Interpretable evaluation for the Brunnstrom recovery stage of the lower limb based on wearable sensors

Data on stroke epidemiology and availability of hospital-based stroke services around the world are important for guiding policy decisions and healthcare planning. To provide the most current incidence, mortality and case–fatality data on stroke and describe current availability of stroke units around the world by country. We searched multiple databases (based on our existing search strategy) to identify new original manuscripts and review articles published between 1 June 2016 and 31 October 2018 that met the ideal criteria for data on stroke incidence and case–fatality. For data on the availability of hospital-based stroke services, we searched PubMed for all literature published up until 31 June 2018. We further screened reference lists, citation history of manuscripts and gray literature for this information. Mortality codes for International Classification of Diseases-9 and International Classification of Diseases-10 were extracted from the World Health Organization mortality database for each country providing these data. Population denominators were obtained from the World Health Organization, and when these were unavailable within a two-year period of mortality data, population denominators within a two-year period were obtained from the United Nations. Using country-specific population denominators and the most recent years of mortality data available for each country, we calculated both the crude mortality from stroke and mortality adjusted to the World Health Organization world population. Since our last report in 2017, there were two countries with new incidence studies, China ( n = 1) and India ( n = 2) that met the ideal criteria. New data on case–fatality were found for Estonia and India. The most current mortality data were available for the year 2015 (39 countries), 2016 (43 countries), and 2017 (7 countries). No new data on mortality were available for six countries. Availability of stroke units was noted for 63 countries, and the proportion of patients treated in stroke units was reported for 35/63 countries. Up-to-date data on stroke incidence, case–fatality, and mortality statistics provide evidence of variation among countries and changing magnitudes of burden among high and low–middle income countries. Reporting of hospital-based stroke units remains limited and should be encouraged.

Machine learning methods have been widely used for gait assessment through the estimation of spatio-temporal parameters. As a further step, the objective of this work is to propose and validate a general probabilistic modeling approach for the classification of different pathological gaits. Specifically, the presented methodology was tested on gait data recorded on two pathological populations (Huntington’s disease and post-stroke subjects) and healthy elderly controls using data from inertial measurement units placed at shank and waist. By extracting features from group-specific Hidden Markov Models (HMMs) and signal information in time and frequency domain, a Support Vector Machines classifier (SVM) was designed and validated. The 90.5% of subjects was assigned to the right group after leave-one-subject–out cross validation and majority voting. The long-term goal we point to is the gait assessment in everyday life to early detect gait alterations.