Maize seeds forecasting with hybrid directional and bi‐directional long short‐term memory models

Food Science & Nutrition, Journal Year: 2023, Volume and Issue: 12(2), P. 786 - 803

Published: Nov. 9, 2023

The purity of the seeds is one important factors that increase yield. For this reason, classification maize cultivars constitutes a significant problem. Within scope study, six different models were designed to solve A special dataset was created be used in for study. contains total 14,469 images four classes. Images belong types, BT6470, CALIPOS, ES_ARMANDI, and HIVA, taken from BIOTEK company. AlexNet ResNet50 architectures, with transfer learning method, image classification. In order improve success, LSTM (Directional Long Short-Term Memory) BiLSTM (Bi-directional algorithms architectures hybridized. As result classifications, highest success obtained ResNet50+BiLSTM model 98.10%.

Language: Английский

---

Principal component analysis and fine-tuned vision transformation integrating model explainability for breast cancer prediction

Visual Computing for Industry Biomedicine and Art, Journal Year: 2025, Volume and Issue: 8(1)

Published: March 10, 2025

Abstract Breast cancer, which is the most commonly diagnosed cancers among women, a notable health issues globally. cancer result of abnormal cells in breast tissue growing out control. Histopathology, refers to detection and learning diseases, has appeared as solution for treatment it plays vital role its diagnosis classification. Thus, considerable research on histopathology medical computer science been conducted develop an effective method treatment. In this study, vision Transformer (ViT) was employed classify tumors into two classes, benign malignant, Cancer Histopathological Database (BreakHis). To enhance model performance, we introduced novel multi-head locality large kernel self-attention during fine-tuning, achieving accuracy 95.94% at 100× magnification, thereby improving by 3.34% compared standard ViT (which uses self-attention). addition, application principal component analysis dimensionality reduction led improvement 3.34%, highlighting mitigating overfitting reducing computational complexity. final phase, SHapley Additive exPlanations, Local Interpretable Model-agnostic Explanations, Gradient-weighted Class Activation Mapping were used interpretability explainability machine-learning models, aiding understanding feature importance local explanations, visualizing attention. another experiment, ensemble with VGGIN further boosted performance 97.13% accuracy. Our approach exhibited 0.98% 17.13% state-of-the-art methods, establishing new benchmark histopathological image

Language: Английский

---

Advanced deep learning and large language models: Comprehensive insights for cancer detection

Image and Vision Computing, Journal Year: 2025, Volume and Issue: unknown, P. 105495 - 105495

Published: March 1, 2025

Language: Английский

---

Transfer Learning for Accurate Classification of Breast Cancer in Medical Imaging

Published: Nov. 1, 2023

Transfer learning has recently been developed as a powerful technique for accurate classification of medical images. It is predominantly used in deep models to facilitate training on small data sets. based the process leveraging knowledge gained from prior related tasks and transferring it new task. This can be improve accuracy trained images, specifically breast cancer. Such are able provide an improved cancer compared with those standard fashion. Additionally, transfer demonstrate ability increase computational efficiency, reduce over fitting, construct useful representations fewer annotations. particularly imaging due expense difficulty acquiring large annotated datasets purposes. paper explores use imaging, its potential applications diagnosis this disease..

Language: Английский

---

Machine learning and new insights for breast cancer diagnosis

Journal of International Medical Research, Journal Year: 2024, Volume and Issue: 52(4)

Published: April 1, 2024

Breast cancer (BC) is the most prominent form of among females all over world. The current methods BC detection include X-ray mammography, ultrasound, computed tomography, magnetic resonance imaging, positron emission tomography and breast thermographic techniques. More recently, machine learning (ML) tools have been increasingly employed in diagnostic medicine for its high efficiency intervention. subsequent imaging features mathematical analyses can then be used to generate ML models, which stratify, differentiate detect benign malignant lesions. Given marked advantages, radiomics a frequently tool recent research clinics. Artificial neural networks deep (DL) are novel forms that evaluate data using computer simulation human brain. DL directly processes unstructured information, such as images, sounds language, performs precise clinical image stratification, medical record tumour diagnosis. Herein, this review thoroughly summarizes prior investigations on application images intervention radiomics, namely ML. aim was provide guidance scientists regarding use artificial intelligence clinic.

Language: Английский

---