T1-weighted MRI-based brain tumor classification using hybrid deep learning models

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 27, 2025

Health is fundamental to human well-being, with brain health particularly critical for cognitive functions. Magnetic resonance imaging (MRI) serves as a cornerstone in diagnosing issues, providing essential data healthcare decisions. These images represent vast datasets that are increasingly harnessed by deep learning high-performance image processing and classification tasks. In our study, we focus on classifying tumors—such glioma, meningioma, pituitary tumors—using the U-Net architecture applied MRI scans. Additionally, explore effectiveness of convolutional neural networks including Inception-V3, EfficientNetB4, VGG19, augmented through transfer techniques. Evaluation metrics such F-score, recall, precision, accuracy were employed assess model performance. The segmentation architecture, emerged top performer, achieving an 98.56%, along F-score 99%, area under curve 99.8%, recall precision rates 99%. This study demonstrates U-Net, network accurate tumor early detection treatment planning. Achieving 96.01% cross-dataset validation external cohort, exhibited robust performance across diverse clinical scenarios. Our findings highlight potential enhancing diagnostic informing decision-making neuroimaging, ultimately improving patient care outcomes.

Language: Английский

---

Oral cavity carcinoma detection using BAT algorithm-optimized machine learning models with transfer learning and random sampling

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 192, P. 110250 - 110250

Published: May 5, 2025

Language: Английский

---

Hybrid Deep Transfer Learning for Enhanced Brain Tumor Detection through the Integration of MobileNetV2 and InceptionV3

Procedia Computer Science, Journal Year: 2025, Volume and Issue: 258, P. 2968 - 2977

Published: Jan. 1, 2025

Language: Английский

---

An explainable AI-driven deep neural network for accurate breast cancer detection from histopathological and ultrasound images

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: May 20, 2025

Language: Английский

---

Automated Identification of Breast Cancer Type Using Novel Multipath Transfer Learning and Ensemble of Classifier

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 87560 - 87578

Published: Jan. 1, 2024

Language: Английский

---

A hybrid model for post-treatment mortality rate classification of patients with breast cancer

Healthcare Analytics, Journal Year: 2023, Volume and Issue: 4, P. 100254 - 100254

Published: Sept. 9, 2023

Terminal cancer is not curable and eventually results in death. Breast (BC) a prevalent malignancy affecting women. Although there are prognostic indicators, BC prognosis still challenging because of the intricate connections between various survival factors influencing factors. This study proposes an ensemble classifier for predicting survivability using new post-treatment dataset number survivals. However, classes cases skewed, which caused sub-optimal classification performance. Hence, hybrid sampling scheme Synthetic Minority Over-Sampling TEchnique (SMOTE) Wilson's Edited Nearest Neighbor (ENN) employed to treat class imbalance dataset. Random Forest (RF) classifying The proposed framework performs well terms accuracy, recall two classes, Receiver Operating Characteristics (ROC) Kappa Statistic (KS) metric on demonstrated that RF, with 97.0% accuracy holdout sample, best predictor. prediction superior any noted literature, compared Logistic Regression (LR) Bagging classifiers.

Language: Английский

---

Hybrid Deep Learning for Breast Cancer Diagnosis: Evaluating CNN and ANN on BreakHis_v1_400X

Published: April 2, 2024

Language: Английский

---

Integrative hybrid deep learning for enhanced breast cancer diagnosis: leveraging the Wisconsin Breast Cancer Database and the CBIS-DDSM dataset

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Nov. 1, 2024

The objective of this investigation was to improve the diagnosis breast cancer by combining two significant datasets: Wisconsin Breast Cancer Database and DDSM Curated Imaging Subset (CBIS-DDSM). provides a detailed examination characteristics cell nuclei, including radius, texture, concavity, for 569 patients, which 212 had malignant tumors. In addition, CBIS-DDSM dataset-a revised variant Digital Screening Mammography (DDSM)-offers standardized collection 2,620 scanned film mammography studies, cases that are normal, benign, or include verified pathology data. To identify complex patterns trait diagnoses cancer, used hybrid deep learning methodology combines Convolutional Neural Networks (CNNs) with stochastic gradients method. is CNN training, while dataset fine-tuning maximize adaptability across variety investigations. Data integration, feature extraction, model development, thorough performance evaluation main objectives. diagnostic effectiveness algorithm evaluated area under Receiver Operating Characteristic Curve (AUC-ROC), sensitivity, specificity, accuracy. generalizability will be validated independent validation on additional datasets. This research an accurate, comprehensible, therapeutically applicable detection method advance field. These predicted results might greatly increase early diagnosis, could promote improvements in eventually lead improved patient outcomes.

Language: Английский

---

Performance evaluation of federated learning algorithms using breast cancer dataset

Elsevier eBooks, Journal Year: 2024, Volume and Issue: unknown, P. 95 - 114

Published: Jan. 1, 2024

Language: Английский

---

Enhanced Breast Cancer Diagnosis: Leveraging Customized Transfer Learning with Machine Learning and Attention Mechanisms for Histopathology Image Classification

Published: Aug. 8, 2024

Language: Английский

---