دورية أكاديمية
A Practical Roadmap to Implementing Deep Learning Segmentation in the Clinical Neuroimaging Research Workflow.
العنوان: | A Practical Roadmap to Implementing Deep Learning Segmentation in the Clinical Neuroimaging Research Workflow. |
---|---|
المؤلفون: | Pérez Cáceres, Marco, Gauvin, Alexandre, Dumais, Félix, Iorio-Morin, Christian |
المصدر: | World Neurosurg ; ISSN:1878-8769 |
بيانات النشر: | Elsevier Science |
سنة النشر: | 2024 |
المجموعة: | PubMed Central (PMC) |
مصطلحات موضوعية: | Deep Learning, Neuroimaging, Segmentation |
الوصف: | Thanks to the proliferation of open-source tools, we are seeing an exponential growth of machine learning applications, and its integration has become more accessible, particularly for segmentation tools in neuroimaging. This article explores a generalised methodology that harnesses these tools and aims/enables to expedite and enhance the reproducibility of clinical research. Herein, critical re considerations include hardware, software, neural network training strategies and data labelling guidelines. More specifically, we advocate an iterative approach to model training and transfer learning, focusing on internal validation and outlier handling early in the labelling process and fine-tuning later on. The iterative refinement process allows experts to intervene and improve model reliability whilst cutting down on their time spent in manual work. A seamless integration of the final model's predictions into clinical research is proposed to ensure standardized and reproducible results. In short, this article provides a comprehensive framework for accelerating research using machine learning techniques for image segmentation. |
نوع الوثيقة: | article in journal/newspaper |
اللغة: | English |
العلاقة: | https://doi.org/10.1016/j.wneu.2024.06.026Test; https://pubmed.ncbi.nlm.nih.gov/38866234Test |
DOI: | 10.1016/j.wneu.2024.06.026 |
الإتاحة: | https://doi.org/10.1016/j.wneu.2024.06.026Test https://pubmed.ncbi.nlm.nih.gov/38866234Test |
حقوق: | Copyright © 2024. Published by Elsevier Inc. |
رقم الانضمام: | edsbas.95CD4F25 |
قاعدة البيانات: | BASE |
DOI: | 10.1016/j.wneu.2024.06.026 |
---|