Transfer Learning-Based Classification Comparison of Stroke

Abstract

One type of brain disease that significantly harms people's lives and health is stroke. The diagnosis and management of strokes both heavily rely on the quantitative analysis of brain Magnetic Resonance (MR) images. The early diagnosis process is of great importance for the prevention of stroke cases. Stroke prediction is made possible by deep neural networks with the capacity for enormous data learning. Therefore, in thus study, several deep neural network models, including DenseNet121, ResNet50, Xception, MobileNet, VGG16, and EfficientNetB2 are proposed for transfer learning to classify MR images into two categories (stroke and non-stroke) in order to study the characteristics of the stroke lesions and achieve full intelligent automatic detection. The study dataset comprises of 1901 training images, 475 validation images, and 250 testing images. On the training and validation sets, data augmentation was used to increase the number of images to improve the models’ learning. The experimental results outperform all the state of arts that were used the same dataset. The overall accuracy of the best model is 98.8% and the same value for precision, recall, and f1-score using the EfficientNetB2 model for transfer learning.

An intelligent neural network model to detect red blood cells for various blood structure classification in microscopic medical images

“…Developing algorithms capable of precisely segmenting and enumerating red blood cells in microscopy images, and furnishing data on the distribution of minute particles, would be advantageous in ensuring precise clinical analysis. For instance, Khalid [4] and Rusul [5] presented transfer learning-based models for automatically diagnosing skin cancer and brain strokes because manual diagnosis is a challenging task for human beings due to identical properties in images. Similarly, deep learning models are important to automatically diagnose abnormalities in blood cells.…”

Development of deep learning framework for anatomical landmark detection and guided dissection line during laparoscopic cholecystectomy

“…Skin cancer classification, for instance, has seen significant improvements through deep learning algorithms, facilitating better identification of cancerous appearances and improving overall patient prognosis [ 9 ]. Similarly, stroke detection has benefited from this approach, enabling rapid and precise identification of stroke lesions in Magnetic Resonance (MR) images, thereby accelerating patient treatment and potentially improving outcomes [ 10 ]. Deep learning has also been applied for ultrasound (US) image classification, making the analysis of Carotid artery disease, and enhancing diagnostic accuracy [ 11 ].…”

Deep Transfer Learning to Classify Mass and Calcification Pathologies from Screen Film Mammograms

“…Despite these and other successful results in the literature, feature extraction-based machine learning methods can be time-consuming, particularly for medical image-or video-based analysis. Deep learning methods enable learning features from training data, and there are example studies in the literature related to stroke [16], carotid artery [17], [18], skin cancer [19], diabetes [20], Alzheimer's disease [21], and breast cancer [22]. Deep learning methods should be useful for the solution of problems related to other biomedical imaging modalities, so cross-disciplinary studies could possibly lead to new scientific advancements.…”