Upcoming project: Subtyping Hereditary Cerebellar Ataxias based on multimodal disease progression patterns Subtyping Hereditary Cerebellar Ataxias based on multimodal disease progression patterns
Principal investigators: Dr Susmita Saha, Prof Nellie Georgiou-Karistianis, Prof Ian Harding and Dr Thiago Rezende, Monash University (Australia)
Scientific summary:
Structural and diffusion MRI studies have identified distinct patterns of neurodegeneration in SCA3 and CANVAS. SCA3 exhibits white matter degeneration in the cerebellar peduncles, corticospinal tract, and cerebral peduncle, with structural abnormalities primarily affecting the cerebellum, cerebellar nuclei, brainstem, and basal ganglia. RFC1-related disorder (CANVAS) is characterised by widespread and symmetric cerebellar and basal ganglia atrophy, brain stem volumetric reduction, and cerebral white matter involvement, particularly in the corpus callosum and deep tracts, with relatively restricted cerebral cortical damage. Despite these findings, the interplay between structural, microstructural and clinical progression remains poorly understood.
Understanding multimodal progression patterns is crucial for refining disease subtyping, enhancing prognosis and optimising clinical trials. This project aims to bridge this gap by developing a machine learning-driven subtyping framework that integrates longitudinal multimodal biomarkers (structural MRI, diffusion imaging and clinical scores) to identify distinct disease progression subtypes across SCA3 and CANVAS. The researchers also aim to identify shared distinct temporal patterns across structural degeneration and clinical symptoms, highlighting how they evolve in relation to each other over time, and identify whether progression subtypes are meaningful. Finally, the researchers plan to develop a framework that integrates diverse data types and updates with new patient data to enhance stratification. The researchers are also exploring the possibility of also studying SCA1 as part of this project.
Lay summary:
Ataxia is a condition that affects how people move. It can cause problems with balance, walking, speaking, and doing everyday tasks. People with hereditary cerebellar ataxias (HCAs) slowly lose control of their muscles over time. Although genetic tests can tell us what type of ataxia someone has, they do not tell us how fast the condition will get worse or how it might look different from person to person. This makes it hard for doctors to give clear answers or suggest treatments that are right for each person.
In this project, the researchers aim to use artificial intelligence to look closely at brain scans and other health data from people with SCA3 and CANVAS, which are different types of HCAs. By looking at how quickly the brain changes and symptoms get worse, the researchers hope to discover different types (or patterns) of ataxia. In simple terms, this means that not everyone with ataxia is the same. Understanding these differences can help doctors make more accurate diagnoses and give care that’s better suited to each person. They will also look at how factors like age at symptom onset, disease duration, and genetic background influence how the condition develops.
In the future, this work might help improve clinical trials by making sure people are grouped by the kind of ataxia they have. That way, researchers can better test if new treatments work and how they may work on different subtypes of ataxia. The methods the researchers develop could also be used for other types of hereditary ataxia, helping even more people. The researchers are also exploring the possibility of also studying SCA1 as part of this project.
This project is due to begin on 1st November 2025.
