Multi-stream Convolutional Autoencoder and 2D Generative Adversarial Network for Glioma Classification

Ali, Muhaddisa Barat; Gu, Irene Yu‐Hua; Jakola, Asgeir Store

doi:10.1007/978-3-030-29888-3_19

Cited by 11 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As the datasets have no tumor masks, the tumor regions are extracted by fixing a tight rectangular bounding box around ROI of images. These images with only tumor regions are used in a two step training strategy by 2-streams of convolutional autoencoder (CAE) [ 34 ]. During pre-training, phase features are learned from augmented images

(T1ce-MRI and FLAIR-MRI).…”

Section: Overview Of the Proposed Methodsmentioning

confidence: 99%

“…Considering these challenges, our work is focused to propose a robust method by domain mapping to overcome the scanner dependent mismatches that preserves the molecular structural originality of gliomas. In this paper, we propose a novel approach based on CycleGAN [33] and multistream convolutional autoencoder framework [34] as a classifier. Although CycleGAN has been applied for non-medical applications [33] and cross-modality translation of MRIs [26], to the best of our knowledge this is the first work used for domain mapping that retains molecular-subtype information in low grade gliomas.…”

Section: Related Workmentioning

confidence: 99%

“…To investigate the effectiveness of our approach on each case study, first we had to choose which dataset should be mapped to the other. In this regard, we did a primarily test using multi-stream CAE classifier [34] on the USA and France datasets separately and then combined them without applying domain mapping. Then we compared the performance with combined dataset obtained from after domain mapping.…”

Section: Performance Evaluation On the Impact Of Individual Partsmentioning

confidence: 99%

“…It is worth mentioning that although a part of this work has been presented in [34] which was based on classification of low and high grade gliomas, however, this paper applies to classify molecular subtypes in LGGs instead of tumor grading. Furthermore, this paper differs in terms of: mapping multi-source data to a common domain; dealing with the clinical MRIs that are not uniform in size in all 3D directions; avoiding laborious task of defining soft tissue boundaries; and lastly, including empirical tests and evaluation on two clinical datasets from different hospitals.…”

mentioning

confidence: 99%

See 3 more Smart Citations

Domain Mapping and Deep Learning from Multiple MRI Clinical Datasets for Prediction of Molecular Subtypes in Low Grade Gliomas

Ali

Berger

et al. 2020

Brain Sciences

Self Cite

View full text Add to dashboard Cite

Brain tumors, such as low grade gliomas (LGG), are molecularly classified which require the surgical collection of tissue samples. The pre-surgical or non-operative identification of LGG molecular type could improve patient counseling and treatment decisions. However, radiographic approaches to LGG molecular classification are currently lacking, as clinicians are unable to reliably predict LGG molecular type using magnetic resonance imaging (MRI) studies. Machine learning approaches may improve the prediction of LGG molecular classification through MRI, however, the development of these techniques requires large annotated data sets. Merging clinical data from different hospitals to increase case numbers is needed, but the use of different scanners and settings can affect the results and simply combining them into a large dataset often have a significant negative impact on performance. This calls for efficient domain adaption methods. Despite some previous studies on domain adaptations, mapping MR images from different datasets to a common domain without affecting subtitle molecular-biomarker information has not been reported yet. In this paper, we propose an effective domain adaptation method based on Cycle Generative Adversarial Network (CycleGAN). The dataset is further enlarged by augmenting more MRIs using another GAN approach. Further, to tackle the issue of brain tumor segmentation that requires time and anatomical expertise to put exact boundary around the tumor, we have used a tight bounding box as a strategy. Finally, an efficient deep feature learning method, multi-stream convolutional autoencoder (CAE) and feature fusion, is proposed for the prediction of molecular subtypes (1p/19q-codeletion and IDH mutation). The experiments were conducted on a total of 161 patients consisting of FLAIR and T1 weighted with contrast enhanced (T1ce) MRIs from two different institutions in the USA and France. The proposed scheme is shown to achieve the test accuracy of 74 . 81 % on 1p/19q codeletion and 81 . 19 % on IDH mutation, with marked improvement over the results obtained without domain mapping. This approach is also shown to have comparable performance to several state-of-the-art methods.

show abstract

(T1ce-MRI and FLAIR-MRI).…”

Section: Overview Of the Proposed Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Performance Evaluation On the Impact Of Individual Partsmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Domain Mapping and Deep Learning from Multiple MRI Clinical Datasets for Prediction of Molecular Subtypes in Low Grade Gliomas

Ali

Berger

et al. 2020

Brain Sciences

Self Cite

View full text Add to dashboard Cite

show abstract

“…Ge et al (2018) proposed a multistream deep CNN architecture for glioma grading followed by multimodality MRI data fusion, achieving test accuracy of 90.87%. Ali et al (2019) used the generative adversarial networks (GANs) for data augmentation and employed a multistream convolutional autoencoder (CAE) to extract multi-modality MRI features for classification of low/high grade gliomas, achieving test accuracy of 92.04%. Although deep learning methods have shown excellent performance in biomedical domains, their lack of interpretability still remains an issue, especially for clinical practice.…”

Section: Introductionmentioning

confidence: 99%

Detection and Grading of Gliomas Using a Novel Two-Phase Machine Learning Method Based on MRI Images

Chen

Xiao

et al. 2021

Front. Neurosci.

View full text Add to dashboard Cite

The early detection and grading of gliomas is important for treatment decision and assessment of prognosis. Over the last decade numerous automated computer analysis tools have been proposed, which can potentially lead to more reliable and reproducible brain tumor diagnostic procedures. In this paper, we used the gradient-based features extracted from structural magnetic resonance imaging (sMRI) images to depict the subtle changes within brains of patients with gliomas. Based on the gradient features, we proposed a novel two-phase classification framework for detection and grading of gliomas. In the first phase, the probability of each local feature being related to different types (e.g., diseased or healthy for detection, benign or malignant for grading) was calculated. Then the high-level feature representing the whole MRI image was generated by concatenating the membership probability of each local feature. In the second phase, the supervised classification algorithm was used to train a classifier based on the high-level features and patient labels of the training subjects. We applied this framework on the brain imaging data collected from Zhongnan Hospital of Wuhan University for glioma detection, and the public TCIA datasets including glioblastomas (WHO IV) and low-grade gliomas (WHO II and III) data for glioma grading. The experimental results showed that the gradient-based classification framework could be a promising tool for automatic diagnosis of brain tumors.

show abstract

Brain tumor grade classification using multi‐step pre‐training

Mehmood,

Bajwa,

Anwar

2023

Int J Imaging Syst Tech

View full text Add to dashboard Cite

Medical images offer a non‐invasive method to diagnose different diseases, but using them manually produces unreliable results. Modern deep learning architectures and techniques are computed and data‐intensive, making them difficult to use for relatively smaller datasets of medical images. Transfer learning has been used as a remedy for the problem mentioned above. However, the domain difference between the datasets used for pre‐training (e.g., ImageNet) and the target datasets, like medical images, negatively impacts the transfer learning results. Recently, many researchers have used additional pre‐training called domain‐adaptive pre‐training (DAPT) using the data from the target domain (e.g., medical images) before using the model on the target tasks to achieve superior performance. This study proposes a variant of DAPT by performing it on a subset of the architecture. It has achieved state‐of‐the‐art performance for brain tumor grading on the BraTS 2019 while being computationally efficient.

show abstract

Multi-stream Convolutional Autoencoder and 2D Generative Adversarial Network for Glioma Classification

Cited by 11 publications

References 12 publications

Domain Mapping and Deep Learning from Multiple MRI Clinical Datasets for Prediction of Molecular Subtypes in Low Grade Gliomas

Domain Mapping and Deep Learning from Multiple MRI Clinical Datasets for Prediction of Molecular Subtypes in Low Grade Gliomas

Detection and Grading of Gliomas Using a Novel Two-Phase Machine Learning Method Based on MRI Images

Brain tumor grade classification using multi‐step pre‐training

Contact Info

Product

Resources

About