Related Experiment Video
Updated: Jun 19, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.7K
Identifying and training deep learning neural networks on biomedical-related datasets
Alan E Woessner1,2, Usman Anjum1,3,4, Hadi Salman1,4
1Arkansas Integrative Metabolic Research Center, University of Arkansas, Fayetteville, AR.
Briefings in Bioinformatics
|July 23, 2024
Summary
This module introduces deep learning for biomedical data analysis using cloud computing. It covers neural networks for image and ATAC-seq data, aiding researchers in applying AI tools.
Area of Science:
- Computational Biology
- Bioinformatics
- Artificial Intelligence in Medicine
Background:
- Biomedical datasets are growing in size and complexity, challenging traditional interpretation methods.
- Deep learning neural networks offer powerful tools for novel biomedical research but require significant computational resources and expertise.
- Existing deep learning resources often lack specific guidance for biomedical applications and cloud-based implementation.
Purpose of the Study:
- To develop an interactive learning module for implementing deep learning algorithms in biomedical image and ATAC-seq data analysis.
- To provide users with practical knowledge of neural network architectures and their applications in research.
- To demonstrate the utility of cloud computing platforms for accessing data and performing complex analyses.
Main Methods:
- Development of a cloud-based learning module, the NIGMS Sandbox for Cloud-based Learning, hosted on Google Cloud Platform.
- Creation of four submodules focusing on classification, augmentation, segmentation, and regression tasks relevant to biomedical data.
- Integration of detailed code, explanations, quizzes, and challenges to facilitate user training and practical application.
Main Results:
- The module successfully delivers interactive learning materials for deep learning in biomedical contexts.
- It provides practical examples and code for analyzing both image and ATAC-seq data.
- The platform highlights the ease of using cloud resources for deep learning tasks.
Conclusions:
- The NIGMS Sandbox learning module effectively enables users to identify and apply appropriate deep learning neural networks to biomedical data.
- Cloud computing facilitates accessible and efficient implementation of deep learning for researchers and clinicians.
- This resource aims to bridge the gap between deep learning capabilities and their practical application in biomedical research.
