Development of a MRI-Based Radiomics Nomogram for Prediction of Response of Patients With Muscle-Invasive Bladder Cancer to Neoadjuvant Chemotherapy

Zhang, Xinxin; Wang, Yichen; Zhang, Jin; Zhang, Lianyu; Wang, Sicong; Chen, Yan

doi:10.3389/fonc.2022.878499

Cited by 10 publications

(19 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The combined model reached an accuracy of 93% in terms of differentiating muscle invasion preoperatively. Finally, radiomic features have the potential of predicting neoadjuvant chemotherapy response, defined as ≤ pT1 on the pathological report from radical cystectomy ( 20 ). Based on aggressiveness patterns derived from T2WI, ADC and DWI, combined with clinical staging, the proposed nomogram reached sensitivity, specificity, and accuracy of 94.4%, 94.1% and 94.3%, respectively.…”

Section: Discussionmentioning

confidence: 99%

Fully automated bladder tumor segmentation from T2 MRI images using 3D U-Net algorithm

Coroamă

Dioşan²,

Telecan³

et al. 2023

Front. Oncol.

View full text Add to dashboard Cite

IntroductionBladder magnetic resonance imaging (MRI) has been recently integrated in the diagnosis pathway of bladder cancer. However, automatic recognition of suspicious lesions is still challenging. Thus, development of a solution for proper delimitation of the tumor and its separation from the healthy tissue is of primordial importance. As a solution to this unmet medical need, we aimed to develop an artificial intelligence-based decision support system, which automatically segments the bladder wall and the tumor as well as any suspect area from the 3D MRI images.MaterialsWe retrospectively assessed all patients diagnosed with bladder cancer, who underwent MRI at our department (n=33). All examinations were performed using a 1.5 Tesla MRI scanner. All images were reviewed by two radiologists, who performed manual segmentation of the bladder wall and all lesions. First, the performance of our fully automated end-to-end segmentation model based on a 3D U-Net architecture (by considering various depths of 4, 5 or 6 blocks) trained in two data augmentation scenarios (on 5 and 10 augmentation datasets per original data, respectively) was tested. Second, two learning setups were analyzed by training the segmentation algorithm with 7 and 14 MRI original volumes, respectively.ResultsWe obtained a Dice-based performance over 0.878 for automatic segmentation of bladder wall and tumors, as compared to manual segmentation. A larger training dataset using 10 augmentations for 7 patients could further improve the results of the U-Net-5 model (0.902 Dice coefficient at image level). This model performed best in terms of automated segmentation of bladder, as compared to U-Net-4 and U-Net-6. However, in this case increased time for learning was needed as compared to U-Net-4. We observed that an extended dataset for training led to significantly improved segmentation of the bladder wall, but not of the tumor.ConclusionWe developed an intelligent system for bladder tumors automated diagnostic, that uses a deep learning model to segment both the bladder wall and the tumor. As a conclusion, low complexity networks, with less than five-layers U-Net architecture are feasible and show good performance for automatic 3D MRI image segmentation in patients with bladder tumors.

show abstract

Section: Discussionmentioning

confidence: 99%

Fully automated bladder tumor segmentation from T2 MRI images using 3D U-Net algorithm

Coroamă

Dioşan²,

Telecan³

et al. 2023

Front. Oncol.

View full text Add to dashboard Cite

show abstract

“…In recent years, radiomics has become a popular method for preoperative prediction of bladder tumor pathology, thanks to advances in medical technology, and has achieved remarkable success in that regard. 12 , 17 , 18 , 19 Several researchers have applied CT, MRI, and ultrasound to radiomic studies for predicting histological grading and staging features of BLCA and have demonstrated good predictive efficacy. 20 , 21 However, in most studies, only predictive nomogram model that contained imaging features was constructed, and models for other histologies were not.…”

Section: Discussionmentioning

confidence: 99%

Value of the application of computed tomography‐based radiomics for preoperative prediction of unfavorable pathology in initial bladder cancer

et al. 2023

View full text Add to dashboard Cite

ObjectivesTo construct and validate unfavorable pathology (UFP) prediction models for patients with the first diagnosis of bladder cancer (initial BLCA) and to compare the comprehensive predictive performance of these models.Materials and MethodsA total of 105 patients with initial BLCA were included and randomly enrolled into the training and testing cohorts in a 7:3 ratio. The clinical model was constructed using independent UFP‐risk factors determined by multivariate logistic regression (LR) analysis in the training cohort. Radiomics features were extracted from manually segmented regions of interest in computed tomography (CT) images. The optimal CT‐based radiomics features to predict UFP were determined by the optimal feature filter and the least absolute shrinkage and selection operator algorithm. The radiomics model consist with the optimal features was constructed by the best of the six machine learning filters. The clinic‐radiomics model combined the clinical and radiomics models via LR. The area under the curve (AUC), accuracy, sensitivity, specificity, positive and negative predictive value, calibration curve and decision curve analysis were used to evaluate the predictive performance of the models.ResultsPatients in the UFP group had a significantly older age (69.61 vs. 63.93 years, p = 0.034), lager tumor size (45.7% vs. 11.1%, p = 0.002) and higher neutrophil to lymphocyte ratio (NLR; 2.76 vs. 2.33, p = 0.017) than favorable pathologic group in the training cohort. Tumor size (OR, 6.02; 95% CI, 1.50–24.10; p = 0.011) and NLR (OR, 1.50; 95% CI, 1.05–2.16; p = 0.026) were identified as independent predictive factors for UFP, and the clinical model was constructed using these factors. The LR classifier with the best AUC (0.817, the testing cohorts) was used to construct the radiomics model based on the optimal radiomics features. Finally, the clinic‐radiomics model was developed by combining the clinical and radiomics models using LR. After comparison, the clinic‐radiomics model had the best performance in comprehensive predictive efficacy (accuracy = 0.750, AUC = 0.817, the testing cohorts) and clinical net benefit among UFP‐prediction models, while the clinical model (accuracy = 0.625, AUC = 0.742, the testing cohorts) was the worst.ConclusionOur study demonstrates that the clinic‐radiomics model exhibits the best predictive efficacy and clinical net benefit for predicting UFP in initial BLCA compared with the clinical and radiomics model. The integration of radiomics features significantly improves the comprehensive performance of the clinical model.

show abstract

“…CT imaging has a fundamental role in the diagnosis, staging, treatment guidance, and response monitoring in BCa [ 3 ]. As imaging-based features can be easily obtained non-invasively at low costs, radiomics has been increasingly evaluated for initial detection, grading, and local as well as nodal staging of BCa patients [ 24 , 25 , 26 , 27 , 28 , 29 , 30 ], for recurrence assessment [ 31 , 32 , 33 ] and for the response evaluation to neoadjuvant preoperative as well as adjuvant treatments in recent years [ 34 , 35 , 36 , 37 , 38 , 39 , 40 ]. However, to our knowledge, no imaging-based machine-learning models have been investigated for outcome prediction of BCa patients before RC.…”

Section: Discussionmentioning

confidence: 99%

Radiomics for the Prediction of Overall Survival in Patients with Bladder Cancer Prior to Radical Cystectomy

Woźnicki

Laqua

Messmer

et al. 2022

Cancers

View full text Add to dashboard Cite

(1) Background: To evaluate radiomics features as well as a combined model with clinical parameters for predicting overall survival in patients with bladder cancer (BCa). (2) Methods: This retrospective study included 301 BCa patients who received radical cystectomy (RC) and pelvic lymphadenectomy. Radiomics features were extracted from the regions of the primary tumor and pelvic lymph nodes as well as the peritumoral regions in preoperative CT scans. Cross-validation was performed in the training cohort, and a Cox regression model with an elastic net penalty was trained using radiomics features and clinical parameters. The models were evaluated with the time-dependent area under the ROC curve (AUC), Brier score and calibration curves. (3) Results: The median follow-up time was 56 months (95% CI: 48–74 months). In the follow-up period from 1 to 7 years after RC, radiomics models achieved comparable predictive performance to validated clinical parameters with an integrated AUC of 0.771 (95% CI: 0.657–0.869) compared to an integrated AUC of 0.761 (95% CI: 0.617–0.874) for the prediction of overall survival (p = 0.98). A combined clinical and radiomics model stratified patients into high-risk and low-risk groups with significantly different overall survival (p < 0.001). (4) Conclusions: Radiomics features based on preoperative CT scans have prognostic value in predicting overall survival before RC. Therefore, radiomics may guide early clinical decision-making.

show abstract

Development of a MRI-Based Radiomics Nomogram for Prediction of Response of Patients With Muscle-Invasive Bladder Cancer to Neoadjuvant Chemotherapy

Cited by 10 publications

References 30 publications

Fully automated bladder tumor segmentation from T2 MRI images using 3D U-Net algorithm

Fully automated bladder tumor segmentation from T2 MRI images using 3D U-Net algorithm

Value of the application of computed tomography‐based radiomics for preoperative prediction of unfavorable pathology in initial bladder cancer

Radiomics for the Prediction of Overall Survival in Patients with Bladder Cancer Prior to Radical Cystectomy

Contact Info

Product

Resources

About