A CADe System for Gliomas in Brain MRI using Convolutional Neural Networks

Banerjee, Subhasis; Sharma, Anupama; Shankar, B. Uma

doi:10.48550/arxiv.1806.07589

Cited by 3 publications

(3 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…with a random initialization of the parameters) is used as baseline. The baseline in structured as in [45], because of the similarities between our approach and the one designed by authors for detecting brain tumors. The input shape of the network is 224 × 224 × 3, thus compliant with the dataset described in Section II.…”

Section: A Network Architectures and Transfer Learningmentioning

confidence: 99%

Automated Detection of Spinal Lesions From CT Scans via Deep Transfer Learning

Camisa,

Montanari,

Testa

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Convolutional Neural Networks are being increasingly applied to the detection of anomalies in Computed Tomographies (CTs). The goal of this paper is to implement an automated Computer-Aided Detection (CADe) system for spinal lesions using CTs and Convolutional Neural Networks pre-trained on commercial datasets. The proposed pipeline works as follows. The CADe takes in input CT scans and is equipped with an intuitive GUI to allow physicians to use it as a support tool for diagnoses. From the CT scans, the CADe selects volumes containing the vertebrae and extracts 2D slices of these volumes. These slices are pre-processed and then analyzed using a VGG19 Convolutional Neural Network and a tailored, binary classifier. The neural network identifies healthy vertebrae and vertebrae containing lesions (e.g. metastases, primary tumors, lytic and sclerotic lesions). For training and testing purposes, we generated a dataset from CTs of vertebrae from patients treated at IRCCS Rizzoli Orthopaedics Institute of Bologna, Italy, between 2009 and 2019. Both healthy and lesioned vertebrae retrieved with different tomography machines and setups are considered. The dataset has been enlarged using data augmentation techniques and subsequently used to train a wide range of deep learning models. We perform an in-depth benchmark study to assess the performance of the considered classifier against different Convolutional Neural Networks pretrained on the ImageNet dataset and exploiting Transfer Learning techniques. Leveraging Transfer Learning techniques, we reached 93.43% accuracy and a recall of 92.99%.

show abstract

Section: A Network Architectures and Transfer Learningmentioning

confidence: 99%

Automated Detection of Spinal Lesions From CT Scans via Deep Transfer Learning

Camisa,

Montanari,

Testa

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The study of Xu et al 47 aimed to reduce FPs in the segmentation task rather than object detection. Banerjee et al 48 suggested an image processing-based two-stage method reduce FP cases in their study that segmented with brain MRI. The study of Naceur et al 49 is again a study that reveals the FP problem.…”

Section: Related Workmentioning

confidence: 99%

False positive repression: Data centric pipeline for object detection in brain MRI

Terzı

Azgınoglu

Terzi

2021

Concurrency and Computation

View full text Add to dashboard Cite

One of the problems that often arise during the application of medical research to real life is the high number of false positive cases. This situation causes experts to be warned with false alarms unnecessarily and increases their workload. This study proposes a new data centric approach to reduce bias-based false positive predictions in brain MRI-specific medical object detection applications. The proposed method has been tested using two different datasets: Gazi Brains 2020 and BraTS 2020, and three different deep learning-based object detection models: Mask R-CNN, YOLOv5, and EfficientDet. According to the results, the proposed pipeline outperformed the classical pipeline, up to 18% on the Gazi Brains 2020 dataset, and up to 24% on the BraTS 2020 dataset for mean specificity value without much change in sensitivity metric. It means that the proposed pipeline reduces false positive rates due to bias in real-life applications and it can help to reduce the workload of experts.

show abstract

“…Nevertheless, it is a fact that these models struggle with performance issues when they are not sufficiently balanced. So, a compelling need for more robust methods to enhance their performance is required [8]. Existing methods to improve the performance of CNN models in the medical domain include domain adaptation and transfer learning [9].…”

Section: Introductionmentioning

confidence: 99%

A Novel Method for Myocardial Image Classification using Data Augmentation

Zhu¹

2023

IJACSA

View full text Add to dashboard Cite

Myocarditis is an important public health concern since it can cause heart failure and abrupt death. It can be diagnosed with magnetic resonance imaging (MRI) of the heart, a non-invasive imaging technology with the potential for operator bias. The study provides a deep learning-based model for myocarditis detection using CMR images to support medical professionals. The proposed architecture comprises a convolutional neural network (CNN), a fully-connected decision layer, a generative adversarial network (GAN)-based algorithm for data augmentation, an enhanced DE for pre-training weights, and a reinforcement learning-based method for training. We present a new method of employing produced images for data augmentation based on GAN to improve the classification performance of the provided CNN. Unbalanced data is one of the most significant classification issues, as negative samples are more than positive, decimating system performance. To solve this issue, we offer an RL-based training method that learns minority class examples with attention. In addition, we tackle the challenges associated with the training step, which typically relies on gradient-based techniques for the learning process; however, these methods often face issues like sensitivity to initialization. To start the BP process, we present an improved differential evolution (DE) technique that leverages a clustering-based mutation operator. It recognizes a successful cluster for DE and applies an original updating strategy to produce potential solutions. We assess our suggested model on the Z-Alizadeh Sani myocarditis dataset and show that it outperforms other methods.

show abstract

A CADe System for Gliomas in Brain MRI using Convolutional Neural Networks

Cited by 3 publications

References 45 publications

Automated Detection of Spinal Lesions From CT Scans via Deep Transfer Learning

Automated Detection of Spinal Lesions From CT Scans via Deep Transfer Learning

False positive repression: Data centric pipeline for object detection in brain MRI

A Novel Method for Myocardial Image Classification using Data Augmentation

Contact Info

Product

Resources

About