Master Thesis - Information Technology

Job Description

Uppsala Monitoring Centre is a dynamic non-profit organization with the mission of improving the safety of medicines throughout the world. By managing and analyzing VigiBase, WHO’s global database of reported potential side effects of medicinal products, consisting of more than 35 million reports from patients and health care professionals in more than 150 countries, UMC has a unique role in detecting the adverse effects of medicines, for which it is recognized internationally.

UMC is opening master thesis positions for 2024. We are looking for creative and motivated students with a keen interest in data science or software development. We have three ideas for possible theses within the field of information technology.

Please note that these theses proposals are not fully complete! Together with you, we will discuss the ideas further to tailor them for you.

Thesis 1: Automating Adverse Event Reporting in Healthcare
Are you interested in improving the efficiency and accuracy of reporting adverse events in the healthcare sector? Do you want to learn how to use cutting-edge tools to automate the process of digitizing paper forms? If yes, then you might want to check out this thesis project.

The project will compare different tools that can recognize handwritten text from paper forms and convert them into digital data. You would evaluate these tools based on three criteria:

• How well they recognize the handwritten text?
• How well they improve the work situation for the healthcare sector?
• How many languages they support and how well they support them?

By doing this project, you will gain valuable skills and knowledge in using modern technologies to solve real-world problems. You will also contribute to the advancement of healthcare quality and safety by reducing human errors and delays in reporting adverse events.

Thesis 2: Climate-Aware Decision Support for Azure Resource Selection: A Multifaceted Approach

This master thesis aims to bridge the gap between environmental responsibility and operational efficiency within IT organizations. By developing a holistic decision support tool, it empowers organizations to make informed choices that consider climate impact without compromising on their budget and functional needs. Furthermore, it offers a step toward a more sustainable IT landscape, aligning with the growing demand for ecologically responsible business practices. The suggestion is to use the data provided by Gaia Gen to create a decision support for the development teams, as well as architects and infrastructure leads, when making decisions on what services and resources to use at the UMC. This tool would also consider other important aspects of this decision, such as functionality, security, and budget.

Questions to answer include:

• What specific criteria and metrics should be considered in the decision support tool to guide developer teams, architects, and infrastructure leads in selecting services and resources that align with our environmental responsibility goals, while also addressing functionality, security, and budget considerations?
• How can we measure the impact and success of implementing this holistic decision support tool within our IT organization in terms of reducing our climate impact and contributing to a more sustainable IT landscape? What key performance indicators (KPIs) should be used to assess its effectiveness over time?

Thesis 3: Preprocessing drug information

When coding big data sets of drug information, an important first step is preprocessing. Successful preprocessing can eliminate repetitive tasks and allow the experts examining the data more time to review the more complex information.

This master thesis aims to improve our preprocessing of drug information. In one of our applications, we import a lot of drug information. Before our pharmacists can review this information, they need to code active ingredient and Market Authorization Holder to our standardized dictionary. This process often requires the pharmacist to correct spelling mistakes and remove unnecessary information.

Examples include:
paracetamol 200 mg -> Paracetamol
Astra Zeneca AB (Sweden) -> Astra Zeneca

A suggested technique for this problem would be to use a Machine learning algorithm trained on our previously coded drug information.

Please include which theses/thesis you are interested in, in your application

Techniques that may be used depending on the project:

• Neural networks
• Machine learning
• Python, C#, SQL, R
• Agile ways of working

Benefits for you:

• Situated centrally in Uppsala, possibility to work both from the office or other location
• You will be part of a team and be mentored by UMC employees
• Work in an international and inclusive environment
• Possibility to publish your work in a scientific conference or journal
• Compensation on completion of project