We presented a decision support system (DSS) that combines digital spirography and clinical data for differenting between essential tremor (ET) and Parkinson’s disease (PD), including the mixed type tremor (MT) (comorbidity). 122 patients (median age 69 years) with ET, PT, and MT were included in the study. The estimated classification accuracy of the model was 91%, which shows an improvement of 10% compared to the previously reported model. DSS was built using ABML, which combines machine learning with expert knowledge based on the clinical as well as quantitative data from digitalised spirography.
F.22 Improvement to existing health/diagnostic methods/procedures
Tremor is one of the most common disorders in the population of patients diagnosed with movement disorders. In the literature we find several classifications and different types of tremors. Each tremor type has its own characteristics. The most frequently used and widely accepted tremor classification divides tremors according to clinical appearance. First, they are roughly divided into resting tremor and action tremor. Action tremor is then subdivided into postural, kinetic, intention, task specific and isometrictremor. Different types of tremor are further combined into tremor syndromes. Causes and mechanisms of tremor are still unclear. Tremor genesis is explained by four hypothetical mechanisms and one of them is assumed to be dominant for each type of tremor. Correct tremor recognition and diagnosis is necessary for appropriate treatment of tremor patients.
F.22 Improvement to existing health/diagnostic methods/procedures
COBISS.SI-ID: 29488345