Fuzzy Logic in Medical Imaging

Thumbnail Image

Files

URL

Journal Title

Journal ISSN

Volume Title

Perustieteiden korkeakoulu | Bachelor's thesis
Electronic archive copy is available locally at the Harald Herlin Learning Centre. The staff of Aalto University has access to the electronic bachelor's theses by logging into Aaltodoc with their personal Aalto user ID. Read more about the availability of the bachelor's theses.
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.

Authors

Date

2025-09-05

Department

Major/Subject

Data Science

Mcode

SCI3095

Degree programme

Aalto Bachelor’s Programme in Science and Technology

Language

en

Pages

25

Series

Abstract

Fuzzy logic offers a principled way to keep vagueness and partial truth explicit in medical images affected by noise, partial-volume effects, and blurred boundaries. This thesis evaluates when fuzzy methods add value over classical deterministic pipelines and modern learning-based approaches in MRI and CT, and identifies practical contexts where they should be preferred. A literature review synthesises findings from peer-reviewed studies on segmentation, detection, and enhancement, comparing fuzzy techniques (e.g., fuzzy c-means variants, fuzzy connectedness, rule-based filters, and hybrid “fuzzy–deep” designs) against representative baselines using commonly reported outcomes (Dice/AUC for segmentation/detection; PSNR/SSIM for enhancement), runtime characteristics, and interpretability. Across the analysed studies, the following results are observed: (1) In MRI, adaptive/spatial FCM and fuzzy connectedness mitigate bias-field artefacts and capture gradual tissue transitions, yielding smoother borders and more stable volumetry than thresholding or k-means; (2) In CT, local-information fuzzy clustering improves candidate generation for low-dose, low-contrast nodules, while fuzzy rule-based filters preserve edges and deliver real-time denoising without retraining, with PSNR/SSIM competitive with classical filters and some CNN baselines; (3) Relative to state-of-the-art probabilistic or deep models, fuzzy pipelines show mixed accuracy but offer consistent advantages in transparency, low latency, cross-protocol robustness, and data-scarce settings; (4) Hybrids that embed fuzzy priors or entropy-aware gates in neural networks often outperform standalone fuzzy systems while retaining uncertainty visibility. Fuzzy logic complements rather than replaces current practice: it is most beneficial when boundaries are ambiguous, acquisition protocols vary, or explanation and speed are required. Practitioners should carry soft memberships through the pipeline, use fuzzy components as lightweight, interpretable modules, and adopt hybrid designs. Future work should emphasise multi-centre benchmarks, clinically meaningful endpoints, differentiable fuzzy operators, and effective visualisation of soft labels for clinical use.

Description

Supervisor

Korpi-Lagg, Maarit

Thesis advisor

Ferranti, Luca

Keywords

fuzzy logic, medical imaging, uncertainty modelling

Other note

Citation

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-202510218392

Collections

[kand] Perustieteiden korkeakoulu / SCI