Dataset Information

Machine learning in predicting extracapsular extension (ECE) of prostate cancer with MRI: a protocol for a systematic literature review.

ABSTRACT:

Introduction

In patients with prostate cancer (PCa), the detection of extracapsular extension (ECE) and seminal vesicle invasion is not only important for selecting the appropriate therapy but also for preoperative planning and patient prognosis. It is of paramount importance to stage PCa correctly before surgery, in order to achieve better surgical and outcome results. Over the last years, MRI has been incorporated in the classical prostate staging nomograms with clinical improvement accuracy in detecting ECE, but with variability between studies and radiologist's experience.

Methods and analysis

The research question, based on patient, index test, comparator, outcome and study design criteria, was the following: what is the diagnostic performance of artificial intelligence algorithms for predicting ECE in PCa patients, when compared with that of histopathological results after radical prostatectomy. To answer this question, we will use databases (EMBASE, PUBMED, Web of Science and CENTRAL) to search for the different studies published in the literature and we use the QUADA tool to evaluate the quality of the research selection.

Ethics and dissemination

This systematic review does not require ethical approval. The results will be disseminated through publication in a peer-review journal, as a chapter of a doctoral thesis and through presentations at national and international conferences.

Prospero registration number

CRD42020215671.

SUBMITTER: Guerra A

PROVIDER: S-EPMC9083401 | biostudies-literature | 2022 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning in predicting extracapsular extension (ECE) of prostate cancer with MRI: a protocol for a systematic literature review.

Guerra Adalgisa A Negrão Eduardo E Papanikolaou Nickolas N Donato Helena H

BMJ open 20220506 5

<h4>Introduction</h4>In patients with prostate cancer (PCa), the detection of extracapsular extension (ECE) and seminal vesicle invasion is not only important for selecting the appropriate therapy but also for preoperative planning and patient prognosis. It is of paramount importance to stage PCa correctly before surgery, in order to achieve better surgical and outcome results. Over the last years, MRI has been incorporated in the classical prostate staging nomograms with clinical improvement ac ...[more]

PMID: 35523484

Similar Datasets

Project description:The objective of this review is to survey radiomics signatures for detecting pathological extracapsular extension (pECE) on magnetic resonance imaging (MRI) in patients with prostate cancer (PCa) who underwent prostatectomy. Scientific Literature databases were used to search studies published from January 2007 to October 2023. All studies related to PCa MRI staging and using radiomics signatures to detect pECE after prostatectomy were included. Systematic review was performed according to Preferred Reporting Items for Systematic Review and Meta-analyses (PRISMA). The risk of bias and certainty of the evidence was assessed using QUADAS-2 and the radiomics quality score. From 1247 article titles screened, 16 reports were assessed for eligibility, and 11 studies were included in this systematic review. All used a retrospective study design and most of them used 3 T MRI. Only two studies were performed in more than one institution. The highest AUC of a model using only radiomics features was 0.85, for the test validation. The AUC for best model performance (radiomics associated with clinical/semantic features) varied from 0.72-0.92 and 0.69-0.89 for the training and validation group, respectively. Combined models performed better than radiomics signatures alone for detecting ECE. Most of the studies showed a low to medium risk of bias. After thorough analysis, we found no strong evidence supporting the clinical use of radiomics signatures for identifying extracapsular extension (ECE) in pre-surgery PCa patients. Future studies should adopt prospective multicentre approaches using large public datasets and combined models for detecting ECE.Critical relevant statementThe use of radiomics algorithms, with clinical and AI integration, in predicting extracapsular extension, could lead to the development of more accurate predictive models, which could help improve surgical planning and lead to better outcomes for prostate cancer patients.Protocol of systematic review registrationPROSPERO CRD42021272088. Published: https://doi.org/10.1136/bmjopen-2021-052342 .Key pointsRadiomics can extract diagnostic features from MRI to enhance prostate cancer diagnosis performance. The combined models performed better than radiomics signatures alone for detecting extracapsular extension. Radiomics are not yet reliable for extracapsular detection in PCa patients.

Project description:BackgroundExtracapsular extension (ECE) of prostate cancer (PCa) is closely related to the treatment and prognosis of patients, and radiomics has been widely used in the study of PCa. This study aimed to evaluate the value of a combined model considering magnetic resonance imaging (MRI)-based radiomics and clinical parameters for predicting ECE in PCa.MethodsA total of 392 PCa patients enrolled in this retrospective study were randomly divided into the training and validation sets at a ratio of 7:3. Radiologists assessed all lesions by Mehralivand grade. Radiomics features were extracted and selected to build a radiomics model, while clinical parameters were noted to construct the clinical model. The combined model was constructed by the integration of the radiomics model and clinical model. Meanwhile, the nomogram for predicting ECE was constructed based on the combined model. Then, the area under the receiver operating characteristic (ROC) curve (AUC), Delong test and the decision curve analysis (DCA) were used to compare the performance among the combined model, radiomics model, clinical model and Mehralivand grade.ResultsThe AUC of the combined model in the validation set was comparable to that of the radiomics model [AUC =0.894 (95% confidence interval (CI): 0.837-0.950) vs. 0.835 (95% CI: 0.763-0.908), P>0.05]. In addition, the sensitivity of the combined model and radiomics model was 90.7% and 77.8%, with an accuracy of 81.4% and 76.3%, respectively. On the other hand, the AUCs of the Mehralivand grade of radiologists and clinical model were 0.774 (95% CI: 0.691-0.857) and 0.749 (95% CI: 0.658-0.840), respectively, in the validation set, which were lower than those in the combined model (P<0.05). The DCA implied that the combined model could obtain the maximum net clinical benefits compared with the clinical model, the Mehralivand grade and radiomics model.ConclusionsThe combined model has a satisfactory predictive value for ECE in PCa patients compared with the clinical model, Mehralivand grade of radiologists, and the radiomics model.

Project description:BackgroundPrediction of extracapsular extension (ECE) is essential to achieve a balance between oncologic resection and neural tissue preservation. Microultrasound (MUS) is an attractive alternative to multiparametric magnetic resonance imaging (mpMRI) in the staging scenario.ObjectiveTo create a side-specific nomogram integrating clinicopathologic parameters and MUS findings to predict ipsilateral ECE and guide nerve sparing.Design setting and participantsProspective data were collected from consecutive patients who underwent robotic-assisted radical prostatectomy from June 2021 to May 2022 and had preoperative MUS and mpMRI. A total of 391 patients and 612 lobes were included in the analysis.Outcome measurements and statistical analysisECE on surgical pathology was the primary outcome. Multivariate regression analyses were carried out to identify predictors for ECE. The resultant multivariable model's performance was visualized using the receiver-operating characteristic curve. A nomogram was developed based on the coefficients of the logit function for the MUS-based model. A decision curve analysis (DCA) was performed to assess clinical utility.Results and limitationsThe areas under the receiver-operating characteristic curve (AUCs) of the MUS-based model were 81.4% and 80.9% (95% confidence interval [CI] 75.6, 84.6) after internal validation. The AUC of the mpMRI-model was also 80.9% (95% CI 77.2, 85.7). The DCA demonstrated the net clinical benefit of the MUS-based nomogram and its superiority compared with MUS and MRI alone for detecting ECE. Limitations of our study included its sample size and moderate inter-reader agreement.ConclusionsWe developed a side-specific nomogram to predict ECE based on clinicopathologic variables and MUS findings. Its performance was comparable with that of a mpMRI-based model. External validation and prospective trials are required to corroborate our results.Patient summaryThe integration of clinical parameters and microultrasound can predict extracapsular extension with similar results to models based on magnetic resonance imaging findings. This can be useful for tailoring the preservation of nerves during surgery.

Project description:Prostate cancer (PCa) is the most frequent male malignancy and the assessment of PCa aggressiveness, for which a biopsy is required, is fundamental for patient management. Currently, multiparametric (mp) MRI is strongly recommended before biopsy. Quantitative assessment of mpMRI might provide the radiologist with an objective and noninvasive tool for supporting the decision-making in clinical practice and decreasing intra- and inter-reader variability. In this view, high dimensional radiomics features and Machine Learning (ML) techniques, along with Deep Learning (DL) methods working on raw images directly, could assist the radiologist in the clinical workflow. The aim of this study was to develop and validate ML/DL frameworks on mpMRI data to characterize PCas according to their aggressiveness. We optimized several ML/DL frameworks on T2w, ADC and T2w+ADC data, using a patient-based nested validation scheme. The dataset was composed of 112 patients (132 peripheral lesions with Prostate Imaging Reporting and Data System (PI-RADS) score ≥ 3) acquired following both PI-RADS 2.0 and 2.1 guidelines. Firstly, ML/DL frameworks trained and validated on PI-RADS 2.0 data were tested on both PI-RADS 2.0 and 2.1 data. Then, we trained, validated and tested ML/DL frameworks on a multi PI-RADS dataset. We reported the performances in terms of Area Under the Receiver Operating curve (AUROC), specificity and sensitivity. The ML/DL frameworks trained on T2w data achieved the overall best performance. Notably, ML and DL frameworks trained and validated on PI-RADS 2.0 data obtained median AUROC values equal to 0.750 and 0.875, respectively, on unseen PI-RADS 2.0 test set. Similarly, ML/DL frameworks trained and validated on multi PI-RADS T2w data showed median AUROC values equal to 0.795 and 0.750, respectively, on unseen multi PI-RADS test set. Conversely, all the ML/DL frameworks trained and validated on PI-RADS 2.0 data, achieved AUROC values no better than the chance level when tested on PI-RADS 2.1 data. Both ML/DL techniques applied on mpMRI seem to be a valid aid in predicting PCa aggressiveness. In particular, ML/DL frameworks fed with T2w images data (objective, fast and non-invasive) show good performances and might support decision-making in patient diagnostic and therapeutic management, reducing intra- and inter-reader variability.

Project description:BackgroundMultiparametric MRI (mpMRI) is commonly recommended as a triage test prior to any prostate biopsy. However, there exists limited consensus on which patients with a negative prostate mpMRI could avoid prostate biopsy.PurposeTo identify which patient could safely avoid prostate biopsy when the prostate mpMRI is negative, via a radiomics-based machine learning approach.Study typeRetrospective.SubjectsThree hundred thirty patients with negative prostate 3T mpMRI between January 2016 and December 2018 were included.Field strength/sequenceA 3.0 T/T2-weighted turbo spin echo (TSE) imaging (T2 WI) and diffusion-weighted imaging (DWI).AssessmentThe integrative machine learning (iML) model was trained to predict negative prostate biopsy results, utilizing both radiomics and clinical features. The final study cohort comprised 330 consecutive patients with negative mpMRI (PI-RADS < 3) who underwent systematic transrectal ultrasound-guided (TRUS) or MR-ultrasound fusion (MRUS) biopsy within 6 months. A secondary analysis of biopsy naïve subcohort (n = 227) was also conducted.Statistical testsThe Mann-Whitney U test and Chi-Squared test were utilized to evaluate the significance of difference of clinical features between prostate biopsy positive and negative groups. The model performance was validated using leave-one-out cross-validation (LOOCV) and measured by AUC, sensitivity, specificity, and negative predictive value (NPV).ResultsOverall, 306/330 (NPV 92.7%) of the final study cohort patients had negative biopsies, and 207/227 (NPV 91.2%) of the biopsy naïve subcohort patients had negative biopsies. Our iML model achieved NPVs of 98.3% and 98.0% for the study cohort and subcohort, respectively, superior to prostate-specific antigen density (PSAD)-based risk assessment with NPVs of 94.9% and 93.9%, respectively.Data conclusionThe proposed iML model achieved high performance in predicting negative prostate biopsy results for patients with negative mpMRI. With improved NPVs, the proposed model can be used to stratify patients who in whom we might obviate biopsies, thus reducing the number of unnecessary biopsies.Evidence level3 TECHNICAL EFFICACY: Stage 2.

Project description:ObjectivesTo assess the added value of 3D T2-weighted imaging (T2WI) over conventional 2D T2WI in diagnosing extracapsular extension (ECE).MethodsSeventy-five patients undergoing 3-T MRI before radical prostatectomy were included. PI-RADS ≥ 4 lesions were assessed for ECE on 2D T2W images using a 5-point Likert scale (1 = no ECE, 5 = definite ECE) and the length of tumour prostatic capsular contact. A second read using 3D T2W images and reformats evaluated ECE and the maximal 3D capsular contact length and surface.ResultsOne hundred six lesions were identified at MRI. ECE was confirmed by histology in 54% (57/106) of lesions and 64% (48/75) of patients. Sensitivity and specificity for 3D T2 reads were 75.4% versus 64.9% (p = 0.058), respectively, and 83.7% versus 85.7% (p = 0.705) for 2D T2 reads, respectively. 3D T2W reads showed significantly higher mean subjective Likert scores of 3.7 ± 1.4 versus 3.3 ± 1.4 (p = 0.001) in ECE-positive lesions and lower mean Likert score of 1.5 ± 1 versus 1.6 ± 0.9 (p = 0.27) in ECE-negative lesions compared with 2D T2W reads. 3D contact significantly increased sensitivity from 59.6 to 73.7% (p = 0.03), whilst maintaining the same specificity of 87.8% (p = 1). High-grade group tumours (≥ Gleason 4 + 3) showed significantly higher ECE prevalence than low-grade tumours (88% versus 44%, p < 0.001) and a positive predictive value (PPV) for ECE of 90.9% with ≥ 5 mm of contact versus PPV of 90.4% at ≥ 12.5 mm for lower grade tumours.Conclusions3D T2WI significantly increases sensitivity and confidence in calling ECE. The capsular contact length threshold differed between low- and high-grade cancers.Key points• 3D capsular contact length and 3D surface contact significantly increased sensitivity in diagnosing ECE. • 3D T2W reads significantly increased reader confidence in calling ECE. • Thresholds for capsular contact length differed between low-grade and high-grade cancers.

Dataset Information

Machine learning in predicting extracapsular extension (ECE) of prostate cancer with MRI: a protocol for a systematic literature review.

Introduction

Methods and analysis

Ethics and dissemination

Prospero registration number

Publications

Machine learning in predicting extracapsular extension (ECE) of prostate cancer with MRI: a protocol for a systematic literature review.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets