MRI Image Enhancement : a Scoping Review of Traditional Techniques and Deep Learning Approaches
Olabode, Muyiwa (2026)
2026
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2026050810274
https://urn.fi/URN:NBN:fi:amk-2026050810274
Tiivistelmä
Magnetic resonance imaging (MRI) is a valuable tool for diagnostic purposes; however, the quality of images can be limited by noise, artefacts, and poor spatial resolution which might impair the accuracy and the efficiency of diagnostics. Along with conven-tional signal processing algorithms, the application of artificial intelligence (AI) and es-pecially deep learning-based systems has recently gained momentum as a method for MRI image improvement. This master’s thesis is presented in the form of an article and describes the process of research that became the basis for the publication of a scien-tific article. This thesis aims at providing reflections on the methodology used in the conduction of scoping reviews.
This thesis implements the methodology of scoping review based on PRISMA-ScR framework. The systematic search in major scientific databases and a patent repository for articles related to MRI image enhancement, i.e., artefact removal, noise reduction, and resolution enhancement, as well as granted patents, was performed. The review procedure included multi-reviewer searching, data charting, and deductive content qual-itative analysis. In order to ensure high quality and credibility of the methodological ap-proach, the Mixed Methods Appraisal Tool (MMAT) was employed for scientific articles only.
Through an emphasis on the research approach and evaluation framework, the thesis outlines the process through which evidence about traditional and deep learning-based methods for MRI enhancement is produced, appraised, and synthesized through typical image quality measures. Key challenges in conducting such research, from issues re-lated to data availability and generalizability to interpretation and clinical application, are identified. This study contributes to health technology management by illuminating the process behind evidence generation for AI-based MRI enhancement, although scientific insights can be found in the peer-reviewed article.
This thesis implements the methodology of scoping review based on PRISMA-ScR framework. The systematic search in major scientific databases and a patent repository for articles related to MRI image enhancement, i.e., artefact removal, noise reduction, and resolution enhancement, as well as granted patents, was performed. The review procedure included multi-reviewer searching, data charting, and deductive content qual-itative analysis. In order to ensure high quality and credibility of the methodological ap-proach, the Mixed Methods Appraisal Tool (MMAT) was employed for scientific articles only.
Through an emphasis on the research approach and evaluation framework, the thesis outlines the process through which evidence about traditional and deep learning-based methods for MRI enhancement is produced, appraised, and synthesized through typical image quality measures. Key challenges in conducting such research, from issues re-lated to data availability and generalizability to interpretation and clinical application, are identified. This study contributes to health technology management by illuminating the process behind evidence generation for AI-based MRI enhancement, although scientific insights can be found in the peer-reviewed article.