• Corpus ID: 226226750

Encoding Clinical Priori in 3D Convolutional Neural Networks for Prostate Cancer Detection in bpMRI

@article{Saha2020EncodingCP,
  title={Encoding Clinical Priori in 3D Convolutional Neural Networks for Prostate Cancer Detection in bpMRI},
  author={Anindo Saha and Matin Hosseinzadeh and Henkjan J. Huisman},
  journal={ArXiv},
  year={2020},
  volume={abs/2011.00263}
}
We hypothesize that anatomical priors can be viable mediums to infuse domain-specific clinical knowledge into state-of-the-art convolutional neural networks (CNN) based on the U-Net architecture. We introduce a probabilistic population prior which captures the spatial prevalence and zonal distinction of clinically significant prostate cancer (csPCa), in order to improve its computer-aided detection (CAD) in bi-parametric MR imaging (bpMRI). To evaluate performance, we train 3D adaptations of… 
View With Semantic Reader
4 Citations
Methods Citations
1

Figures from this paper

4 Citations

Citation Type

Citation Type

Deep learning for fully automatic detection, segmentation, and Gleason grade estimation of prostate cancer in multiparametric magnetic resonance images
TLDR
A fully automatic system based on Deep Learning that performs localization, segmentation and Gleason grade group (GGG) estimation of PCa lesions from prostate mpMRIs that achieves an excellent lesion-level AUC/sensitivity/specificity.
Anatomy-aided deep learning for medical image segmentation: a review
TLDR
A review of anatomy-aided DL for medical image segmentation which covers systematically summarized anatomical information categories and corresponding representation methods and presents a categorized methodology overview on using anatomical information with DL from over 70 papers.
End-to-end Prostate Cancer Detection in bpMRI via 3D CNNs: Effect of Attention Mechanisms, Clinical Priori and Decoupled False Positive Reduction
Tailoring automated data augmentation to H&E-stained histopathology
TLDR
This work builds upon the RandAugment framework and introduces several domain-specific modifications relevant to histopathological images, increasing generalizability, and outperforms the state-of-the-art manually engineered data augmentation strategy.

References

SHOWING 1-10 OF 28 REFERENCES
Effect of Adding Probabilistic Zonal Prior in Deep Learning-based Prostate Cancer Detection.
TLDR
Results show that fusing zonal prior knowledge improves the baseline detection model with a preference for probabilistic over deterministic zonal segmentation.
Classification of Cancer at Prostate MRI: Deep Learning versus Clinical PI-RADS Assessment.
TLDR
U-Net trained with T2-weighted and diffusion MRI achieves similar performance to clinical Prostate Imaging Reporting and Data System assessment in the task of detection and segmentation of lesions suspicious for sPC.
Prostate Cancer Inference via Weakly-Supervised Learning using a Large Collection of Negative MRI
TLDR
The baseline MRI model is proposed to alternatively learn the appearance of mp-MRI using radiology-confirmed negative MRI cases via weakly supervised learning and to infer the pixel-wise suspiciousness of PCa by comparing the original and synthesized MRI with two distance functions.
Joint Prostate Cancer Detection and Gleason Score Prediction in mp-MRI via FocalNet
TLDR
A novel multi-class CNN, FocalNet, is proposed to jointly detect PCa lesions and predict their aggressiveness using Gleason score (GS), which characterizes lesion aggressiveness and fully utilizes distinctive knowledge from mp-MRI.
USE-Net: incorporating Squeeze-and-Excitation blocks into U-Net for prostate zonal segmentation of multi-institutional MRI datasets
Large scale deep learning for computer aided detection of mammographic lesions
Anatomical Priors in Convolutional Networks for Unsupervised Biomedical Segmentation
TLDR
A generative probabilistic model is introduced that employs the learned prior through a convolutional neural network to compute segmentations in an unsupervised setting to facilitate CNN-based anatomical segmentation in a range of novel clinical problems, where few or no annotations are available and thus standard networks are not trainable.
Attention Gated Networks: Learning to Leverage Salient Regions in Medical Images
DeepNAT: Deep convolutional neural network for segmenting neuroanatomy
Multiparametric Magnetic Resonance Imaging for the Detection of Clinically Significant Prostate Cancer: What Urologists Need to Know. Part 2: Interpretation.
...
1
2
3
...