About
With people living ever longer, age-related conditions such as Alzheimer’s disease (AD) are becoming an urgent healthcare problem. To meet this challenge, we need both effective therapies and widely accessible diagnostic tools that can detect individuals at early disease stages—as they are most likely to benefit from these therapies. The project in hand aims to tackle an important challenge in AD diagnostics: two of the most trusted tools used to diagnose AD: a brain scan (amyloid PET) and spinal-fluid test (CSF) disagree in a meaningful portion of patients—about 5–20%. This discordance is likely rooted in the fact that the two methods capture different aspects of the same disease process. Amyloid PET visualizes the accumulation of insoluble amyloid plaques in the brain, whereas CSF reflects soluble amyloid dynamics, including how amyloid is produced and cleared.
At the same time, blood tests for AD are nearing implementation in clinical practice. Compared with CSF tap and especially with PET, plasma biomarkers are less invasive, more accessible, and less costly. Highly accurate assays, such as plasma p-tau217, therefore hold great promise as scalable diagnostic tools. However, important questions remain regarding how these markers relate to established CSF and PET measures, particularly in cases where CSF and PET results disagree. In addition, the new assays’ multiplexing capacity (i.e., NULISA) enables the simultaneous measurement of many proteins, offering a unique opportunity to explore the biological mechanisms underlying PET–CSF discordance.
In this project, we will leverage plasma biomarkers alongside CSF and PET to address key questions related to diagnostic uncertainty, biological mechanisms and heterogeneity, as well as optimal patient management.
To whom this project might be a good fit:
• you enjoy independent problem-solving
• you have an affinity for data and coding
• you are not afraid of handling data through the command line
• you enjoy working in an international environment
The project may include the following tasks:
• handling MRI and PET scans, including image quality control (QC)
• clinical data analysis and group comparisons (e.g., discordant vs concordant patients)
• diagnostic accuracy evaluation (ROC curves, AUROC)
• Bayesian modelling of diagnostic probability
• multivariable regression

What we offer:
• broad training opportunities in clinical and translational Alzheimer’s disease research
• supervision and support throughout the project
• opportunities to participate in seminars, journal clubs, and scientific discussions
• experience working in a friendly, collaborative, and international research environment

The student will be based at Karolinska campus South in central Flemingsberg. Though, it will be possible to perform part of the work remotely.

The details and specific aims of the project will be developed together with the student to accommodate specific interests.