A Lightweight Convolutional Neural Network for Classification of Brain Tumors Using Magnetic Resonance Imaging DOI Creative Commons

Alper Özatılgan,

Mahır Kaya

Sakarya University Journal of Computer and Information Sciences, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 25, 2024

The brain, which controls important vital functions such as vision, hearing and movement, negatively affects our lives when it is sick. Of these diseases, the deadliest undoubtedly brain tumor, can occur in all age groups be benign or malignant. Therefore, early diagnosis prognosis are very important. Magnetic Resonance (MR) images used for detection treatment of tumor types. Successful results diseases from medical with Convolutional Neural Networks (CNN) depend on optimum creation number layers other hyper-parameters. In this study, we propose a CNN model that will achieve highest accuracy least layers. A public data set consisting 4 different classes (Meningioma, Glioma, Pituitary Normal) obtained use training models was trained tested 50 deep learning designed, better result compared existing studies literature 99.47% 99.44% F1 score values.

Language: Английский

Real-time obstacle perception method for UAVs with an RGB-D camera in low-light environments DOI
Hua Wang, Hao Wang, Yan‐Jun Liu

et al.

Signal Image and Video Processing, Journal Year: 2025, Volume and Issue: 19(3)

Published: Jan. 28, 2025

Language: Английский

Citations

0
Enhancing Intermittent Spare Part Demand Forecasting: A Novel Ensemble Approach with Focal Loss and SMOTE DOI Creative Commons
Saskia Puspa Kenaka, Andi Cakravastia, Anas Ma’ruf

et al.

Logistics, Journal Year: 2025, Volume and Issue: 9(1), P. 25 - 25

Published: Feb. 8, 2025

Background: Accurate inventory management of intermittent spare parts requires precise demand forecasting. The sporadic and irregular nature demand, characterized by long intervals between occurrences, results in a significant data imbalance, where events are vastly outnumbered zero-demand periods. This challenge has been largely overlooked forecasting research for parts. Methods: proposed model incorporates the Synthetic Minority Oversampling Technique (SMOTE) to balance dataset uses focal loss enhance sensitivity deep learning models rare events. approach was empirically validated comparing model’s Mean Squared Error (MSE) performance Area Under Curve (AUC). Results: ensemble achieved 47% reduction MSE 32% increase AUC, demonstrating substantial improvements accuracy. Conclusions: findings highlight effectiveness method addressing imbalance improving prediction part providing valuable tool management.

Language: Английский

Citations

0
Efficient Diagnosis of Retinal Diseases Using Convolutional Neural Networks DOI Open Access
Mahır Kaya

Gazi University Journal of Science Part A Engineering and Innovation, Journal Year: 2025, Volume and Issue: 12(1), P. 15 - 35

Published: March 26, 2025

The eye is a vital sensory organ that enables us to fulfill all our life’s needs. Diseases affecting such can have detrimental impact on lives. Although certain conditions are easily managed, others may result in lasting damage or loss of sight if not identified promptly. Problems within the retina improper image focus eyesight. Optical Coherence Tomography (OCT) identify diseases using retinal images taken from side-angle view. Medical analyzed Convolutional Neural Networks (CNNs) automatically diagnose diseases. Doctors reach varying conclusions when diagnosing based medical images. These even contain human error. challenges be overcome with use CNNs. When creating CNN architecture, many hyperparameter values need determined at beginning before training phase. A well-structured design crucial for successful performance lengthy time CNNs makes testing every combination very time-intensive process. This research best hyperparameters by means Bayesian optimization. study employed dataset comprising four categories: DME, CNV, DRUSEN, and NORMAL. With optimization, this proposed model reached an accuracy F1 score 99.69%, outperforming existing findings. will also help doctors make decisions speed up decision-making

Language: Английский

Citations

0
Automated Healthcare Medical Imaging through EOA Optimized Hyperparameter in CNN DOI Open Access
Umang Kumar Agrawal, Nibedan Panda,

Debashreet Das

et al.

Procedia Computer Science, Journal Year: 2025, Volume and Issue: 259, P. 1106 - 1114

Published: Jan. 1, 2025

Language: Английский

Citations

0
Equilibrium Optimization-Based Ensemble CNN Framework for Breast Cancer Multiclass Classification Using Histopathological Image DOI Creative Commons
Yasemın Çetın-Kaya

Diagnostics, Journal Year: 2024, Volume and Issue: 14(19), P. 2253 - 2253

Published: Oct. 9, 2024

: Breast cancer is one of the most lethal cancers among women. Early detection and proper treatment reduce mortality rates. Histopathological images provide detailed information for diagnosing staging breast disease.

Language: Английский

Citations

2
Pixel Embedding for Grayscale Medical Image Classification DOI Creative Commons

Wensu Liu,

Na Lv, Jing Wan

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(16), P. e36191 - e36191

Published: Aug. 1, 2024

In our paper, we present an extension of text embedding architectures for grayscale medical image classification. We introduce a mechanism that combines n-gram features with efficient pixel flattening technique to preserve spatial information during feature representation generation. Our approach involves all pixels in images using combination column-wise, row-wise, diagonal-wise, and anti-diagonal-wise orders. This ensures dependencies are captured effectively the representations. To evaluate effectiveness method, conducted benchmark 5 datasets varying sizes complexities. 10-fold cross-validation showed achieved test accuracy score 99.92 % on Medical MNIST dataset, 90.06 Chest X-ray Pneumonia 96.94 Curated Covid CT 79.11 MIAS dataset 93.17 Ultrasound dataset. The framework reproducible code can be found GitHub at https://github.com/xizhou/pixel_embedding.

Language: Английский

Citations

1
Optimized attention-based lightweight CNN using particle swarm optimization for brain tumor classification DOI
Okan Güder, Yasemın Çetın-Kaya

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 100, P. 107126 - 107126

Published: Nov. 14, 2024

Language: Английский

Citations

1
Advanced CNN-Based Classification and Segmentation for Enhanced Breast Cancer Ultrasound Imaging DOI

Jehad Cheyi,

Yasemın Çetın-Kaya

Gazi University Journal of Science Part A Engineering and Innovation, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 7, 2024

Breast cancer (BC) is one of the primary causes mortality in women globally. Thus, early and exact identification critical for effective treatment. This work investigates deep learning, more especially convolutional neural networks (CNNs), to classify BC from ultrasound images. We worked with a collection breast images 600 patients. Our approach included extensive image preprocessing techniques, such as enhancement overlay methods, before training various learning models particular reference VGG16, VGG19, ResNet50, DenseNet121, EfficientNetB0, custom CNNs. proposed model achieved remarkable classification accuracy 97%, significantly outperforming established like MobileNet, Inceptionv3. research demonstrates ability advanced CNNs, when paired good preprocessing, enhance further used Grad-CAM make interpretable so we may see which parts CNNs focus on making decisions.

Language: Английский

Citations

1
Research on Online Interactive Teaching Platform of College English Combined with Semantic Association Network Modeling DOI Creative Commons

Fengqin Wu

Applied Mathematics and Nonlinear Sciences, Journal Year: 2024, Volume and Issue: 9(1)

Published: Jan. 1, 2024

Abstract The emergence of semantic association networks has injected a new impetus for the development online English teaching and provided model reference design education platforms. In this paper, research an interactive platform college draws on algorithmic advantages associative network utilizes self-operation to realize functions autonomous addition, deletion, modification, checking. text similarity is predicted by word embedding model, convolutional neural network, other algorithms so as better achieve integration resources, connecting knowledge highlighting focus in process English. Dynamic load balancing are used solve problems short-term surges number visits concentration call requests, optimization further realized through genetic finally complete platform. Comparison experiments concluded that proposed paper could hold more stable repair effect when cleaning inconsistent data dataset, effectiveness paper. designed also performs well performance test, with only 0.01% abnormality rate concurrency ability test achieves expected effect.

Language: Английский

Citations

0
Predicting Employee Turnover Through Genetic Algorithm DOI

Vincent Jake Recilla,

Mohn Romy A. Enonaria,

Reyper John Florida

et al.

Published: Aug. 7, 2024

Language: Английский

Citations

0