Portable Medical Imaging in Modern Healthcare: Fundamentals, AI-Based Taxonomy, Image Quality, and Open Challenges

TL;DR

This review discusses recent advances in portable medical imaging, focusing on AI techniques for improving image quality and addressing challenges in resource-limited healthcare settings.

Contribution

It introduces an AI-based taxonomy for PMI methods and emphasizes the relationship between image quality, AI robustness, and clinical usability.

Findings

AI methods improve image reconstruction and quality assessment in PMI.

Portable imaging devices face challenges from motion artifacts and environmental factors.

The review identifies gaps and future directions for reliable PMI systems.

Abstract

Portable medical imaging (PMI) has emerged as an important solution for point-of-care diagnosis in emergency, rural, and resource-limited settings where conventional imaging infrastructure is not readily available. Modalities such as portable computed tomography, portable magnetic resonance imaging, portable ultrasound, and wireless capsule endoscopy improve access to timely diagnosis, but they remain highly vulnerable to image-quality degradation caused by motion artifacts, environmental interference, hardware limitations, and unstable acquisition conditions. This review provides a systematic and quality-centered synthesis of recent advances in PMI. It introduces a taxonomy of AI-based PMI methods spanning machine learning, deep learning, transfer learning, and Transformer-based approaches, and examines their roles in image enhancement, reconstruction, quality assessment, detection, and classification. The review also analyzes PMI devices, sensing pipelines, modality-specific distortions, evaluation metrics, and publicly available datasets. In contrast to existing surveys that are mainly modality-driven or application-focused, this work emphasizes the relationship between image quality, AI robustness, and clinical usability in portable settings. Finally, it identifies current research gaps and outlines future directions toward reliable, interpretable, and clinically deployable PMI systems.