ML in limited medical datasets usually means doing precision guesswork on unreliable data provided by those with high expectations. The first part of this talk will focus on issues that data scientists and engineers have to address when working with this kind of data (e.g. unreliable labels, effect of covariates, necessity of clinical interpretability, difficulties with fusing more data sets), then some common ML approaches in this field will be described (yes, sometimes you have to go from deep neural networks to logistic regression) and finally some specific ML-based research studies in the field of neurology and psychology (diagnosis and prediction of motor/non-motor deficits in Parkinson’s disease, assessment of graphomotor disabilities in children with developmental dysgraphia) will be provided.
Jiri Mekyska is the head of the BDALab (Brain Diseases Analysis Laboratory) at the Brno University of Technology, where he leads a group of data scientists and biomedical engineers. He deals especially with the research of non-invasive quantitative analysis of neurodegenerative and neurodevelopmental disorders based on speech and handwriting processing. In cooperation with neurologists/psychologists from different countries, he develops diagnostic/monitoring systems focused on Parkinson’s disease and developmental dysgraphia.
ML in limited medical datasets usually means doing precision guesswork on unreliable data provided by those with high expectations. The first part of this talk will focus on issues that data scientists and engineers have to address when working with this kind of data (e.g. unreliable labels, effect of covariates, necessity of clinical interpretability, difficulties with fusing more data sets), then some common ML approaches in this field will be described (yes, sometimes you have to go from deep neural networks to logistic regression) and finally some specific ML-based research studies in the field of neurology and psychology (diagnosis and prediction of motor/non-motor deficits in Parkinson’s disease, assessment of graphomotor disabilities in children with developmental dysgraphia) will be provided.
Jiri Mekyska is the head of the BDALab (Brain Diseases Analysis Laboratory) at the Brno University of Technology, where he leads a group of data scientists and biomedical engineers. He deals especially with the research of non-invasive quantitative analysis of neurodegenerative and neurodevelopmental disorders based on speech and handwriting processing. In cooperation with neurologists/psychologists from different countries, he develops diagnostic/monitoring systems focused on Parkinson’s disease and developmental dysgraphia.
ML in limited medical datasets usually means doing precision guesswork on unreliable data provided by those with high expectations. The first part of this talk will focus on issues that data scientists and engineers have to address when working with this kind of data (e.g. unreliable labels, effect of covariates, necessity of clinical interpretability, difficulties with fusing more data sets), then some common ML approaches in this field will be described (yes, sometimes you have to go from deep neural networks to logistic regression) and finally some specific ML-based research studies in the field of neurology and psychology (diagnosis and prediction of motor/non-motor deficits in Parkinson’s disease, assessment of graphomotor disabilities in children with developmental dysgraphia) will be provided.
Jiri Mekyska is the head of the BDALab (Brain Diseases Analysis Laboratory) at the Brno University of Technology, where he leads a group of data scientists and biomedical engineers. He deals especially with the research of non-invasive quantitative analysis of neurodegenerative and neurodevelopmental disorders based on speech and handwriting processing. In cooperation with neurologists/psychologists from different countries, he develops diagnostic/monitoring systems focused on Parkinson’s disease and developmental dysgraphia.
ML in limited medical datasets usually means doing precision guesswork on unreliable data provided by those with high expectations. The first part of this talk will focus on issues that data scientists and engineers have to address when working with this kind of data (e.g. unreliable labels, effect of covariates, necessity of clinical interpretability, difficulties with fusing more data sets), then some common ML approaches in this field will be described (yes, sometimes you have to go from deep neural networks to logistic regression) and finally some specific ML-based research studies in the field of neurology and psychology (diagnosis and prediction of motor/non-motor deficits in Parkinson’s disease, assessment of graphomotor disabilities in children with developmental dysgraphia) will be provided.
Jiri Mekyska is the head of the BDALab (Brain Diseases Analysis Laboratory) at the Brno University of Technology, where he leads a group of data scientists and biomedical engineers. He deals especially with the research of non-invasive quantitative analysis of neurodegenerative and neurodevelopmental disorders based on speech and handwriting processing. In cooperation with neurologists/psychologists from different countries, he develops diagnostic/monitoring systems focused on Parkinson’s disease and developmental dysgraphia.
