.27, 95% CI: 0.12 - 0.58, P = 0.001). Only 5% (13/243) of those who had no cancer detected on TBx had csCaP on non-targeted SBx compared to 45% (70/155) of those who had csCaP on TBx (P< 0.001).

CONCLUSIONS

The utility of SBx in detecting csCaP decreases after accounting for overlap into the MRI region of interest, especially in men with a prior negative biopsy. Overlapping systematic cores improve the csCaP detection rate on TBx.

Multiparametric MRI-ultrasonography software fusion prostate biopsy: initial results using a stereotactic robotic-assisted transperineal prostate biopsy platform comparing systematic vs targeted biopsy

OBJECTIVE

To compare the detection rates of prostate cancer between systematic biopsy and targeted biopsy using a stereotactic robot-assisted transperineal prostate platform.

MATERIALS AND METHODS

We identified consecutive patients with suspicious lesion(s) on multiparametric magnetic resonance imaging (mpMRI), who underwent both systematic and MRI-transrectal ultrasonography (US) fusion targeted biopsy using our proprietary transperineal robot-assisted prostate biopsy platform between January 2015 and January 2019 at our institution, for retrospective analysis. Comparative analysis was performed between systematic and targeted biopsy using McNemar's test, and the cohort was further stratified by prior biopsy status and Prostate Imaging Reporting and Data System (PI-RADS) v2.0 score. International Society of Urological Pathology (ISUP) grade group (GG) ≥2 cancers (previously known as Gleason grade ≥7) were considered to be clinically significant.

RESULTS

A total of 500 patients were included in our final analysis, of whom 67 (13%) were patients with low-risk cancer on active surveillance. Of the 433 patients without prior diagnosis of cancer, 288 (67%) were biopsy-naïve. A total of 248 (57%) were diagnosed with prostate cancer, with 199 (46%) having clinically significant prostate cancer (ISUP GG ≥2). There were no statistically significant differences in the overall prostate cancer and clinically significant prostate cancer detection rate between systematic and targeted biopsy (51% vs 49% and 40% vs 38% respectively; P = 0.306 and P = 0.609). Of the 248 prostate cancers detected, 75% (187/248) were detected on both systematic and targeted biopsy, 14% (35/248) were detected on systematic biopsy alone and 11% (26/248) were detected on targeted biopsy alone. Of the 199 clinically significant cancers detected, 69% (138/199) were detected on both systematic and targeted biopsy, 17% (33/199) on systematic biopsy alone and 14% (28/199) on targeted biopsy alone. There were no statistically significant differences in the detection rate between systematic and targeted biopsy for both overall and clinically significant prostate cancer, even when the cohort was stratified by prior biopsy status and PI-RADS score. Targeted biopsy has greater sampling efficiency compared to systematic biopsy for both overall and clinically significant prostate cancer (23.2% vs 9.8%, P < 0.001 and 14.8% vs 5.6%, P < 0.001).

CONCLUSIONS

Using our robot-assisted transperineal prostate platform, combined MRI-US targeted biopsy with concurrent systematic prostate systematic biopsy probably represents the optimal method for the detection of clinically significant prostate cancer.

Multiparametric magnetic resonance imaging and clinical variables: Which is the best combination to predict reclassification in active surveillance patients?

Introduction & objectives

We tested the role of multiparametric magnetic resonance imaging (mpMRI) in disease reclassification and whether the combination of mpMRI and clinicopathological variables could represent the most accurate approach to predict the risk of reclassification during active surveillance.

Materials & methods

Three-hundred eighty-nine patients (pts) underwent mpMRI and subsequent confirmatory or follow-up biopsy according to the Prostate Cancer Research International Active Surveillance (PRIAS) protocol. Pts with negative (−) mpMRI underwent systematic random biopsy. Pts with positive (+) mpMRI [Prostate Imaging Reporting and Data System, version 2 (PI-RADS-V2) score ≥3] underwent targeted + systematic random biopsies. Multivariate analyses were used to create three models predicting the probability of reclassification [International Society of Urological Pathology ≥ Grade Group 2 (GG2)]: a basic model including only clinical variables (age, prostate-specific antigen density, and number of positive cores at baseline), an Magnetic resonance imaging (MRI) model including only the PI-RADS score, and a full model including both the previous ones. The predictive accuracy (PA) of each model was quantified using the area under the curve.

Results

mpMRI negative (−) was recorded in 127 (32.6%) pts; mpMRI positive (+) was recorded in 262 pts: 72 (18.5%) had PI-RADS 3, 150 (38.6%) PI-RADS 4, and 40 (10.3%) PI-RADS 5 lesions. At a median follow-up of 12 months, 125 pts (32%) were reclassified to GG2 prostate cancer. The rate of reclassification to GG2 prostate cancer was 17%, 35%, 38%, and 52% for mpMRI (−), PI-RADS 3, 4, and 5, respectively (P < 0.001). The PA was 69% and 64% in the basic and MRI models, respectively. The full model had the best PA of 74%: older age (P = 0.023; Odds ratio (OR) = 1.040), prostate-specific antigen density (P = 0.037; OR = 1.324), number of positive cores at baseline (P = 0.001; OR = 1.441), and PI-RADS 3, 4, and 5 (overall P = 0.001; OR = 2.458, 3.007, and 3.898, respectively) were independent predictors of reclassification.

Conclusions

Disease reclassification increased according to the PI-RADS score increase, at confirmatory or follow-up biopsy. However, a no-negligible rate of reclassification was found also in cases of mpMRI (−). The combination of mpMRI and clinicopathological variables still represents the most accurate approach to pts on active surveillance.

The Use of Multiparametric Magnetic Resonance Imaging for Follow-up of Patients Included in Active Surveillance Protocol. Can PSA Density Discriminate Patients at Different Risk of Reclassification?

INTRODUCTION

The objective of this study was to test Prostate Imaging Reporting and Data System (PI-RADS) classification on multiparametric magnetic resonance imaging (mpMRI) and MRI-derived prostate-specific antigen density (PSAD) in predicting the risk of reclassification in men in active surveillance (AS), who underwent confirmatory or per-protocol follow-up biopsy.

MATERIALS AND METHODS

Three hundred eighty-nine patients in AS underwent mpMRI before confirmatory or follow-up biopsy. Patients with negative (-) mpMRI underwent systematic random biopsy. Patients with positive (+) mpMRI underwent targeted fusion prostate biopsies + systematic random biopsies. Different PSAD cutoff values were tested (< 0.10, 0.10-0.20, ≥ 0.20). Multivariable analyses assessed the risk of reclassification, defined as clinically significant prostate cancer of grade group 2 or more, during follow-up according to PSAD, after adjusting for covariates.

RESULTS

One hundred twenty-seven (32.6%) patients had mpMRI(-); 72 (18.5%) had PI-RADS 3, 150 (38.6%) PI-RADS 4, and 40 (10.3%) PI-RADS 5 lesions. The rate of reclassification to grade group 2 PCa was 16%, 22%, 31%, and 39% for mpMRI(-) and PI-RADS 3, 4, and 5, respectively, in case of PSAD < 0.10 ng/mL²; 16%, 25%, 36%, and 44%, in case of PSAD 0.10 to 0.19 ng/mL²; and 25%, 42%, 55%, and 67% in case of PSAD ≥ 0.20 ng/mL². PSAD ≥ 0.20 ng/mL² (odds ratio [OR], 2.45; P = .007), PI-RADS 3 (OR, 2.47; P = .013), PI-RADS 4 (OR, 2.94; P < .001), and PI-RADS 5 (OR, 3.41; P = .004) were associated with a higher risk of reclassification.

CONCLUSION

PSAD ≥ 0.20 ng/mL² may improve predictive accuracy of mpMRI results for reclassification of patients in AS, whereas PSAD < 0.10 ng/mL² may help selection of patients at lower risk of harboring clinically significant prostate cancer. However, the risk of reclassification is not negligible at any PSAD cutoff value, also in the case of mpMRI(-).

A comparison of magnetic resonance imaging techniques used to secure biopsies in prostate cancer patients

Introduction: Prostate cancer (PCa) is the most common diagnosed malignancy among the male population in the United States. The incidence is increasing with an estimated amount of 175.000 cases in 2019. Areas covered: Primarily, PCa is generally detected by an elevated or rising serum prostate-specific antigen (PSA) and digital rectal examination (DRE) followed by pathological examination. Histopathology ultimately confirms the presence of PCa and determines a Gleason score. However, PSA and DRE have low specificity and sensitivity, respectively. Subsequently, accurate assessment of the aggressiveness of PCa is essential to prevent overdiagnosis and thus overtreatment of low-risk or indolent cancers. By visualizing PCa suspicious lesions and sampling them during the targeted biopsy, it is likely that the diagnostic accuracy of significant PCa improves. This article reviews the current imaging techniques used to secure biopsies in patients with a suspicion of PCa. The advantages and limitations of each technique are described. Expert opinion: Multiparametric magnetic resonance imaging (mpMRI) and subsequent-targeted biopsy have improved the diagnostic accuracy of PCa detection in men with an elevated or rising serum PSA. Prostate lesions visible on mpMRI are easily targeted during either in-bore MRI-guided biopsy, cognitive fusion biopsy or MRI-TRUS fusion biopsy.

Multicentre evaluation of magnetic resonance imaging supported transperineal prostate biopsy in biopsy-naïve men with suspicion of prostate cancer

OBJECTIVES

To analyse the detection rates of primary magnetic resonance imaging (MRI)-fusion transperineal prostate biopsy using combined targeted and systematic core distribution in three tertiary referral centres.

PATIENTS AND METHODS

In this multicentre, prospective outcome study, 807 consecutive biopsy-naïve patients underwent MRI-guided transperineal prostate biopsy, as the first diagnostic intervention, between 10/2012 and 05/2016. MRI was reported following the Prostate Imaging-Reporting and Data System (PI-RADS) criteria. In all, 236 patients had 18-24 systematic transperineal biopsies only, and 571 patients underwent additional targeted biopsies either by MRI-fusion or cognitive targeting if PI-RADS ≥3 lesions were present. Detection rates for any and Gleason score 7-10 cancer in targeted and overall biopsy were calculated and predictive values were calculated for different PI-RADS and PSA density (PSAD) groups.

RESULTS

Cancer was detected in 68% of the patients (546/807) and Gleason score 7-10 cancer in 49% (392/807). The negative predictive value of 236 PI-RADS 1-2 MRI in combination with PSAD of <0.1 ng/mL/mL for Gleason score 7-10 was 0.91 (95% confidence interval ± 0.07, 8% of study population). In 418 patients with PI-RADS 4-5 lesions using targeted plus systematic biopsies, the cancer detection rate of Gleason score 7-10 was significantly higher at 71% vs 59% and 61% with either approach alone (P < 0.001). For 153 PI-RADS 3 lesions, the detection rate was 31% with no significant difference to systematic biopsies with 27% (P > 0.05). Limitations include variability of multiparametric MRI (mpMRI) reading and Gleason grading.

CONCLUSION

MRI-based transperineal biopsy performed at high-volume tertiary care centres with a significant experience of prostate mpMRI and image-guided targeted biopsies yielded high detection rates of Gleason score 7-10 cancer. Prostate biopsies may not be needed for men with low PSAD and an unsuspicious MRI. In patients with high probability lesions, combined targeted and systematic biopsies are recommended.

Retrospective comparison of direct in-bore magnetic resonance imaging (MRI)-guided biopsy and fusion-guided biopsy in patients with MRI lesions which are likely or highly likely to be clinically significant prostate cancer

Purpose

To compare clinically significant prostate cancer (csPCa) detection rates between magnetic resonance imaging (MRI)–transrectal ultrasound (TRUS) fusion-guided prostate biopsy (FGB) and direct in-bore MRI-guided biopsy (MRGB).

Methods

We performed a comparison of csPCa detection rates between FGB and MRGB. Included patients had (1) at least one prior negative TRUS biopsy; (2) a Prostate Imaging Reporting and Data System (PI-RADS) 4 or 5 lesion and (3) a lesion size of ≥8 mm measured in at least one direction. We considered a Gleason score ≥7 being csPCa. Descriptive statistics with 95% confidence intervals (CI) were used to determine any differences.

Results

We included 51 patients with FGB (59 PI-RADS 4 and 41% PI-RADS 5) and 227 patients with MRGB (34 PI-RADS 4 and 66% PI-RADS 5). Included patients had a median age of 69 years (IQR, 65–72) and a median PSA level of 11.0 ng/ml (IQR, 7.4–15.1) and a median age of 67 years (IQR, 61–70), the median PSA 12.8 ng/ml (IQR, 9.1–19.0) within the FGB and the MRGB group, respectively. Detection rates of csPCA did not differ significantly between FGB and MRGB, 49 vs. 61%, respectively.

Conclusion

We did not detect significant differences between FGB and MRGB in the detection of csPCa. The differences in detection ratios between both biopsy techniques are narrow with an increasing lesion size. This study warrants further studies to optimize selection of best biopsy modality.

Magnetic Resonance and Ultrasound Image Fusion Supported Transperineal Prostate Biopsy Using the Ginsburg Protocol: Technique, Learning Points, and Biopsy Results

BACKGROUND

Prostate biopsy supported by transperineal image fusion has recently been developed as a new method to the improve accuracy of prostate cancer detection.

OBJECTIVE

To describe the Ginsburg protocol for transperineal prostate biopsy supported by multiparametric magnetic resonance imaging (mpMRI) and transrectal ultrasound (TRUS) image fusion, provide learning points for its application, and report biopsy results. The article is supplemented by a Surgery in Motion video.

DESIGN, SETTING, AND PARTICIPANTS

This single-centre retrospective outcome study included 534 patients from March 2012 to October 2015. A total of 107 had no previous prostate biopsy, 295 had benign TRUS-guided biopsies, and 159 were on active surveillance for low-risk cancer.

SURGICAL PROCEDURE

A Likert scale reported mpMRI for suspicion of cancer from 1 (no suspicion) to 5 (cancer highly likely). Transperineal biopsies were obtained under general anaesthesia using BiopSee fusion software (Medcom, Darmstadt, Germany). All patients had systematic biopsies, two cores from each of 12 anatomic sectors. Likert 3-5 lesions were targeted with a further two cores per lesion.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Any cancer and Gleason score 7-10 cancer on biopsy were noted. Descriptive statistics and positive predictive values (PPVs) and negative predictive values (NPVs) were calculated.

RESULTS AND LIMITATIONS

The detection rate of Gleason score 7-10 cancer was similar across clinical groups. Likert scale 3-5 MRI lesions were reported in 378 (71%) of the patients. Cancer was detected in 249 (66%) and Gleason score 7-10 cancer was noted in 157 (42%) of these patients. PPV for detecting 7-10 cancer was 0.15 for Likert score 3, 0.43 for score 4, and 0.63 for score 5. NPV of Likert 1-2 findings was 0.87 for Gleason score 7-10 and 0.97 for Gleason score ≥4+3=7 cancer. Limitations include lack of data on complications.

CONCLUSIONS

Transperineal prostate biopsy supported by MRI/TRUS image fusion using the Ginsburg protocol yielded high detection rates of Gleason score 7-10 cancer. Because the NPV for excluding Gleason score 7-10 cancer was very high, prostate biopsies may not be needed for all men with elevated prostate-specific antigen values and nonsuspicious mpMRI.

PATIENT SUMMARY

We present our technique to sample (biopsy) the prostate by the transperineal route (the area between the scrotum and the anus) to detect prostate cancer using a fusion of magnetic resonance and ultrasound images to guide the sampling.

Elastic Versus Rigid Image Registration in Magnetic Resonance Imaging-transrectal Ultrasound Fusion Prostate Biopsy: A Systematic Review and Meta-analysis

CONTEXT

The main difference between the available magnetic resonance imaging-transrectal ultrasound (MRI-TRUS) fusion platforms for prostate biopsy is the method of image registration being either rigid or elastic. As elastic registration compensates for possible deformation caused by the introduction of an ultrasound probe for example, it is expected that it would perform better than rigid registration.

OBJECTIVE

The aim of this meta-analysis is to compare rigid with elastic registration by calculating the detection odds ratio (OR) for both subgroups. The detection OR is defined as the ratio of the odds of detecting clinically significant prostate cancer (csPCa) by MRI-TRUS fusion biopsy compared with systematic TRUS biopsy. Secondary objectives were the OR for any PCa and the OR after pooling both registration techniques.

EVIDENCE ACQUISITION

The electronic databases PubMed, Embase, and Cochrane were systematically searched for relevant studies according to the Preferred Reporting Items for Systematic Review and Meta-analysis Statement. Studies comparing MRI-TRUS fusion and systematic TRUS-guided biopsies in the same patient were included. The quality assessment of included studies was performed using the Quality Assessment of Diagnostic Accuracy Studies version 2.

EVIDENCE SYNTHESIS

Eleven papers describing elastic and 10 describing rigid registration were included. Meta-analysis showed an OR of csPCa for elastic and rigid registration of 1.45 (95% confidence interval [CI]: 1.21-1.73, p<0.0001) and 1.40 (95% CI: 1.13-1.75, p=0.002), respectively. No significant difference was seen between the subgroups (p=0.83). Pooling subgroups resulted in an OR of 1.43 (95% CI: 1.25-1.63, p<0.00001).

CONCLUSIONS

No significant difference was identified between rigid and elastic registration for MRI-TRUS fusion-guided biopsy in the detection of csPCa; however, both techniques detected more csPCa than TRUS-guided biopsy alone.

PATIENT SUMMARY

We did not identify any significant differences in prostate cancer detection between two distinct magnetic resonance imaging-transrectal ultrasound fusion systems which vary in their method of compensating for prostate deformation.