ABSTRACT
OBJECTIVES: To identify factors influencing the diagnostic performance of the quantitative imaging biomarkers ADC and ADCratio in prostate cancer (PCa) detection. MATERIALS AND METHODS: A systematic literature search was conducted in Embase, Medline and Web of Science, for studies evaluating ADC values and ADCratio for PCa diagnosis, using the same patient cohorts and using histopathological references as ground truth. Pooled sensitivities, specificities, summary ROC curves and AUCs were calculated from constructed contingency data tables. Diagnostic performance (AUC) was quantitatively pooled using a bivariate mixed effects model. For identifying influencing factors, subgroup analysis, publication bias and heterogeneity assessment were investigated. RESULTS: Thirteen studies, involving 1038 patients and 1441 lesions, were included. For ADC, the pooled sensitivity and specificity was 80% (95% CI: 74-85%) and 78% (95% CI: 70-85%), respectively. For ADCratio pooled sensitivity and specificity was 80% (95% CI: 74-84%) and 80% (95% CI: 71-87%). Summary ROC analysis revealed AUCs of 0.86 (95% CI: 0.83-0.89) and 0.86 (95% CI: 0.83-0.89), respectively. Meta-regression showed heterogeneity between both imaging biomarkers. Subgroup analysis showed that ADCratio improved diagnostic performance in comparison to ADC when including both peripheral and transitional zone lesions (AUC: 0.87 [95% CI: 0.84-0.90] and 0.82 [95% CI: 0.79-0.85], respectively). CONCLUSION: Both ADC and ADCratio imaging biomarkers showed good and comparable diagnostic performance in PCa diagnosis. However, ADCratio shows better diagnostic performance than ADC in diagnosing transition zone cancers. CLINICAL RELEVANCE STATEMENT: In quantitative MRI-based PCa diagnosis, the imaging biomarker ADCratio is useful in challenging MRI readings of lesions. Understanding the performance of quantitative imaging biomarkers better can aid diagnostic MRI protocols, enhancing the precision of PCa assessments. KEY POINTS: MRI diffusion-weighted imaging-based ADC and ADCratio have comparable diagnostic performance in PCa assessment. In contrast to ADC, the ADCratio improves diagnostic performance, when assessing whole gland lesions. Compared to ADCratio, the ADC demonstrates enhanced diagnostic performance when evaluating peripheral zone lesions.
ABSTRACT
The classification of extramural vascular invasion status using baseline magnetic resonance imaging in rectal cancer has gained significant attention as it is an important prognostic marker. Also, the accurate prediction of patients achieving complete response with primary staging MRI assists clinicians in determining subsequent treatment plans. Most studies utilised radiomics-based methods, requiring manually annotated segmentation and handcrafted features, which tend to generalise poorly. We retrospectively collected 509 patients from 9 centres, and proposed a fully automated pipeline for EMVI status classification and CR prediction with diffusion weighted imaging and T2-weighted imaging. We applied nnUNet, a self-configuring deep learning model, for tumour segmentation and employed learned multiple-level image features to train classification models, named MLNet. This ensures a more comprehensive representation of the tumour features, in terms of both fine-grained detail and global context. On external validation, MLNet, yielding similar AUCs as internal validation, outperformed 3D ResNet10, a deep neural network with ten layers designed for analysing spatiotemporal data, in both CR and EMVI tasks. For CR prediction, MLNet showed better results than the current state-of-the-art model using imaging and clinical features in the same external cohort. Our study demonstrated that incorporating multi-level image representations learned by a deep learning based tumour segmentation model on primary MRI improves the results of EMVI classification and CR prediction with good generalisation to external data. We observed variations in the contributions of individual feature maps to different classification tasks. This pipeline has the potential to be applied in clinical settings, particularly for EMVI classification.