Enhanced EfficientNet Model for Multiclass Brain Tumor Prognostication Using Advanced MR Image Analysis Techniques

doi:10.21203/rs.3.rs-4809509/v1

Download PDF

Research Article

Enhanced EfficientNet Model for Multiclass Brain Tumor Prognostication Using Advanced MR Image Analysis Techniques

https://doi.org/10.21203/rs.3.rs-4809509/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

The prognosis of brain tumor diseases is essential for effective treatment planning and patient management. This study investigates the use of Dense EfficientNet models, specifically an enhanced EfficientNet-B1, for the prognostication of multiclass brain tumor diseases. A dataset comprising 6462 MR images, including T1-W, T2-W, and FLAIR sequences, was classified into four categories: glioma, meningioma, no tumor, and pituitary tumors. The proposed method incorporates advanced data augmentation techniques, image cropping, and pixel resizing to improve training accuracy. Additionally, modifications to the EfficientNet architecture layers and the application of normalization and histogram equalization further enhance model performance.The results indicate that the enhanced EfficientNet-B1 model achieves a superior training accuracy of 98%, outperforming the EfficientNet-B0 model, with the highest accuracy observed in glioma tumor classification. Compared with other CNN architectures, such as ResNet50 and VGG-16, the EfficientNet-B1 model demonstrates higher performance and computational efficiency with fewer parameters.The study concludes that the enhanced EfficientNet-B1 model offers a robust and efficient solution for brain tumor detection and prognostication using MR images. Its innovative modifications and advanced preprocessing techniques significantly contribute to its high performance, making it a valuable tool for developing clinically useful applications for MR image analysis in brain tumor management.

EfficientNet

Deep learning

Brain tumor

CNN models

Brain tumors are a type of disease that affects specific parts of the brain and are caused by abnormal brain cells. The two types of brain tumors that can be identified are benign(nonharmful) and malignant(harmful)(Abiwinanda et al. 2019; Brindha et al. 2021). There are four distinct grades of brain tumors: grades I and II are called "low grade gliomas" while grades III and IV are called "high grade gliomas." Malignant tumors, also known as Grade III or IV tumors, grow quickly and must be identified as soon as possible(Özyurt et al. 2019). The most common type of malignant brain tumor is glioma which grows in the glial cells of the brain has the potential to invade nearby healthy tissues and has immediate side effects such as decreased life expectancy(Ayadi et al. 2021). Meningioma’s, on the other hand, do not spread rapidly and may be surgically removed, however diagnosing them early on might be challenging. Pituitary tumors are benign tumors that affect the pituitary glands and do not spread to other regions of the body. Loss of eyesight and hormone deficiencies can result from pituitary tumors(Ayadi et al. 2021). When diagnosing and treating brain tumors, early detection, accurate grading, and categorization are essential. Throughout the history of medical image analysis, different techniques have been employed to identify and categorize brain tumors. Many medical experts use a variety of MR images to design effective treatments for diagnosing brain tumors (Gu et al. 2021). Brain tumors are currently diagnosed and classified via histological analysis of tissue samples.Radiologists use MRS metabolite ratios to assess tumor grades manually. Nonetheless, this treatment is unique because radiologists evaluate the reports manually using the metabolites peaks ratios which allows a possibility of human error. These limitations make it imperative to develop a completely automated method for identifying and classifying brain tumors(Irmak 2021). In recent years, brain tumor classification has been addressed by medical professionals through the use of automated computer-aided techniques. Many machine learning and deep learning techniques have been used by previous researchers. Many models, including DenseNeT, ResNET, Google Net, VSM, and Inception V3, have been used for classification tasks. Milica M. Badza et al. classified of brain tumors (meningioma, glioma and pituitary tumors) using convolutional neural network (CNNS) 10 fold cross-validation methods. He used pretrained models on T1-weighted contrast-enhanced magnetic resonance images. He achieved 96.56% accuracy on training data(Badža and Barjaktarović 2020). Ayadi et al. used computer-assisted diagnosis for the classification of brain tumors into different grades. They ran three different types of MR images into a model and achieved 94.74% accuracy (Ayadi et al. 2021).Pereira et al. used CNN models for the visualization of glioblastoma multiforme. Using this approach, he was able to make predictions directly from MRI scans, eliminating the requirement for a region of interest. He had a grade prediction from the entire brain of 85% and a tumor ROI of 92.8%(Pereira et al. 2018). Abiwinanda et al. proposed CNN models to classify three different types of brain tumors (meningioma, pituitary and glioma tumors). He trained a CNN model on 3064 T1-weighted CE-MR images. He achieved 98.5% training accuracy and 84.19% validation accuracy(Abiwinanda et al. 2019).

Hossam et al. proposed a CNN architecture to classify different grades of brain tumours such as Grade II, Grade III and Grade IV. 3588 MR T1-weighted contrast-enhanced images were obtained. He achieved overall accuracies of 96.13% and 98.7% in two different studies(Sultan, Salem, and Al-Atabany 2019).Cinar and Yildirim et al. used pretrained models (ResNet50,AlexNet, DenseNet201,etc.) for the detection of brain tumors. Eight additional layers were added to ResNet50, replacing the last five layers. The total accuracy of this model was 97.2% (Çinar and Yildirim 2020). Khawaldeh et al. used the ConvNet model for classifying brain tumors. He used a modified version of AlexNet .He achieved 91.6% accuracy in the classification of brain tumors(Khawaldeh et al. 2017).

Talo et al. used a deep transfer learning approach for the classification of brain tumors. For this purpose he used the ResNet34 model. He used 613 MR images and achieved 100% 5- fold accuracy(Talo et al. 2019).Deepak and Ameer et al. used pretrained Google-Net models to extract features from M images of the brain. He obtained 98% accuracy in three different grades of brain tumors(Deepak and Ameer 2019). Mohsen et al. used a deep learning architecture for categorizing of four different grades of tumors. He used the DWT model for feature extraction and achieved 96.97% accuracy(Mohsen et al. 2018).

Deep learning has proven to be highly efficient in the field of biomedical sciences, surpassing traditional biomedical models by offering superior data descriptions(Rao, Sarabi, and Jaiswal 2015). The conventional method is more complex because the brain has high-density metabolites that are difficult to access manually. Consequently, many computer-based techniques are used to diagnose tumors (Reddy et al. 2020). These developments are vital resources for imaging experts as well as other medical specialties. MRI images are processed using a variety of algorithms based on deep learning, particularly for segmentation and image classification, which provides radiologists with valuable second opinions(Amin et al. 2018). The most popular deep learning method for brain tumor classification is convolutional neural network model (Nayak et al. 2022). Convolutional neural network models reduce the lengthy and time consuming manual image evaluation process. Most CNNs extract pertinent features from tumor images which leads to a substantial improvement in performance (Tripathy, Singh, and Ray 2023). The objective of this research is to effectively categorize tumors using Dense EfficientNet models which refers to CNNs models. This is the first time that EfficientNet models is used for classification of brain tumors into four different classes. EfficientNet models are more advanced network models and provide greater training accuracy on large datasets(Tan and Le 2019). Comprehensive image data are acquired to generate densely reconstructed segmentation masks for classifying tumors into four categories.

2.1 Dataset Preparation

In this study, a total of 6462 MR images were used for brain tumor classification. The data are publicly available on “Mendely Data”, and the computational codes are also available on the same website as dataset (dataset 2024). The dataset is divided into 4 classes for training, testing and validation phases. Three main kinds of tumor images were collected in three planes: namely, the sagittal, coronal and axial planes which included three sequences, T1-W, T2-W and FLAIR images. The proposed model is based on pretrained EfficientNet models, i.e., EfficientNetB0 and EfficientNetB1(Deng et al. 2009).

2.2 Dense EfficientNet CNN Model

This article introduces a groundbreaking dense convolutional neural network model that combines a pretrained EfficientNet-B0 model with dense layers. The MR images were obtained using EfficientNetB0, which encompasses 230 layers and 7 blocks. Within the dense EfficientNet network, there is an integration of alternating dense and dropout layers for enhanced performance. In this structure, a dense layer acts as a fundamental unit, transmitting all outputs from the previous layer to each of its neurons, with each neuron providing a single output to the subsequent layer. To prevent overfitting, a dropout layer is strategically used to reduce network capacity during training(Tan and Le 2019). However, the second model, EfficientNet B1, boasts 340 layers, and due to its complexity, an expanded version of its architecture is available in supplementary materials. The optimization process employs the Adam algorithm, which is known for its simplicity, computational efficiency, and effectiveness. This algorithm is applied to minimize the categorical cross-entropy loss function associated with model training(Tan and Le 2019). Figure 1shows the performance of the Dense EfficientNet model.

2.2.1 Image Preprocessing

The first step of image preprocessing was standardizing the data into a categorized format. Grayscale image conversion was the first step in the process. Subsequently, Gaussian blurring was applied to mitigate noise. Further processing included the utilization of image enhancement techniques such as erosion and dilation, employing a structuring element to eliminate pixel intensities in small regions. Erosion involves the removal of pixels from object edges, and dilation adds pixels to structure edges. In the final step, black portions were removed from each image(Shah et al. 2022).

2.2.2 Data Augmentation

The primary purpose of data augmentation is to enhance model performance and prevent overfitting by expanding the dataset without the need for additional data collection. Diverse techniques, such as (rotation, and flipping in the horizontal and vertical axes) are implemented. Figure 2 shows the augmented images of different types of brain tumors.

2.3 Model Scaling

ConvNets are frequently designed under specific computational constraints, and their scalability is enhanced for improved accuracy when additional resources become accessible. To improve this approach, we incorporate neural architecture search to establish a novel foundational network that surpasses the performance of prior ConvNets in both accuracy and efficiency. This novel architecture, known as EfficientNet, is meticulously crafted by finding an optimal equilibrium between network depth, width, and resolution(Bello et al. 2021).

The EfficientNet model scaling method for convolutional neural networks (CNNs) is a technique designed to optimize the trade-off between model size and performance. It was introduced to address the challenge of finding the right balance between model capacity and computational efficiency(Han et al. 2021).

Compound scaling is a strategy that involves simultaneously scaling different aspects of a CNN architecture, such as the depth (number of layers), width (number of channels or filters), and resolution (input image size)(Tan and Le 2021). This approach aims to strike a balance between these dimensions to achieve better performance while still maintaining computational efficiency

Depth: d = α^Ø

Width: = β^Ø

Resolution: = ɤ^Ø

s.t α.β².ɤ² = 2

α ≥ 1, β ≥ 1, ɤ≥1

Here, α, β, and γ represent constants that can be determined through a small grid search (Masters et al. 2021). Figure 3 shows the structural difference between the models channels.

In this study, all experiments were carried out utilizing Python version 3.10.1. The EfficientNet-B0 model consisted of 25 epochs, utilizing a batch size of 32. In terms of model accuracy, the initial validation accuracy was approximately 0.75, but after the first epoch, it notably increased to approximately 0.82. Similarly, the initial validation loss exceeded 0.49, but after one epoch, it decreased to below 0.40. Figure 4 illustrates a positive trend wherein the validation accuracy improves to 98%, and the loss diminishes.

The EfficientNet-B0 model spanned 50 epochs, employing a batch size of 32. In terms of model accuracy, the initial validation accuracy started at approximately 0.27 but experienced a notable increase to approximately 0.81 after the first epoch. Similarly, the initial validation loss exceeded 1.20, but after 10 epochs, it decreased to below 0.63. Figure 5 visually depicts a positive trend with an improvement in the 90% validation accuracy.

Based on the observed model accuracy and model loss graphs, it is evident that for EfficientNet B0, the bar graph appears disordered, and there is a considerable gap between loss and accuracy. Consequently, the accuracy of EfficientNet-B0 is lower than that of EfficientNet B1. Specifically, the training accuracy and training loss for EfficientNet-B1 are 98% and 0.047%, respectively, while those for EfficientNet-B0, are 90% and 0.248% respectively. A comprehensive performance analysis, illustrated in Fig. 6, provides a visual representation of this comparison.

In this study, the effectiveness of the proposed approach involving the fine-tuning of EfficientNetB1 is evaluated against various innovative methods using performance metrics. Most researchers have used many other CNN models such as ResNet50 and VGG16 to classify brain tumors. These models face vanishing gradient problems over large datasets. However, EfficientNet models use a compound scaling method to overcome the vanishing gradient problem over large datasets. The best performance was shown for glioma tumors by EfficientNet-B1. Figure 7 illustrates the class-specific evaluation of brain tumors and related data are given in Table 1.

Table 1

provides a comprehensive analysis of assessment metrics, including accuracy and loss values, for the classification of brain tumors using various convolutional neural network (CNN) models
CNN Models	Accuracy	Loss	Glioma	Meningioma	Pituitary	No tumour
EfficientNet-B1(proposed)	98%	0.047%	100%	98%	50%	70%
EfficientNet-B0(Proposed)	90%	0.248%	76%	90%	45%	65%
ResNet50(Sahaai et al. 2022)	75%	25%	78%	88%	78%	55%
VGG-16(Younis et al. 2022)	85%	15%	80%	78%	60%	65%

In this study, our focus was on utilizing unique convolutional neural network models to evaluate high-grade gliomas. In previous studies, many alternative convolutional neural networks, such as ResNET, Densenet, VGG, SqueezeNET, and MobileNET, were used to categorize brain glioma data. Other convolutional algorithms provide similar accuracy but require more time and scale up only one of the three dimensions, width, height, and resolution. At the same time, the EfficientNet model can be efficiently scaled up with fewer parameters and FLOPS. ConvNet models are mostly based on depth scaling. Using the compound scaling method, this research shows the significance of balancing all network width, depth, and resolution dimensions. Ullah, N. et al. used the ResNet50 model to classify brain tumors. ResNet50 has several layers, and more layers require more time to process the input. As the number of layers increases, we will encounter a vanishing gradient problem. ResNet-50, on the other hand, eliminates the saturation problem, although it requires more parameters and delivers less accuracy than does EfficientNet (Ullah et al. 2022). Habiba, S.U. et al. proposed several pretrained models for tumor classification, such as VGG16, Inception V3, and ResNet50. All these models were analyzed for brain tumor classification. Although the accuracy of these models is quite good, all these models can be scaled up on one of the three dimensions. EfficientNet integrates different model scaling dimensions to balance the speed and accuracy of the neural network model: network depth, network width, and image resolution. To discover the optimal mix of these three interdependent dimensions, EfficientNet applies a compound scaling technique. The CNN base model (Efficinetnet-B0) provides higher accuracy as the number of channels and layers increases. The compound scaling method can improve the accuracy by up to 2.5% compared with other single dimensional scales (Habiba et al. 2022). Senan, E.M. et al. trained a dataset using an AlexNet hybrid model to classify brain tumors. He categorized brain MR images into three types of tumors: meningioma, glioma, pituitary, and healthy. The CNN EfficientNet model is compared to the AlexNet hybrid model to visualize the results. This shows that the AlexNet hybrid model has a significant computational cost. In contrast, when applying the EfficientNet model, the computational cost was low. The AlexNet model trained the dataset in 47 minutes and 35 seconds on the GPU processor, and EfficentNetB0 only required 7 seconds of processing time on the GPU processor (Senan et al. 2022).

In this study, we utilized dense EfficientNet models with min-max normalization to effectively categorize brain tumours into four categories. Within this framework, image resizing, cropping, and augmentation techniques were employed to further improve the training accuracy of the system. Notably, the EfficientNet-B1 model exhibited a remarkable 98% accuracy in detecting glioma, while the EfficientNet-B0 model displayed a greater loss at 90% accuracy. This research gap might be filled by focusing on minimizing the number of parameters and the computing time needed to run the suggested model to maintain optimal performance. The proposed model is compared with recent CNN architectures to evaluate the performance of the proposed model. The results demonstrate that EfficientNetB1 operates with fewer parameters and is computationally efficient, demonstrating a better performance in tumor detection.

Ethics Approval and consent to participate:

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Competing Interest

The authors have no relevant financial or non-financial interests to disclose

Funding

The author confirms that this research will receive funding from the “Faculty of Resilience, Rabdan Academy, Abu Dhabi, United Arab Emirates”. The funder contribute to analyse the results.

Author Contribution

A.G: Conceptualization, methodology, software, data validation, writing original draft, and editing the manuscript. Dr.A.J: evaluate the manuscript critically and made contribution in the result. Dr. W.A: handles the funding sources and did critical analysis on the results. S.A: ran the code using python, gathered data and analyzed the data. All authors read and approve the final manuscript.

Acknowledgements

Not applicable

Data Availability

Data is published and can be accessed through "Mendely website " and Github.

Abiwinanda N, Hanif M, Hesaputra ST, Handayani A, Tati Rajab Mengko (2019) Brain tumor classification using convolutional neural network. World Congress on Medical Physics and Biomedical Engineering 2018: June 3–8, 2018, Prague, Czech Republic (Vol. 1). Springer, pp 183–189
Alanazi M, Faleh MU, Ali SJ, Hussain A, Zafar M, Mohatram M, Irfan R, AlRuwaili M, Alruwaili NH, Ali, Anas Mohammad Albarrak (2022) Brain tumor/mass classification framework using magnetic-resonance-imaging-based isolated and developed transfer deep-learning model. Sensors 22:372
Amin J, Sharif M, Yasmin M, and Steven Lawrence Fernandes (2018) Big data analysis for brain tumor detection: Deep convolutional neural networks. Future Generation Comput Syst 87:290–297
Ayadi W, Elhamzi W, Charfi I, Atri M (2021) Deep CNN for brain tumor classification. Neural Process Lett 53:671–700
Badža MM, Marko ČBarjaktarović (2020) 'Classification of brain tumors from MRI images using a convolutional neural network', Applied Sciences, 10: 1999
Bello I, Fedus W, Du X, Cubuk ED, Srinivas A, Lin T-Y Jonathon Shlens, and Barret Zoph. 2021. 'Revisiting resnets: Improved training and scaling strategies'. Adv Neural Inf Process Syst, 34: 22614–22627
Brindha P, Gokila M, Kavinraj P, Manivasakam (2021) and P Prasanth. Brain tumor detection from MRI images using deep learning techniques. In IOP conference series: materials science and engineering, 012115. IOP Publishing
Çinar A, and Muhammed Yildirim (2020) Detection of tumors on brain MRI images using the hybrid convolutional neural network architecture. Med Hypotheses 139:109684
dataset K (2024) Brain Tumor MRI images. In. Mendely Dataset
Deepak S, Ameer PM (2019) Brain tumor classification using deep CNN features via transfer learning. Comput Biol Med 111:103345
Deng J, Dong W, Socher R, Li L-J, Li K, Li F-F (2009) Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, 248 – 55. Ieee
Gu X, Shen Z, Xue J, Yiqing Fan, and, Tongguang, Ni (2021) Brain tumor MR image classification using convolutional dictionary learning with local constraint. Front NeuroSci 15:679847
Habiba S, Umme MK, Islam L, Nahar F, Tasnim MS, Hossain, Andersson K (2022) Brain-DeepNet: a deep learning based classifier for brain tumor detection and classification. In International Conference on Intelligent Computing & Optimization, 550 – 60. Springer
Han D, Yun S, Heo B, YoungJoon, Yoo (2021) Rethinking channel dimensions for efficient model design. In Proceedings of the IEEE/CVF conference on Computer Vision and Pattern Recognition, 732 – 41
Irmak E (2021) Multi-classification of brain tumor MRI images using deep convolutional neural network with fully optimized framework. Iran J Sci Technol Trans Electr Eng 45:1015–1036
Khawaldeh S, Pervaiz U, Rafiq A, Rami S, Alkhawaldeh (2017) Noninvasive grading of glioma tumor using magnetic resonance imaging with convolutional neural networks. Appl Sci 8:27
Masters D, Labatie A (2021) Zach Eaton-Rosen, and Carlo Luschi. 'Making EfficientNet more efficient: Exploring batch-independent normalization, group convolutions and reduced resolution training', arXiv preprint arXiv:2106.03640
Mohsen H, El-Sayed A, El-Dahshan, El-Sayed M, El-Horbaty, Abdel-Badeeh M, Salem (2018) Classification using deep learning neural networks for brain tumors. Future Comput Inf J 3:68–71
Nayak D, Ranjan N, Padhy PK, Mallick DK, Bagal, and Sachin Kumar (2022). 'Brain tumour classification using noble deep learning approach with parametric optimization through metaheuristics approaches', Computers, 11: 10
Özyurt F, Sert E, Avci E, and Esin Dogantekin (2019) Brain tumor detection based on Convolutional Neural Network with neutrosophic expert maximum fuzzy sure entropy. Measurement 147:106830
Pereira Sérgio, Meier R, Alves V, Reyes M, Carlos A, Silva (2018) Automatic brain tumor grading from MRI data using convolutional neural networks and quality assessment. In Understanding and Interpreting Machine Learning in Medical Image Computing Applications: First International Workshops, MLCN 2018, DLF 2018, and iMIMIC 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 16–20, 2018, Proceedings 1, 106 – 14. Springer
Pillai R, Sharma A, Sharma N, Rupesh Gupta (2023) Brain tumor classification using VGG 16, ResNet50, and inception V3 transfer learning models. In 2023 2nd International Conference for Innovation in Technology (INOCON), 1–5. IEEE
Rao, Vinay MS, Sarabi, Jaiswal A (2015) Brain tumor segmentation with deep learning. MICCAI multimodal brain tumor segmentation Chall (BraTS) 59:1–4
Rasheed Z, Ma Y-K, Ullah I, Ghadi YY, Khan MZ, Khan MA, Abdusalomov A, Alqahtani F, Ahmed MS (2023) Brain tumor classification from MRI using image enhancement and convolutional neural network techniques. Brain Sci 13:1320
Rasool M, Ismail NA, Boulila W, Ammar A, Samma H, Wael MS, Yafooz, Abdel-Hamid M, Emara (2022) A hybrid deep learning model for brain tumour classification. Entropy 24:799
Reddy AVN, Krishna CP, Mallick PK, Satapathy SK, Tiwari P, Zymbler M, Sachin Kumar (2020) Analyzing MRI scans to detect glioblastoma tumor using hybrid deep belief networks. J Big Data 7:1–17
Sahaai MB, Jothilakshmi GR, Ravikumar D, Prasath R, Singh S (2022) ResNet-50 based deep neural network using transfer learning for brain tumor classification. In AIP Conference Proceedings. AIP Publishing
Senan E, Mohammed ME, Jadhav, Taha H, Rassem, Zeyad Ghaleb Al-Mekhlafi (2022) Abdulaziz Salamah Aljaloud, Badiea Abdulkarem Mohammed, and. 'Early diagnosis of brain tumour mri images using hybrid techniques between deep and machine learning', Computational and mathematical methods in medicine, 2022
Shah H, Ali F, Saeed S, Yun J-H, Park A Paul, and, Jae-Mo K (2022) A robust approach for brain tumor detection in magnetic resonance images using finetuned efficientnet. IEEE Access 10:65426–65438
Sharma A, Kumar A, Nandal A, Dhaka L, Zhou A, Alhudhaif F, Alenezi, Kemal Polat (2023) Brain tumor classification using the modified ResNet50 model based on transfer learning. Biomed Signal Process Control 86:105299
Sultan, Hossam H, Nancy M, Salem, Walid Al-Atabany (2019) Multi-classification of brain tumor images using deep neural network. IEEE access 7:69215–69225
Talo M, Baloglu UB, Yıldırım Özal, Rajendra Acharya U (2019) Application of deep transfer learning for automated brain abnormality classification using MR images. Cogn Syst Res 54:176–188
Tan M (2019) and Quoc Le. Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning, 6105-14. PMLR
Tan M (2021) Efficientnetv2: Smaller models and faster training. In International conference on machine learning, 10096-106. PMLR
Tripathy S, Singh R, Mousim Ray (2023) Automation of Brain Tumor Identification using EfficientNet on Magnetic Resonance Images. Procedia Comput Sci 218:1551–1560
Ullah N, Khan JA, Khan MS, Khan W, Hassan I, Obayya M, Negm N, Ahmed SS (2022) An effective approach to detect and identify brain tumors using transfer learning. Appl Sci 12:5645
Younis A, Qiang L, Nyatega CO, Adamu MJ, Halima Bello Kawuwa (2022) Brain tumor analysis using deep learning and VGG-16 ensembling learning approaches. Appl Sci 12:7282

No competing interests reported.

EffcientNetB1.pdf

Download PDF

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

Enhanced EfficientNet Model for Multiclass Brain Tumor Prognostication Using Advanced MR Image Analysis Techniques

Status:

Version 1

Abstract

Figures

1 Background

2 Proposed Methodology

2.1 Dataset Preparation

2.2 Dense EfficientNet CNN Model

2.2.1 Image Preprocessing

2.2.2 Data Augmentation

2.3 Model Scaling

3 Results

4 Discussion

5 Conclusion

Declarations

Funding

Author Contribution

Acknowledgements

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1