(Hedvig Sidenbladh) Kjellstrom Kjellström
Hedvig Kjellstr
o
m
Home

Open Master Project Opportunities

These projects are primarily geared towards KTH Machine Learning, Computer Science, and Robotics Master students. If you are interested in a project, please contact me for more information. We can also formulate a new project relating to my ongoing research.

Causal Inference for Achilles Tendon Rupture Rehabilitation

Machine Learning

In this project, machine learning tools such as causal identification and causal inference will be applied for a health care application.
Despite advancements in treatment modalities, recovery after musculoskeletal injuries, eg. tendon rupture, is still a prolonged process with very variable and often unsatisfactory outcome. Still at one year after their injury patients suffer from symptoms like pain, fatigue, and weakness. Thus, patients are often hindered to return to work.
Identifying key factors causing the different outcomes after the injuries is extremely important for the treatment of such injuries. For example, identifying for which patient groups early load bearing mobilization causes better outcome is unclear but essential.
Causal inference [2][3] which is a machine learning subject that identify causal relationships among observed factors. This project aims to identify causal relationship among observed factors and validated patient reported outcome after Achilles tendon rupture. We will use causal inference to identify potential changes in the outcome wrt different factors.
The student working on this project will also interact with MD.PhD. Ass.Prof. Paul Ackermann from KI, and Dr. Cheng Zhang. Moreover, the student may also have chance to get advice from Ass. Prof Kun Zhang from Carnegie Mellon University.
We are looking for someone that are interested in machine learning and are comfortable with mathematics and coding. Students who are interested in this project are encouraged to read the following related work for this project:

[1] Bridging Medical Data Inference to Achilles Tendon Rupture Rehabilitation. https://arxiv.org/abs/1612.02490
[2] A survey on conditional independence-based causal discovery and functional causal model-based causal discovery. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4841209/
[3] The state-of-the-art theoretical identifiability of the causal direction between two variables. https://arxiv.org/pdf/1205.2599.pdf

Achilles tendon rupture rehabilitation outcome prediction with Probablistic Graphical Models

Machine Learning

In this project, a probabilistic model will be designed and applied for a health care application.
Despite advancements in treatment modalities, recovery after musculoskeletal injuries, eg. tendon rupture, is still a prolonged process with very variable and often unsatisfactory outcome. Still at one year after their injury patients suffer from symptoms like pain, fatigue, and weakness. Thus, patients are often hindered to return to work.
Machine learning has shown great potential for healthcare applications [1][2]. A probabilistic interpretation of machine leaning output is important for medical doctors to use the results. In this project, a student will continue with [1] and build a probabilistic model that can encode sequential structures for sequential measurements and be able to predict rehabilitation outcome given a new patient.
The student working on this project will also interact with MD.PhD. Ass.Prof. Paul Ackermann from KI, and Dr. Cheng Zhang. We are looking for someone that are interested in machine learning and are comfortable with mathematics and coding.
Students who are interested in this project are encouraged to read the following related work for this project.

[1] Bridging Medical Data Inference to Achilles Tendon Rupture Rehabilitation. https://arxiv.org/abs/1612.02490
[2] A Framework for Individualizing Predictions of Disease Trajectories by Exploiting Multi-Resolution Structure. https://arxiv.org/pdf/1601.04674.pdf