University of Twente Student Theses
Comparing Supervised and Unsupervised Models for Disease Detection
Biju, Vivan (2024) Comparing Supervised and Unsupervised Models for Disease Detection.
|
PDF
480kB |
| Abstract: | The automation of detecting rare diseases accurately has proven to be challenging due to the lack of enough data on cases with such diseases. This research explores and compares the effectiveness of supervised and unsupervised machine learning methods to identify rare disease patterns in medical imaging. Supervised learning methods are known to be very accurate but unsupervised methods are adept at handling unlabelled data and identifying the anomalies that exist. Thus, a thorough comparison will bring more clarity on what methods to prefer for anomaly detection. This research also focuses primarily on using different types of autoencoders to detect anomalies in medical images. To thoroughly assess the supervised models(Resnet50 and Densenet121) and the unsupervised models(Autoencoders and Variational Autoencoders), this study will make use of multiple datasets: Retinal OCT Images, Brain Tumor MRI Scans, COVID-19 Radiography images and ISIC 2018 HAM10000 dataset. By conducting this comparative analysis, this research aims to shed light on suitable machine learning models for the use of detection of diseases in medical images. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science BSc (56964) |
| Link to this item: | https://purl.utwente.nl/essays/100774 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page