Medical Rehabilitation Support

Medical Rehabilitation Support

We participated in a study on patient rehabilitation after a stroke. The authors were Department of Rehabilitation Medicine, First Faculty of Medicine, Charles University in Prague and General University Hospital in Prague. The aim of the study were patients with hindered locomotion of the upper arm.

The study compares activity of affected and unaffected arm based on three different criteria.

The goal of the study was to compare affected and unaffected arm. We used three sensors to record activities of one patient. Two in the form of an armband and one attached to a belt. The system we developed is able to determine whether a detected activity is a result of a movement of an arm or of the whole body (like walking or driving).
Read more

Rehabilitation – Illustrative photo

Study parameters

The study was performed by doc. Švestková’s team from the Department of Rehabilitation Medicine, with 21 patients. Data was acquired for four week for each patient. The data was sent continually using a simple PC application or a single purpose docking station that was developed for this study.


Three different criteria were used to assess upper arm activity. Using more criteria allowed us to define a relevance of various movement types for the result of the study. One criteria was sensitive only to significant physical activity, while other emphasized fine motor skills.

Data scrubbing

A significant problem with these types of studies is to remove irrelevant measurements. For example a sensor attached to a fixed hand will still detect movement while the patient is walking, although the patient does not move the hand. Similar problems arise in transit, elevator, etc.

A solution that is often used is an activity journal that the patient has to write himself during the whole course of the study. This solution is difficult for the participants and limits the amount of patients willing to enter the study. It is also difficult to assure the correctness of this data.

Our system is able to detect above mentioned problems and remove any irrelevant data automatically. it uses a third sensor on a belt.