CNN-Based Prostate Zonal Segmentation on T2-Weighted MR Images: A Cross-Dataset Study

Rundo, Leonardo; Han, Changhee; Zhang, Jin; Hataya, Ryuichiro; Nagano, Yudai; Militello, Carmelo; Ferretti, Claudio; Nobile, Marco S.; Tangherloni, Andrea; Gilardi, Maria Carla; Vitabile, Salvatore; Nakayama, Hideki; Mauri, Giancarlo

doi:10.1007/978-981-13-8950-4_25

Cited by 35 publications

(28 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Before conducting the comparison, we evaluate the segmentation performance of our backbone network. To the best of our knowledge, Tian et al [26] and Rundo et al [9] achieved the state-of-the-art segmentation performance on the same three datasets that we utilize. We directly reference their reported numbers in our paper to demonstrate that our implemented segmentation backbones are valid.…”

Section: Effectiveness Of Our Multi-site Learning Methodsmentioning

confidence: 91%

“…We conduct pre-processing for the three datasets. Following [9], we first center-cropped the images from Site C with roughly same view as images from the other two sites, since the raw images of Site C are scanned from whole body, rather than prostate surrounding area. We then resized all samples of site A, B and C with size of 384 × 384 in axial plane.…”

Section: A Datasets and Evaluation Metricmentioning

confidence: 99%

“…John et al [8] aggregate two different prostate MRI datasets, and present that augmenting training data with heterogeneous images cannot bring explicit improvement. Rundo et al [9] conduct extensive experiments on inter-site prostate MRIs, and find that directly mixing two datasets leads to performance degradation. Similar observations have also been reported on other medical image analysis tasks [10], [11].…”

Section: Introductionmentioning

confidence: 99%

“…To compensate for the inter-site heterogeneity, we propose a Domain-Specific Batch Normalization (DSBN) layer in the network backbone, enabling the network to estimate statistics and perform feature normalization for each site individually. Based on previous observations [9], it is not straight-forward for a network to explore robust representations from multiple datasets. Therefore, we propose a novel learning paradigm, i.e., Multi-site-guided Knowledge Transfer (MSKT) to enhance the learning of shared kernels by conducting knowledge transfer from multiple auxiliary branches.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

MS-Net: Multi-Site Network for Improving Prostate Segmentation With Heterogeneous MRI Data

Liu

Dou

et al. 2020

IEEE Trans. Med. Imaging

204

154

View full text Add to dashboard Cite

Automated prostate segmentation in MRI is highly demanded for computer-assisted diagnosis. Recently, a variety of deep learning methods have achieved remarkable progress in this task, usually relying on large amounts of training data. Due to the nature of scarcity for medical images, it is important to effectively aggregate data from multiple sites for robust model training, to alleviate the insufficiency of single-site samples. However, the prostate MRIs from different sites present heterogeneity due to the differences in scanners and imaging protocols, raising challenges for effective ways of aggregating multi-site data for network training. In this paper, we propose a novel multisite network (MS-Net) for improving prostate segmentation by learning robust representations, leveraging multiple sources of data. To compensate for the inter-site heterogeneity of different MRI datasets, we develop Domain-Specific Batch Normalization layers in the network backbone, enabling the network to estimate statistics and perform feature normalization for each site separately. Considering the difficulty of capturing the shared knowledge from multiple datasets, a novel learning paradigm, i.e., Multi-site-guided Knowledge Transfer, is proposed to enhance the kernels to extract more generic representations from multi-site data. Extensive experiments on three heterogeneous prostate MRI datasets demonstrate that our MS-Net improves the performance across all datasets consistently, and outperforms state-of-the-art methods for multi-site learning.

show abstract

Section: Effectiveness Of Our Multi-site Learning Methodsmentioning

confidence: 91%

Section: A Datasets and Evaluation Metricmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

MS-Net: Multi-Site Network for Improving Prostate Segmentation With Heterogeneous MRI Data

Liu

Dou

et al. 2020

IEEE Trans. Med. Imaging

204

154

View full text Add to dashboard Cite

show abstract

“…However, no literature method so far coped with the generalization ability among multi-institutional MRI datasets, making their clinical applicability difficult [53]. In a previous work [54], we compared existing CNN-based architectures-namely, SegNet [51], U-Net [44], and pix2pix [52]-on two multi-institutional MRI datasets. According to our results, U-Net generally achieves the most accurate performance.…”

Section: Related Workmentioning

confidence: 99%

USE-Net: Incorporating Squeeze-and-Excitation blocks into U-Net for prostate zonal segmentation of multi-institutional MRI datasets

et al. 2019

Self Cite

View full text Add to dashboard Cite

Prostate cancer is the most common malignant tumors in men but prostate Magnetic Resonance Imaging (MRI) analysis remains challenging. Besides whole prostate gland segmentation, the capability to differentiate between the blurry boundary of the Central Gland (CG) and Peripheral Zone (PZ) can lead to differential diagnosis, since the frequency and severity of tumors differ in these regions. To tackle the prostate zonal segmentation task, we propose a novel Convolutional Neural Network (CNN), called USE-Net, which incorporates Squeeze-and-Excitation (SE) blocks into U-Net, i.e., one of the most effective CNNs in biomedical image segmentation. Especially, the SE blocks are added after every Encoder (Enc USE-Net) or Encoder-Decoder block (Enc-Dec USE-Net). This study evaluates the generalization ability of CNN-based architectures on three T2-weighted MRI datasets, each one consisting of a different number of patients and heterogeneous image characteristics, collected by different institutions. The following mixed scheme is used for training/testing: (i ) training on either each individual dataset or multiple prostate MRI datasets and (ii ) testing on all three datasets with all possible training/testing combinations. USE-Net is compared against three stateof-the-art CNN-based architectures (i.e., U-Net, pix2pix, and Mixed-Scale Dense Network), along with a semi-automatic continuous max-flow model. The results show that training on the union of the datasets generally outperforms training on each dataset separately, allowing for both intra-/crossdataset generalization. Enc USE-Net shows good overall generalization under any training condition, while Enc-Dec USE-Net remarkably outperforms the other methods when trained on all datasets. These findings reveal that the SE blocks' adaptive feature recalibration provides excellent cross-dataset generalization when testing is performed on samples of the datasets used during training. Therefore, we should consider multi-dataset training and SE

show abstract

Automatic Multiparametric Magnetic Resonance Imaging‐Based Prostate Lesions Assessment with Unsupervised Domain Adaptation

Dai

Wang

et al. 2023

Advanced Intelligent Systems

View full text Add to dashboard Cite

Multiparametric magnetic resonance imaging (mpMRI) has emerged as a valuable diagnostic tool in prostate lesion assessment. However, training convolutional neural networks (CNNs) inevitably involves magnetic resonance (MR) images from multiple cohorts. There always exists variation in scanning protocol among cohorts, inducing significant changes in data distribution between source and target domains. This challenge has greatly limited clinical adoption on a large scale. Herein, a coarse mask‐guided deep domain adaptation network (CMD2A‐Net) is proposed to develop a fully automated framework for prostate lesion detection and classification (PLDC). No category or mask label is required from the target domain. A coarse segmentation module is trained to cover the possible lesion‐related regions, so that attention maps can be generated to dedicate the local feature extraction of lesions within those regions. Experiments are performed on 512 mpMRI sets from datasets of PROSTATEx (330 sets) and two cohorts, A (74 sets) and B (108 sets). Using ensemble learning, CMD2A‐Net accomplishes an AUC of 0.921 in cohort A and 0.913 in cohort B, demonstrating its transferability from a large‐scale public dataset PROSTATEx to small‐scale target domains. Results from an ablation study also support its effectiveness in classification between benign and malignant lesions, compared to the state‐of‐the‐art models. An interactive preprint version of the article can be found here: https://doi.org/10.22541/au.166081031.11420810/v1.

show abstract

CNN-Based Prostate Zonal Segmentation on T2-Weighted MR Images: A Cross-Dataset Study

Cited by 35 publications

References 37 publications

MS-Net: Multi-Site Network for Improving Prostate Segmentation With Heterogeneous MRI Data

MS-Net: Multi-Site Network for Improving Prostate Segmentation With Heterogeneous MRI Data

USE-Net: Incorporating Squeeze-and-Excitation blocks into U-Net for prostate zonal segmentation of multi-institutional MRI datasets

Automatic Multiparametric Magnetic Resonance Imaging‐Based Prostate Lesions Assessment with Unsupervised Domain Adaptation

Contact Info

Product

Resources

About