University of Twente Student Theses


An automatic Segmentation method and prediction method for skin prick test results

Geessinck, M.S.M. (2021) An automatic Segmentation method and prediction method for skin prick test results.

Full text not available from this repository.

Full Text Status:Access to this publication is restricted
Embargo date:2 July 2025
Abstract:The Skin Prick Test (SPT) is the first step in allergy diagnostics and it is a widely used tool for over many decades. The inter-observer variability and measurement errors of this test lead to a lack of objectivity and reproducibility, which makes inter-institutional comparison of test results challenging. Besides, the current prediction model of SPT results lead to a relatively high amount of unnecessary follow-up diagnosis with additional costs. The objectives of this work are to automate the SPT result reading and to improve the patient outcome prediction. In order to automate the SPT reading process, a deep learning network is proposed. This network extracts the wheal areas from photos taken from the patient’s forearm. To fulfill the aim of an improved prediction model, multiple clinical predictors are incorporated into more complex predictive models. Four different machine learning models are compared to the current way of classifying used in clinic. The two studies can be integrated by the development of a tool that automates the SPT result reading and predicts the patient outcome. The conclusions from this study will support the development of such a tool, leading to a more accurate, quantitative, objective and reproducible allergy diagnosis.
Item Type:Essay (Master)
Faculty:TNW: Science and Technology
Subject:44 medicine, 54 computer science
Programme:Biomedical Engineering MSc (66226)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page