Disability prediction in multiple sclerosis using performance outcome measures and demographic data

Subhrajit Roy*, Diana Mincu*, Negar Rostamzadeh, Chintan Ghate, Nenad Tomasev, Jessica Schrouff, Natalie Harris, Christina Chen, Fletcher Lee Hartsell, Katherine Heller.

Disability prediction in multiple sclerosis using performance outcome measures and demographic data

Subhrajit Roy, Diana Mincu, Negar Rostamzadeh, Chintan Ghate, Nenad Tomasev, Jessica Schrouff, Natalie Harris, Christina Chen, Fletcher Lee Hartsell, Katherine Heller.

Abstract: Literature on machine learning (ML) for multiple sclerosis (MS) has primarily focused on using neuroimaging data such as MRI and clinical laboratory tests for disease identification. However studies have shown that these modalities are not consistent with disease activity such as symptoms or disease progression. Furthermore the cost of collecting data from these modalities is high, leading to scarce evaluations. In this work, we used multi-dimensional and affordable physical and smartphone-based performance outcome measures (POMs) and demographic data to predict multiple sclerosis disease progression. We perform a benchmarking exercise on two datasets (MSOAC and Floodlight) and present results across 8 endpoints and 6 ML models. Our results show that it is possible to predict disease progression using POMs for both a clinical and non-clinical dataset, in the absence of neuroimaging data or clinical laboratory tests.

Poster

Abstract: Literature on machine learning (ML) for multiple sclerosis (MS) has primarily focused on using neuroimaging data such as MRI and clinical laboratory tests for disease identification. However studies have shown that these modalities are not consistent with disease activity such as symptoms or disease progression. Furthermore the cost of collecting data from these modalities is high, leading to scarce evaluations. In this work, we used multi-dimensional and affordable physical and smartphone-based performance outcome measures (POMs) and demographic data to predict multiple sclerosis disease progression. We perform a benchmarking exercise on two datasets (MSOAC and Floodlight) and present results across 8 endpoints and 6 ML models. Our results show that it is possible to predict disease progression using POMs for both a clinical and non-clinical dataset, in the absence of neuroimaging data or clinical laboratory tests.