MERGE: A model for multi-input biomedical federated learning DOI Creative Commons
Bruno Casella, Walter Riviera, Marco Aldinucci

и другие.

Patterns, Год журнала: 2023, Номер 4(11), С. 100856 - 100856

Опубликована: Окт. 6, 2023

Driven by the deep learning (DL) revolution, artificial intelligence (AI) has become a fundamental tool for many biomedical tasks, including analyzing and classifying diagnostic images. Imaging, however, is not only source of information. Tabular data, such as personal genomic data blood test results, are routinely collected but rarely considered in DL pipelines. Nevertheless, requires large datasets that often must be pooled from different institutions, raising non-trivial privacy concerns. Federated (FL) cooperative paradigm aims to address these issues moving models instead across institutions. Here, we present federated multi-input architecture using images tabular methodology enhance model performance while preserving privacy. We evaluated it on two showcases: prognosis COVID-19 patients' stratification Alzheimer's disease, providing evidence enhanced accuracy F1 scores against single-input improved generalizability non-federated models.

Язык: Английский

Deep Learning Approaches for Object Detection in Autonomous Driving: Smart Cities Perspective DOI
Othman Omran Khalifa, Hanita Daud, Elmustafa Sayed Ali

и другие.

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
Automatic colorectal cancer detection using machine learning and deep learning based on feature selection in histopathological images DOI

Hazrat Junaid,

Fatemeh Daneshfar, Mahmud Abdulla Mohammad

и другие.

Biomedical Signal Processing and Control, Год журнала: 2025, Номер 107, С. 107866 - 107866

Опубликована: Март 27, 2025

Язык: Английский

Процитировано

0
EPDTNet + -EM: Advanced Transfer Learning and SubNet Architecture for Medical Image Diagnosis DOI

K. Dhivya,

K Sangamithrai,

Indra Priyadharshini S

и другие.

Cognitive Computation, Год журнала: 2025, Номер 17(2)

Опубликована: Апрель 1, 2025

Язык: Английский

Процитировано

0
An End-to-End Deep Learning Framework for Accurate Tumor Detection and Segmentation in Brain MRI Scans DOI
A. V. Kalpana,

B. Bharathi

Advances in medical diagnosis, treatment, and care (AMDTC) book series, Год журнала: 2025, Номер unknown, С. 113 - 132

Опубликована: Фев. 14, 2025

This chapter presents an automated biomedical image classification system, HBDL-FBTA (Hybrid Bio-inspired Deep Learning with Fusion Brain Tumor Analysis), focused on brain tumors—abnormal cell growths in the or surrounding tissues that require early, accurate detection for effective treatment. The employs pre-processing to enhance quality, Swin-UNet-based segmentation precise region delineation, and fusion-based feature extraction robust acquisition. It uses Humpback Whale Optimization Simulated Annealing (HSSA) parameter tuning a Gated Recurrent Unit (GRU) reliable classification. Simulations benchmark datasets, including BraTS2017, demonstrate superior performance, achieving accuracies of 94.51% ISIC 2017 95.38% 2020 datasets. Future work will focus evaluating computational complexity large-scale integrating multi-modal imaging data, developing interpretable deep learning models clinical adoption reliability.

Язык: Английский

Процитировано

0
MERGE: A model for multi-input biomedical federated learning DOI Creative Commons
Bruno Casella, Walter Riviera, Marco Aldinucci

и другие.

Patterns, Год журнала: 2023, Номер 4(11), С. 100856 - 100856

Опубликована: Окт. 6, 2023

Driven by the deep learning (DL) revolution, artificial intelligence (AI) has become a fundamental tool for many biomedical tasks, including analyzing and classifying diagnostic images. Imaging, however, is not only source of information. Tabular data, such as personal genomic data blood test results, are routinely collected but rarely considered in DL pipelines. Nevertheless, requires large datasets that often must be pooled from different institutions, raising non-trivial privacy concerns. Federated (FL) cooperative paradigm aims to address these issues moving models instead across institutions. Here, we present federated multi-input architecture using images tabular methodology enhance model performance while preserving privacy. We evaluated it on two showcases: prognosis COVID-19 patients' stratification Alzheimer's disease, providing evidence enhanced accuracy F1 scores against single-input improved generalizability non-federated models.

Язык: Английский

Процитировано

8