Prostate segmentation in MRI using a convolutional neural network architecture and training strategy based on statistical shape models

Karimi, Davood; Samei, Golnoosh; Kesch, Claudia; Nir, Guy; Salcudean, Septimiu E.

doi:10.1007/s11548-018-1785-8

Cited by 78 publications

(59 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…CNN-based methods are relative new-comers to the field of medical image segmentation, but they have already proved to be highly versatile and effective [25], [3]. These methods usually produce a dense (i.e., pixel-or voxel-wise) segmentation probability map of the organ or volume of interest, although there are some exceptions [26], [27]. The early CNN-based image segmentation methods applied a soft-max function to the output layer activations and defined a loss function in terms of the negative log-likelihood, which is equivalent to a crossentropy loss for binary segmentation [28].…”

Section: Introductionmentioning

confidence: 99%

Reducing the Hausdorff Distance in Medical Image Segmentation With Convolutional Neural Networks

Karimi

Salcudean

2020

IEEE Trans. Med. Imaging

Self Cite

365

220

View full text Add to dashboard Cite

The Hausdorff Distance (HD) is widely used in evaluating medical image segmentation methods. However, existing segmentation methods do not attempt to reduce HD directly. In this paper, we present novel loss functions for training convolutional neural network (CNN)-based segmentation methods with the goal of reducing HD directly. We propose three methods to estimate HD from the segmentation probability map produced by a CNN. One method makes use of the distance transform of the segmentation boundary. Another method is based on applying morphological erosion on the difference between the true and estimated segmentation maps. The third method works by applying circular/spherical convolution kernels of different radii on the segmentation probability maps. Based on these three methods for estimating HD, we suggest three loss functions that can be used for training to reduce HD. We use these loss functions to train CNNs for segmentation of the prostate, liver, and pancreas in ultrasound, magnetic resonance, and computed tomography images and compare the results with commonly-used loss functions. Our results show that the proposed loss functions can lead to approximately 18 − 45% reduction in HD without degrading other segmentation performance criteria such as the Dice similarity coefficient. The proposed loss functions can be used for training medical image segmentation methods in order to reduce the large segmentation errors.

show abstract

Section: Introductionmentioning

confidence: 99%

Reducing the Hausdorff Distance in Medical Image Segmentation With Convolutional Neural Networks

Karimi

Salcudean

2020

IEEE Trans. Med. Imaging

Self Cite

365

220

View full text Add to dashboard Cite

show abstract

“…7 As a result, CNN-based auto-contouring systems for computed tomography (CT) images have been developed for various body sites, such as the head and neck, [8][9][10][11][12] thoracic region, [13][14][15][16] abdomen, [17][18][19] and pelvis. [20][21][22][23][24][25][26][27][28][29][30][31][32] Although these approaches have generally been very successful, they are not yet accessible to cancer treatment centers where they would be most usefulthose with limited resources that see a large number of cervical cancer patients, such as in South Africa and other low-and middle-income countries (LMICs). In fact, cervical cancer is the second most common cancer in women in Africa, 33,34 and the most cost-effective treatment that increases the survival rate of cervical cancer patients in LMICs is radiation treatment.…”

Section: Introductionmentioning

confidence: 99%

“…Although the potential of deep learning-based auto-contouring systems for pelvic structures has been explored in several previous studies, most of them were focused on prostate cancer, [21][22][23][24]31,32 and only a few papers have published results for the female pelvis. 25,26 In this study, we developed an auto-contouring system that can contour the clinical treatment volumes (CTVs) and normal structures that are necessary for various cervical cancer radiation treatment planning techniques.…”

Section: Introductionmentioning

confidence: 99%

Automatic contouring system for cervical cancer using convolutional neural networks

et al. 2020

View full text Add to dashboard Cite

To develop a tool for the automatic contouring of clinical treatment volumes (CTVs) and normal tissues for radiotherapy treatment planning in cervical cancer patients. Methods: An auto-contouring tool based on convolutional neural networks (CNN) was developed to delineate three cervical CTVs and 11 normal structures (seven OARs, four bony structures) in cervical cancer treatment for use with the Radiation Planning Assistant, a web-based automatic plan generation system. A total of 2254 retrospective clinical computed tomography (CT) scans from a single cancer center and 210 CT scans from a segmentation challenge were used to train and validate the CNN-based auto-contouring tool. The accuracy of the tool was evaluated by calculating the Sørensen-dice similarity coefficient (DSC) and mean surface and Hausdorff distances between the automatically generated contours and physician-drawn contours on 140 internal CT scans. A radiation oncologist scored the automatically generated contours on 30 external CT scans from three South African hospitals.

show abstract

“…It is worth noting that the trainable parameters in Resnet-50 in the network are pretrained on the ImageNet dataset [40]. The transfer learning from natural images to medical images is often used to improve the small-scale defects of prostate image datasets [41,42].…”

Section: Ra-fasterrcnnmentioning

confidence: 99%

CDA-Net for Automatic Prostate Segmentation in MR Images

Zhao

Pang

2020

Applied Sciences

View full text Add to dashboard Cite

Automatic and accurate prostate segmentation is an essential prerequisite for assisting diagnosis and treatment, such as guiding biopsy procedures and radiation therapy. Therefore, this paper proposes a cascaded dual attention network (CDA-Net) for automatic prostate segmentation in MRI scans. The network includes two stages of RAS-FasterRCNN and RAU-Net. Firstly, RAS-FasterRCNN uses improved FasterRCNN and sequence correlation processing to extract regions of interest (ROI) of organs. This ROI extraction serves as a hard attention mechanism to focus the segmentation of the subsequent network on a certain area. Secondly, the addition of residual convolution block and self-attention mechanism in RAU-Net enables the network to gradually focus on the area where the organ exists while making full use of multiscale features. The algorithm was evaluated on the PROMISE12 and ASPS13 datasets and presents the dice similarity coefficient of 92.88% and 92.65%, respectively, surpassing the state-of-the-art algorithms. In a variety of complex slice images, especially for the base and apex of slice sequences, the algorithm also achieved credible segmentation performance.

show abstract

Prostate segmentation in MRI using a convolutional neural network architecture and training strategy based on statistical shape models

Cited by 78 publications

References 17 publications

Reducing the Hausdorff Distance in Medical Image Segmentation With Convolutional Neural Networks

Reducing the Hausdorff Distance in Medical Image Segmentation With Convolutional Neural Networks

Automatic contouring system for cervical cancer using convolutional neural networks

CDA-Net for Automatic Prostate Segmentation in MR Images

Contact Info

Product

Resources

About