Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9434022
I. Sirazitdinov, M. Lenga, Ivo M. Baltruschat, D. Dylov, A. Saalbach
The placement of a central venous catheter (CVC) for venous access is a common clinical routine. Nonetheless, various clinical studies report that CVC insertions are unsuccessful in up to 20% of all cases. Among other, typical complications include the incidence of a pneumothorax, hemothorax, arterial puncture, venous air embolism, arrhythmias or catheter knotting. In order to detect the CVC tip in chest X-ray (CXR) images, and to evaluate the catheter placement, we propose a HRNet-based key point detection approach in combination with a probabilistic constellation model. In a cross-validation study, we show that our approach not only enables the exact localization of the CVC tip, but also of relevant anatomical landmarks. Moreover, the probabilistic model provides a likelihood score for tip position which allows us to identify malpositioned CVCs.
{"title":"Landmark Constellation Models For Central Venous Catheter Malposition Detection","authors":"I. Sirazitdinov, M. Lenga, Ivo M. Baltruschat, D. Dylov, A. Saalbach","doi":"10.1109/ISBI48211.2021.9434022","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9434022","url":null,"abstract":"The placement of a central venous catheter (CVC) for venous access is a common clinical routine. Nonetheless, various clinical studies report that CVC insertions are unsuccessful in up to 20% of all cases. Among other, typical complications include the incidence of a pneumothorax, hemothorax, arterial puncture, venous air embolism, arrhythmias or catheter knotting. In order to detect the CVC tip in chest X-ray (CXR) images, and to evaluate the catheter placement, we propose a HRNet-based key point detection approach in combination with a probabilistic constellation model. In a cross-validation study, we show that our approach not only enables the exact localization of the CVC tip, but also of relevant anatomical landmarks. Moreover, the probabilistic model provides a likelihood score for tip position which allows us to identify malpositioned CVCs.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"26 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114039539","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9434007
Tianqing Li, Leihao Wei, William Hsu
While quantitative image features (radiomic) can be employed as informative indicators of disease progression, they are sensitive to variations in acquisition and reconstruction. Prior studies have demonstrated the ability to normalize heterogeneous scans using per-pixel metrics (e.g., mean squared error) and qualitative reader studies. However, the generalizability of these techniques and the impact of normalization on downstream tasks (e.g., classification) have been understudied. We present a multi-pronged evaluation by assessing image normalization techniques using 1) per-pixel image quality and perceptual metrics, 2) variability in radiomic features, and 3) task performance differences using a machine learning (ML) model. We evaluated a previously reported 3D generative adversarial network-based (GAN) approach, investigating its performance on low-dose computed tomography (CT) scans acquired at a different institution with varying dose levels and reconstruction kernels. While the 3D GAN achieved superior metric results, its impact on quantitative image features and downstream task performance did not result in universal improvement. These results suggest a more complicated relationship between CT acquisition and reconstruction parameters and their effect on radiomic features and ML model performance, which are not fully captured using per-pixel metrics alone. Our approach provides a more comprehensive picture of the effect of normalization.
{"title":"A Multi-Pronged Evaluation For Image Normalization Techniques","authors":"Tianqing Li, Leihao Wei, William Hsu","doi":"10.1109/ISBI48211.2021.9434007","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9434007","url":null,"abstract":"While quantitative image features (radiomic) can be employed as informative indicators of disease progression, they are sensitive to variations in acquisition and reconstruction. Prior studies have demonstrated the ability to normalize heterogeneous scans using per-pixel metrics (e.g., mean squared error) and qualitative reader studies. However, the generalizability of these techniques and the impact of normalization on downstream tasks (e.g., classification) have been understudied. We present a multi-pronged evaluation by assessing image normalization techniques using 1) per-pixel image quality and perceptual metrics, 2) variability in radiomic features, and 3) task performance differences using a machine learning (ML) model. We evaluated a previously reported 3D generative adversarial network-based (GAN) approach, investigating its performance on low-dose computed tomography (CT) scans acquired at a different institution with varying dose levels and reconstruction kernels. While the 3D GAN achieved superior metric results, its impact on quantitative image features and downstream task performance did not result in universal improvement. These results suggest a more complicated relationship between CT acquisition and reconstruction parameters and their effect on radiomic features and ML model performance, which are not fully captured using per-pixel metrics alone. Our approach provides a more comprehensive picture of the effect of normalization.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123082482","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9433968
T. Kline
Generalizing automated medical image segmentation methods to new image domains is inherently difficult. We have previously developed a number of automated segmentation methods that perform at the level of human readers on images acquired under similar conditions to the original training data. We are interested in exploring techniques that will improve model generalization to new imaging domains. In this study we explore a method to limit the inherent bias of these models to intensity and textural information. Using a dataset of 100 T2-weighted MR images with fat-saturation, and 100 T2-weighted MR images without fat-saturation, we explore the use of neural style transfer to induce shape preference and improve model performance on the task of segmenting the kidneys in patients affected by polycystic kidney disease. We find that using neural style transfer images improves the average dice value by $sim0.2$. In addition, visualizing individual network kernel responses highlights a drastic difference in the optimized networks. Biasing models to invoke shape preference is a promising approach to create methods that are more closely aligned with human perception.
{"title":"Improving Domain Generalization in Segmentation Models with Neural Style Transfer","authors":"T. Kline","doi":"10.1109/ISBI48211.2021.9433968","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9433968","url":null,"abstract":"Generalizing automated medical image segmentation methods to new image domains is inherently difficult. We have previously developed a number of automated segmentation methods that perform at the level of human readers on images acquired under similar conditions to the original training data. We are interested in exploring techniques that will improve model generalization to new imaging domains. In this study we explore a method to limit the inherent bias of these models to intensity and textural information. Using a dataset of 100 T2-weighted MR images with fat-saturation, and 100 T2-weighted MR images without fat-saturation, we explore the use of neural style transfer to induce shape preference and improve model performance on the task of segmenting the kidneys in patients affected by polycystic kidney disease. We find that using neural style transfer images improves the average dice value by $sim0.2$. In addition, visualizing individual network kernel responses highlights a drastic difference in the optimized networks. Biasing models to invoke shape preference is a promising approach to create methods that are more closely aligned with human perception.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124548966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9433987
Rituparna Sarkar, Daniel Darby, Héloise Foucambert, S. Meilhac, J. Olivo-Marin
In developmental biology, quantification of cellular morphological features is a critical step to get insight into tissue morphogenesis. The efficacy of the quantification tools rely heavily on robust segmentation techniques which can delineate individual cells/nuclei from cluttered environment. Application of popular neural network methods is restrained by the availability of ground truth necessary for 3D nuclei segmentation. Consequently, we propose a convolutional neural network method, combined with graph theoretic approach for 3D nuclei segmentation of mouse embryo cardiomyocytes, imaged by light-sheet microscopy. The designed neural network architecture encapsulates both membrane and nuclei cues for 2D detection. A global association of the 2D nuclei detection is performed by solving a linear optimization with second order constraint to obtain 3D nuclei reconstruction.
{"title":"Nu3D: 3D Nuclei Segmentation from Light-Sheet Microscopy Images of the Embryonic Heart","authors":"Rituparna Sarkar, Daniel Darby, Héloise Foucambert, S. Meilhac, J. Olivo-Marin","doi":"10.1109/ISBI48211.2021.9433987","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9433987","url":null,"abstract":"In developmental biology, quantification of cellular morphological features is a critical step to get insight into tissue morphogenesis. The efficacy of the quantification tools rely heavily on robust segmentation techniques which can delineate individual cells/nuclei from cluttered environment. Application of popular neural network methods is restrained by the availability of ground truth necessary for 3D nuclei segmentation. Consequently, we propose a convolutional neural network method, combined with graph theoretic approach for 3D nuclei segmentation of mouse embryo cardiomyocytes, imaged by light-sheet microscopy. The designed neural network architecture encapsulates both membrane and nuclei cues for 2D detection. A global association of the 2D nuclei detection is performed by solving a linear optimization with second order constraint to obtain 3D nuclei reconstruction.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127666307","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9434009
Shiyu Wang, N. Dvornek
Traditional regression models do not generalize well when learning from small and noisy datasets. Here we propose a novel metamodel structure to improve the regression result. The metamodel is composed of multiple classification base models and a regression model built upon the base models. We test this structure on the prediction of autism spectrum disorder (ASD) severity as measured by the ADOS communication (ADOS_COMM) score from resting-state fMRI data, using a variety of base models. The metamodel outperforms traditional regression models as measured by the Pearson correlation coefficient between true and predicted scores and stability. In addition, we found that the metamodel is more flexible and more generalizable.
{"title":"A Metamodel Structure For Regression Analysis: Application To Prediction Of Autism Spectrum Disorder Severity","authors":"Shiyu Wang, N. Dvornek","doi":"10.1109/ISBI48211.2021.9434009","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9434009","url":null,"abstract":"Traditional regression models do not generalize well when learning from small and noisy datasets. Here we propose a novel metamodel structure to improve the regression result. The metamodel is composed of multiple classification base models and a regression model built upon the base models. We test this structure on the prediction of autism spectrum disorder (ASD) severity as measured by the ADOS communication (ADOS_COMM) score from resting-state fMRI data, using a variety of base models. The metamodel outperforms traditional regression models as measured by the Pearson correlation coefficient between true and predicted scores and stability. In addition, we found that the metamodel is more flexible and more generalizable.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128802509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9433955
Li Lin, Pujin Cheng, Zhonghua Wang, Meng Li, Kai Wang, Xiaoying Tang
Precise quantification of the corneal nerve plexus morphology is of great importance in diagnosing peripheral diabetic neuropathy and assessing the progression of various eye-related systemic diseases, wherein segmentation of corneal nerves is an essential component. In this paper, we proposed and validated a novel pipeline for corneal nerve segmentation, comprising corneal confocal microscopy (CCM) image synthesis, image quality enhancement and nerve segmentation. Our goal was to address three major problems existing in most CCM datasets, namely inaccurate annotations, non-uniform illumination and contrast variations. In our synthesis and enhancement steps, we employed multilayer and patchwise contrastive learning based Generative Adversarial Network (GAN) frameworks, which took full advantage of multi-scale local features. Through both qualitative and quantitative experiments on two publicly available CCM datasets, our pipeline has achieved overwhelming enhancement performance compared to several state-of-the-art methods. Moreover, the segmentation results showed that models trained on our synthetic images performed much better than those trained on a real CCM dataset, which clearly identified the effectiveness of our synthesis method. Overall, our proposed pipeline can achieve satisfactory segmentation performance for poor-quality CCM images without using any manual labels and can effectively enhance those images.
{"title":"Automated Segmentation Of Corneal Nerves In Confocal Microscopy Via Contrastive Learning Based Synthesis And Quality Enhancement","authors":"Li Lin, Pujin Cheng, Zhonghua Wang, Meng Li, Kai Wang, Xiaoying Tang","doi":"10.1109/ISBI48211.2021.9433955","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9433955","url":null,"abstract":"Precise quantification of the corneal nerve plexus morphology is of great importance in diagnosing peripheral diabetic neuropathy and assessing the progression of various eye-related systemic diseases, wherein segmentation of corneal nerves is an essential component. In this paper, we proposed and validated a novel pipeline for corneal nerve segmentation, comprising corneal confocal microscopy (CCM) image synthesis, image quality enhancement and nerve segmentation. Our goal was to address three major problems existing in most CCM datasets, namely inaccurate annotations, non-uniform illumination and contrast variations. In our synthesis and enhancement steps, we employed multilayer and patchwise contrastive learning based Generative Adversarial Network (GAN) frameworks, which took full advantage of multi-scale local features. Through both qualitative and quantitative experiments on two publicly available CCM datasets, our pipeline has achieved overwhelming enhancement performance compared to several state-of-the-art methods. Moreover, the segmentation results showed that models trained on our synthetic images performed much better than those trained on a real CCM dataset, which clearly identified the effectiveness of our synthesis method. Overall, our proposed pipeline can achieve satisfactory segmentation performance for poor-quality CCM images without using any manual labels and can effectively enhance those images.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116026467","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9433941
Jiaqi Yang, Xiaoling Hu, Chao Chen, Chialing Tsai
We propose a new topology-preserving method for 3D image segmentation. We treat the image as a stack of 2D images so that the topological computation can be carried only within 2D in order to achieve computational efficiency. To enforce the continuity between slices, we propose a compound multi-slice representation and a compound multi-slice topological loss that incorporates rich topological information from adjacent slices. The quantitative and qualitative results show that our proposed method outperforms various strong baselines, especially for structure-related evaluation metrics.
{"title":"3D Topology-Preserving Segmentation with Compound Multi-Slice Representation","authors":"Jiaqi Yang, Xiaoling Hu, Chao Chen, Chialing Tsai","doi":"10.1109/ISBI48211.2021.9433941","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9433941","url":null,"abstract":"We propose a new topology-preserving method for 3D image segmentation. We treat the image as a stack of 2D images so that the topological computation can be carried only within 2D in order to achieve computational efficiency. To enforce the continuity between slices, we propose a compound multi-slice representation and a compound multi-slice topological loss that incorporates rich topological information from adjacent slices. The quantitative and qualitative results show that our proposed method outperforms various strong baselines, especially for structure-related evaluation metrics.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"95 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116717000","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9433952
K. B. D. Raad, Karin A. van Garderen, M. Smits, S. V. D. Voort, Fatih Incekara, E. Oei, J. Hirvasniemi, S. Klein, M. P. Starmans
In recent years, deep learning has become the leading method for medical image segmentation. While the majority of studies focus on developments of network architectures, several studies have shown that non-architectural factors also play a substantial role in performance improvement. An important factor is preprocessing. However, there is no agreement on which preprocessing steps work best for different applications. The aim of this study was to investigate the effect of preprocessing on model performance. To this end, we conducted a systematic evaluation of 24 preprocessing configurations on three clinical application datasets (brain, liver, and knee). Different configurations of normalization, region of interest selection, bias field correction, and resampling methods were applied before training one convolutional neural network. Performance varied up to 64 percentage points between configurations within one dataset. Across the three datasets, different configurations performed best. In conclusion, to improve model performance, preprocessing should be tuned to the specific segmentation application.
{"title":"The Effect of Preprocessing on Convolutional Neural Networks for Medical Image Segmentation","authors":"K. B. D. Raad, Karin A. van Garderen, M. Smits, S. V. D. Voort, Fatih Incekara, E. Oei, J. Hirvasniemi, S. Klein, M. P. Starmans","doi":"10.1109/ISBI48211.2021.9433952","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9433952","url":null,"abstract":"In recent years, deep learning has become the leading method for medical image segmentation. While the majority of studies focus on developments of network architectures, several studies have shown that non-architectural factors also play a substantial role in performance improvement. An important factor is preprocessing. However, there is no agreement on which preprocessing steps work best for different applications. The aim of this study was to investigate the effect of preprocessing on model performance. To this end, we conducted a systematic evaluation of 24 preprocessing configurations on three clinical application datasets (brain, liver, and knee). Different configurations of normalization, region of interest selection, bias field correction, and resampling methods were applied before training one convolutional neural network. Performance varied up to 64 percentage points between configurations within one dataset. Across the three datasets, different configurations performed best. In conclusion, to improve model performance, preprocessing should be tuned to the specific segmentation application.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117009038","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9433871
Tomé Albuquerque, J. S. Cardoso
Cervical cancer ranks as the fourth most common cancer among females worldwide with roughly 528,000 new cases yearly. Significant progress in the realm of artificial intelligence particularly in neural networks and deep learning help physicians to diagnose cervical cancer more accurately. In this paper, we address a classification problem with the widely used VGG16 architecture. In addition to classification error, our model considers a regularization part during tuning of the weights, acting as prior knowledge of the colposcopic image. This embedded regularization approach, using a 2D Gaussian kernel, has enabled the model to learn which sections of the medical images are more crucial for the classification task. The experimental results show an improvement compared with standard transfer learning and multimodal approaches of cervical cancer classification in literature.
{"title":"Embedded Regularization For Classification Of Colposcopic Images","authors":"Tomé Albuquerque, J. S. Cardoso","doi":"10.1109/ISBI48211.2021.9433871","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9433871","url":null,"abstract":"Cervical cancer ranks as the fourth most common cancer among females worldwide with roughly 528,000 new cases yearly. Significant progress in the realm of artificial intelligence particularly in neural networks and deep learning help physicians to diagnose cervical cancer more accurately. In this paper, we address a classification problem with the widely used VGG16 architecture. In addition to classification error, our model considers a regularization part during tuning of the weights, acting as prior knowledge of the colposcopic image. This embedded regularization approach, using a 2D Gaussian kernel, has enabled the model to learn which sections of the medical images are more crucial for the classification task. The experimental results show an improvement compared with standard transfer learning and multimodal approaches of cervical cancer classification in literature.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117031994","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-13DOI: 10.1109/ISBI48211.2021.9433870
Md. Aminur Rab Ratul, Kun Yuan, Won-Sook Lee
Despite the advancement of the deep neural network, the 3D CT reconstruction from its correspondence 2D X-ray is still a challenging task in computer vision. To tackle this issue here, we proposed a new class-conditioned network, namely CCX-rayNet, which is proficient in recapturing the shapes and textures with prior semantic information in the resulting CT volume. Firstly, we propose a Deep Feature Transform (DFT) module to modulate the 2D feature maps of semantic segmentation spatially by generating the affine transformation parameters. Secondly, by bridging 2D and 3D features (Depth-Aware Connection), we heighten the feature representation of the X-ray image. Particularly, we approximate a 3D attention mask to be employed on the enlarged 3D feature map, where the contextual association is emphasized. Furthermore, in the biplanar view model, we incorporate the Adaptive Feature Fusion (AFF) module to relieve the registration problem that occurs with unrestrained input data by using the similarity matrix. As far as we are aware, this is the first study to utilize prior semantic knowledge in the 3D CT reconstruction. Both qualitative and quantitative analyses manifest that our proposed CCX-rayNet outperforms the baseline method.
{"title":"CCX-rayNet: A Class Conditioned Convolutional Neural Network For Biplanar X-Rays to CT Volume","authors":"Md. Aminur Rab Ratul, Kun Yuan, Won-Sook Lee","doi":"10.1109/ISBI48211.2021.9433870","DOIUrl":"https://doi.org/10.1109/ISBI48211.2021.9433870","url":null,"abstract":"Despite the advancement of the deep neural network, the 3D CT reconstruction from its correspondence 2D X-ray is still a challenging task in computer vision. To tackle this issue here, we proposed a new class-conditioned network, namely CCX-rayNet, which is proficient in recapturing the shapes and textures with prior semantic information in the resulting CT volume. Firstly, we propose a Deep Feature Transform (DFT) module to modulate the 2D feature maps of semantic segmentation spatially by generating the affine transformation parameters. Secondly, by bridging 2D and 3D features (Depth-Aware Connection), we heighten the feature representation of the X-ray image. Particularly, we approximate a 3D attention mask to be employed on the enlarged 3D feature map, where the contextual association is emphasized. Furthermore, in the biplanar view model, we incorporate the Adaptive Feature Fusion (AFF) module to relieve the registration problem that occurs with unrestrained input data by using the similarity matrix. As far as we are aware, this is the first study to utilize prior semantic knowledge in the 3D CT reconstruction. Both qualitative and quantitative analyses manifest that our proposed CCX-rayNet outperforms the baseline method.","PeriodicalId":372939,"journal":{"name":"2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)","volume":"284 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115213338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：