Bridging Data Gaps: Multimodal Text-Image Fusion for Improved Medical Diagnosis in Low-Resource Settings
Machines and Algorithms · 16.03.2016 · 5 (01) · 44-56
DOI: 10.66108/mna.v5i01.89 ↗Kurz gefasst
Fachartikel in der Zeitschrift Machines and Algorithms, online erschienen am 16.03.2016. Verfasst von Muhammad Usama Abrar. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Medical diagnosis often requires the integration of heterogeneous information sources such as clinical text and medical images. While artificial intelligence (AI)–based multimodal diagnostic systems have shown promising results in data-rich environments, their applicability in low-resource settings remains limited due to data scarcity, weak infrastructure, and limited expert availability. This paper surveys existing multimodal text–image fusion techniques and proposes data-efficient strategies tailored for low-resource healthcare environments. We discuss challenges related to data acquisition, annotation, alignment, and computational constraints, and analyze fusion architectures including early, late, and intermediate fusion. Ethical, practical, and deployment considerations are also examined. By focusing on lightweight, interpretable, and clinically grounded multimodal fusion approaches, this study highlights pathways for improving diagnostic accuracy and accessibility in underserved healthcare systems.
Zitieren
APA
Abrar, M. U. (2016). Bridging Data Gaps: Multimodal Text-Image Fusion for Improved Medical Diagnosis in Low-Resource Settings. Machines and Algorithms, 5(01), 44-56. https://doi.org/10.66108/mna.v5i01.89
Vancouver
Abrar MU. Bridging Data Gaps: Multimodal Text-Image Fusion for Improved Medical Diagnosis in Low-Resource Settings. Machines and Algorithms. 2016;5(01):44-56. doi:10.66108/mna.v5i01.89
BibTeX
@article{Abrar2016,
author = {Abrar, Muhammad Usama},
title = {Bridging Data Gaps: Multimodal Text-Image Fusion for Improved Medical Diagnosis in Low-Resource Settings},
journal = {Machines and Algorithms},
year = {2016},
volume = {5},
number = {01},
pages = {44--56},
doi = {10.66108/mna.v5i01.89}
}
EUVIMED