CWCheng Wei

Cheng Wei

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Cheng Wei has a total of 56 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Subsequent Detection of Clinically Significant Prostate Cancer After Suspicious MRI and an Initial Biopsy Negative for csPCa: A Longitudinal Subgroup Analysis of the MULTIPROS Trial

    by Wafa D. Aloufi, Magdalena Szewczyk‐Bieda, Ghulam Nabi, Xinyu Zhang, Yilong Zhang, Cheng Wei, Luigi Manfredi

    Background: Management after suspicious multiparametric MRI (mpMRI; PI‐RADS ≥ 3) and an initial biopsy negative for clinically significant prostate cancer (csPCa) remains uncertain. We described the frequency of subsequent csPCa detection and explored baseline factors. Methods: This post hoc longitudinal analysis of MULTIPROS trial data included 153 men with PI‐RADS ≥ 3 lesions and no csPCa detected on baseline transrectal systematic biopsy, with or without targeted biopsy. Participants contributed follow‐up from the date of the initial mpMRI until csPCa detection or their last available clinical follow‐up, with records reviewed through May 2025. Follow‐up was not protocol‐mandated, and repeat biopsy was guided by clinical findings. The primary outcome was time to histological csPCa detection. Kaplan–Meier and multivariable logistic regression analyses were exploratory. Results: Over a median observed time to csPCa detection or censoring of 64 months, csPCa was detected in 37/153 men (24.2%). Of these, 28/37 (75.7%) were detected within 6 months, and the median time to detection was 3 months. This early clustering was compatible with possible baseline sampling limitations, although timing alone could not establish this mechanism. Overall, 84/153 men underwent repeat biopsy and 69/153 did not; 69/116 men without detected csPCa lacked repeat histological confirmation. The between‐group Kaplan–Meier comparison was not significant ( p = 0.210) and was difficult to interpret because repeat‐biopsy rates differed. Transition‐zone location, anterior fibromuscular stroma location, and atypical small acinar proliferation were associated with detection. The hypothesis‐generating model had moderate apparent discrimination (AUC 0.756) but low sensitivity at a 0.50 threshold (24.3%). Conclusions: Most detections occurred early, a pattern compatible with possible baseline sampling limitations, although the biological timing of disease onset cannot be established from these data. Selective follow‐up limits the interpretation of the detection proportion and between‐group comparison. The identified associations require external validation, and the model should not be used clinically.

  • Open Access

    Baseline Whole-Gland MRI Radiomics for Risk Stratification After Suspicious mpMRI and Negative Biopsy: A MULTIPROS Sub-Analysis

    by Wafa D. Aloufi, Magdalena Szewczyk-Bieda, Asma Aloufi, Cheng Wei, Luigi Manfredi

    Background/Objectives: Patients with suspicious prostate mpMRI but negative baseline biopsy remain challenging to manage, as some are later diagnosed with clinically significant prostate cancer (csPCa). This study evaluated whether baseline whole-gland radiomic features from T2-weighted imaging and apparent diffusion coefficient (ADC) maps could predict subsequent csPCa detection. Methods: This retrospective radiomics sub-analysis of the MULTIPROS trial included men with suspicious baseline mpMRI findings (PI-RADS v2.0 categories 3–5) and no csPCa detected at baseline biopsy. The primary outcome was subsequent csPCa detection during follow-up, defined as ISUP Grade Group ≥2. Whole-gland prostate segmentation was performed on baseline T2-weighted images using a semi-automated approach with manual refinement, and the resulting T2-derived masks were applied to the spatially corresponding ADC maps. Radiomic features were extracted using PyRadiomics version 3.1.0a2 after standardised preprocessing. Feature robustness was assessed in a 20-patient reproducibility subset using intra- and inter-reader intraclass correlation coefficients, followed by correlation filtering and LASSO-based feature prioritisation. Radiomics-only, clinicoradiological-only, and clinicoradiological–radiomic logistic regression models were evaluated using stratified five-fold cross-validation. Results: T2-based modelling included 151 patients, of whom 36/151 (23.8%) had subsequent csPCa detected; the median time to csPCa detection or last follow-up was 64 months. ADC analyses included 144 patients, of whom 34/144 (23.6%) had subsequent csPCa detected. The T2 clinicoradiological–radiomic model achieved a mean AUC of 0.696 ± 0.060, compared with 0.713 ± 0.072 for the T2 clinicoradiological-only model and 0.456 ± 0.091 for the T2 radiomics-only model. In the ADC-available cohort, ADC radiomics-only, ADC clinicoradiological–radiomic, and combined T2 + ADC clinicoradiological–radiomic models achieved mean AUCs of 0.684 ± 0.052, 0.635 ± 0.157, and 0.663 ± 0.106, respectively. Calibration was imperfect, with deviations at higher predicted probabilities and overprediction in the highest predicted-risk bin. Conclusions: Baseline whole-gland MRI radiomics showed limited and inconsistent predictive value for subsequent csPCa detection after negative biopsy and did not demonstrate consistent incremental improvement over clinicoradiological predictors. These findings are exploratory and hypothesis-generating, and external validation is required before clinical use.

  • Open Access

    Design and Characterisation of a Polyvinyl Chloride (PVC) Tissue-Mimicking Polymer Phantom for Quantitative Shear Wave Elastography Validation

    by Wadhhah Aldehani, Sarah Louise Savaridas, Cheng Wei, Luigi Manfredi

    A polyvinyl chloride (PVC)-based tissue-mimicking polymer phantom was developed and mechanically characterised to replicate stiffness ranges relevant to breast elastography and to provide a controlled platform for evaluating shear wave elastography (SWE) measurements. SWE provides quantitative stiffness information that complements B-mode ultrasound in breast imaging. However, measurement variability related to operator technique and tissue continues to limit confidence in clinical interpretation. This study evaluates the reproducibility of SWE using custom-fabricated PVC-based breast phantoms with mechanically defined stiffness properties. Two PVC-based breast phantoms with identical geometry and different background stiffnesses were scanned using a single ultrasound system under a fixed SWE protocol. Each phantom contained four embedded inclusions representing clinically relevant stiffness categories. Six breast imagers independently acquired repeated SWE measurements in transverse and longitudinal planes, blinded to lesion identity and ground truth. Inter-operator reproducibility was assessed using intraclass correlation coefficients, and was high across both phantom backgrounds, with low intra-operator variability following quality assurance exclusion of one dataset due to sampling error. Measurement variability was lowest for solid inclusions and increased for the cyst-like inclusion in the stiffer background. SWE measurements consistently preserved the relative stiffness ordering of inclusions, although absolute values differed systematically from mechanically derived ground-truth stiffness. These findings demonstrate that PVC-based polymer phantoms provide a stable and reproducible platform for evaluating SWE measurement behaviour under controlled conditions. By isolating operator and acquisition effects from biological variability, this polymer-based framework supports methodological standardisation and structured operator training in breast elastography.

  • Open Access

    Long‐Term Observed Risk of Clinically Significant Prostate Cancer After Negative mpMRI and Systematic Biopsy in Biopsy‐Naïve Men: A MULTIPROS Subanalysis

    by Wafa D. Aloufi, Xinyu Zhang, Magdalena Szewczyk-Bieda, Ghulam Nabi, Cheng Wei, Luigi Manfredi

    Background: mpMRI is now widely used before prostate biopsy and has improved the detection of clinically significant prostate cancer (csPCa). However, the subsequent risk of csPCa among men with both negative mpMRI and negative systematic biopsy (SBx) remains uncertain. Purpose: To evaluate the observed medium‐term risk of csPCa in biopsy‐naïve men with initially negative mpMRI and negative SBx. Methods: This post hoc longitudinal follow‐up subanalysis included biopsy‐naïve men from the prospective multicenter MULTIPROS trial who had negative baseline mpMRI (PI‐RADS ≤ 2) and negative SBx. Of the 582 enrolled men, 169 had negative mpMRI; 138 underwent baseline SBx, of whom 8 had csPCa, and 31 had < 12 months of follow‐up, leaving 99 men for analysis. Follow‐up was PSA‐based, with repeat mpMRI and biopsy performed when clinically indicated. Kaplan–Meier analysis estimated csPCa‐free survival. Results: During a mean follow‐up of 61.8 ± 22.1 months, four patients were diagnosed with csPCa (4/99; 4.0%). The mean time from initial negative mpMRI to csPCa diagnosis was 68.3 ± 22.4 months. Repeat biopsy was performed in 16 patients. csPCa‐free survival was 100% at 24 months and 97.3% at 60 months (95% CI: 93.7%–100%). Conclusion: In this selected cohort of biopsy‐naïve men with both negative mpMRI and negative SBx, the observed incidence of csPCa during medium‐term follow‐up was low but based on only four events. Because repeat imaging and biopsy were selectively triggered rather than systematically performed, the true incidence of csPCa may be underestimated. Continued PSA‐based monitoring, with repeat mpMRI and biopsy when clinically indicated, remains important.

  • Open Access

    Machine Learning Model Development for Malignant Prostate Lesion Prediction Using Texture Analysis Features from Ultrasound Shear-Wave Elastography

    by Adel Jawli, Ghulam Nabi, Zhihong Huang, Abeer J. Alhusaini, Cheng Wei, Benjie Tang

    Introduction: Artificial intelligence (AI) is increasingly utilized for texture analysis and the development of machine learning (ML) techniques to enhance diagnostic accuracy. ML algorithms are trained to differentiate between normal and malignant conditions based on provided data. Texture feature analysis, including first-order and second-order features, is a critical step in ML development. This study aimed to evaluate quantitative texture features of normal and prostate cancer tissues identified through ultrasound B-mode and shear-wave elastography (SWE) imaging and to develop and assess ML models for predicting and classifying normal versus malignant prostate tissues. Methodology: First-order and second-order texture features were extracted from B-mode and SWE imaging, including four reconstructed regions of interest (ROIs) from SWE images for normal and malignant tissues. A total of 94 texture features were derived, including features for intensity, Gray-Level Co-Occurrence Matrix (GLCM), Gray-Level Dependence Length Matrix (GLDLM), Gray-Level Run Length Matrix (GLRLM), and Gray-Level Size Zone Matrix (GLSZM). Five ML models were developed and evaluated using 5-fold cross-validation to predict normal and malignant tissues. Results: Data from 62 patients were analyzed. All ROIs, except those derived from B-mode imaging, exhibited statistically significant differences in features between normal and malignant tissues. Among the developed models, Support Vector Machines (SVM), Random Forest (RF), and Naive Bayes (NB) demonstrated the highest performance across all ROIs. These models consistently achieved strong predictive accuracy for classifying normal versus malignant tissues. Gray Pure SWE and Gray Reconstructed images Provided the highest sensitivity and specificity in PCa prediction by 82%, 90%, and 98%, 96%, respectively. Conclusions: Texture analysis with machine learning on SWE-US and reconstructed images effectively differentiates malignant from benign prostate lesions, with features like contrast, entropy, and correlation playing a key role. Random Forest, SVM, and Naïve Bayes showed the highest classification performance, while grayscale reconstructions (GPSWE and GRRI) enhanced detection accuracy.


  • Open Access

    A Comparative Analysis of U-Net and Vision Transformer Architectures in Semi-Supervised Prostate Zonal Segmentation

    by Guantian Huang, Bixuan Xia, Haoming Zhuang, Bohan Yan, Cheng Wei, Shouliang Qi, Wei Qian, Dianning He

    The precise segmentation of different regions of the prostate is crucial in the diagnosis and treatment of prostate-related diseases. However, the scarcity of labeled prostate data poses a challenge for the accurate segmentation of its different regions. We perform the segmentation of different regions of the prostate using U-Net- and Vision Transformer (ViT)-based architectures. We use five semi-supervised learning methods, including entropy minimization, cross pseudo-supervision, mean teacher, uncertainty-aware mean teacher (UAMT), and interpolation consistency training (ICT) to compare the results with the state-of-the-art prostate semi-supervised segmentation network uncertainty-aware temporal self-learning (UATS). The UAMT method improves the prostate segmentation accuracy and provides stable prostate region segmentation results. ICT plays a more stable role in the prostate region segmentation results, which provides strong support for the medical image segmentation task, and demonstrates the robustness of U-Net for medical image segmentation. UATS is still more applicable to the U-Net backbone and has a very significant effect on a positive prediction rate. However, the performance of ViT in combination with semi-supervision still requires further optimization. This comparative analysis applies various semi-supervised learning methods to prostate zonal segmentation. It guides future prostate segmentation developments and offers insights into utilizing limited labeled data in medical imaging.


  • Multicenter Randomized Trial Assessing MRI and Image-guided Biopsy for Suspected Prostate Cancer: The MULTIPROS Study

    by Cheng Wei, Magdalena Szewczyk-Bieda, Anthony S. Bates, Peter T. Donnan, Petra Rauchhaus, Stephen Gandy, Senthil Kumar Arcot Ragupathy, Paras Singh, Katherine Coll, Jonathan Serhan, et al.

    Background The optimal diagnostic pathway for prostate cancer (PCa) is evolving, requiring further evaluation in a randomized controlled trial. Purpose To assess the diagnostic accuracy of prebiopsy multiparametric MRI in the identification of clinically significant PCa (csPCa) using radical prostatectomy (RP) specimens as the reference standard, and to test the diagnostic accuracy of combined US and MRI fusion-targeted biopsy with systematic biopsies. Materials and Methods In a prospective randomized controlled trial including university hospitals, men with suspected PCa were recruited between January 2015 and August 2020 to assess the diagnostic accuracy of multiparametric MRI before biopsy in detection of csPCa at biopsy and RP histopathologic structure (primary outcome). Men with lesions suspicious for cancer (Prostate Imaging and Reporting Data System [PI-RADS] ≥3) at multiparametric MRI were first randomized to either systematic random prostate biopsies alone (control group) or US and MRI fusion-targeted biopsies with systematic random prostate biopsies (intervention group) at a one-to-one ratio to compare the diagnostic accuracy of systematic random versus combined fusion with systematic random biopsies (secondary outcome). A subset of recruited participants (n = 89) underwent RP and histologic sectioning. Results There were 582 participants who were eligible to undergo multiparametric MRI (mean age, 65 years ± 6 [SD]). In total, 413 had a PI-RADS score of at least 3 and were randomized into either the intervention group (207 of 413; 50.1%) or control group (206 of 413; 49.9%). The csPCa detection rate in the intervention group was higher, with an adjusted odds ratio of 1.79 (95% CI: 1.14, 2.79; P = .01). A subgroup of 89 men underwent RP (21.5%; 89 of 413). Multiparametric MRI helped correctly identify 131 of 182 csPCa foci in 89 men (sensitivity, 72%; 95% CI: 65, 78). The specificity, positive predictive value, and negative predictive value were 71% (91 of 128), 78% (131 of 168), and 64% (91 of 142), respectively. Conclusion Prebiopsy multiparametric MRI was accurate in the depiction of clinically significant PCa. Combining US and MRI fusion-targeted biopsies with systematic biopsies helped detect more clinically significant lesions than did systematic biopsies alone. Clinical trial registration no. NCT02745496 © RSNA, 2023 Supplemental material is available for this article.


  • Open Access

    Radiomics Approach to the Detection of Prostate Cancer Using Multiparametric MRI: A Validation Study Using Prostate-Cancer-Tissue-Mimicking Phantoms

    by Faisal Alshomrani, Basim S. O. Alsaedi, Cheng Wei, Magdalena Szewczyk-Bieda, Stephen Gandy, Jennifer Wilson, Zhihong Huang, Ghulam Nabi

    Over the last few years, a number of studies have quantified the role of radiomics, dynamic contrast enhancement and standard MRI (T2WI + DWI) in detecting prostate cancer; however, the aim of this paper was to assess the advantage of combining radiomics with other multiparametric magnetic resonance imaging (mpMRI) (T2-DWI-DCE) in improving the detection of prostate cancer. This study used 10 prostate-cancer-tissue-mimicking phantoms to obtain preclinical data. We then focused on 46 patients who underwent mpMRI and Transrectal Ultrasound (TRUS) guided biopsy between September 2016 and December 2017. The texture analysis parameters combined with the mpMRI and compared with the histopathology of TRUS biopsy have been assessed statistically by principal component analysis (PCA) and discriminant component analysis (DCA). The prediction model and goodness-of-fit were examined with the Akaike information criterion (AIC) and McFadden pseudo-R-squared. In the PCA, there was a higher separation between cancerous and noncancerous tissue in the preclinical compared with the clinical data. Both AIC and R2 showed an improvement in the model in cancer prediction by adding the radiomics to mpMRI. The discriminant analysis showed an accuracy of cancer prediction of 81% compared with 100% in the pre-clinical phantom data. Combining radiomics with mpMRI showed an improvement in prostate cancer prediction. The ex vivo experiments validated the findings of this study.


  • Open Access

    Cost-Effectiveness of Prostate Cancer Detection in Biopsy-Naïve Men: Ultrasound Shear Wave Elastography vs. Multiparametric Diagnostic Magnetic Resonance Imaging

    by Mehdi Shiva, Cheng Wei, Hassan Molana, Ghulam Nabi

    This exploratory study investigates the cost-effectiveness of ultrasound shear wave elastography (SWE) imaging in comparison to pre-biopsy multiparametric magnetic resonance imaging (mpMRI) in men with suspected prostate cancer. This research is motivated by the early evidence of the good performance of SWE in distinguishing cancerous from benign prostate tissues. We used a decision analysis model representing the care-pathways of men referred with a high prostate specific antigen (PSA) and/or abnormal digital rectal examination (DRE) in a UK setting from the payer’s perspective with results reported in 2016 GBP. We then appraised the cost-effectiveness of a novel approach based on SWE compared to the more conventional and widely practiced mpMRI-based approaches using data reported in the literature. Deterministic and probabilistic sensitivity analyses were used to address uncertainty regarding the parameter values utilised. Our exploratory results implied that SWE approach yielded an additional quality-adjusted life year (QALY) at the cost of GBP 10,048 compared to the standard mpMRI-based approach in the UK. This is lower than the official willingness to pay threshold of GBP 20,000 (the UK healthcare system guidelines) and is therefore a suitable replacement for the current practice. Sensitivity analyses confirmed the robustness of our results.


  • Open Access

    Predicting the Performance of Concurrent Systematic Random Biopsies during Image Fusion Targeted Sampling of Multi-Parametric MRI Detected Prostate Cancer. A Prospective Study (PRESET Study)

    by Saeed Alqahtani, Xinyu Zhang, Cheng Wei, Yilong Zhang, Magdalena Szewczyk-Bieda, Jennifer Wilson, Zhihong Huang, Ghulam Nabi

    The study was aimed to develop a predictive model to identify patients who may benefit from performing systematic random biopsies (SB) in addition to targeted biopsies (TB) in men suspected of having prostate cancer. A total of 198 patients with positive pre-biopsy MRI findings and who had undergone both TB and SB were prospectively recruited into this study. The primary outcome was detection rates of clinically significant prostate cancer (csPCa) in SB and TB approaches. The secondary outcome was net clinical benefits of SB in addition to TB. A logistic regression model and nomogram construction were used to perform a multivariate analysis. The detection rate of csPCa using SB was 51.0% (101/198) compared to a rate of 56.1% (111/198) for TB, using a patient-based biopsy approach. The detection rate of csPCa was higher using a combined biopsy (64.6%; 128/198) in comparison to TB (56.1%; 111/198) alone. This was statistically significant (p < 0.001). Age, PSA density and PIRADS score significantly predicted the detection of csPCa by SB in addition to TB. A nomogram based on the model showed good discriminative ability (C-index; 78%). The decision analysis curve confirmed a higher net clinical benefit at an acceptable threshold.