Feasibility of a smartphone-based balance assessment system for subjects with chronic stroke

作者：Hou You Ruei; Chiu Ya Lan; Chiang Shang Lin; Chen Hui Ya; Sung Wen Hsu^*

来源：Computer Methods and Programs in Biomedicine, 2018, 161: 191-195.

DOI：10.1016/j.cmpb.2018.04.027

• Smartphone
• Balance
• Assessment
• Stroke

Background: Stroke is a cerebral artery disease that may lead to long-term disabilities or death. Patients that survive a stroke usually suffer balance impairments, which affect their performance in activities of daily living (ADLs) and quality of life (QoL). In recent years, smartphones have become very popular and have many capabilities. Smartphone built-in sensors have shown their ability and potential in balance performance assessment. However, the feasibility of smartphones on subjects with chronic strokes remains to be proved. Therefore, the purpose of this study is to evaluate the feasibility of a smartphone-based balance assessment system for subjects with chronic stroke.
Methods: Ten subjects with chronic stroke and thirteen healthy adults were recruited in the study. The smartphone HTC 10 (HTC Corporation, Taiwan) was used to perform the balance assessment, and its builtin accelerometer and gyroscope were used to record data from the subjects. Six postures were tested for thirty seconds each: shoulder-width stance (SWS) with eyes opened (E/O) and eyes closed (E/C), feet-together stance (FTS) with E/O and E/C, and semi-tandem stance (STS) with E/O and E/C. The smartphone was fixed to the back of subjects at the second sacral spine (S2) level. The changes registered in the accelerometer and gyroscope data were used to represent the balance performance, in which higher values indicate more instability. Data was analyzed using the independent t-test with the software SPSS 20, and the statistical significance level was set to alpha<0.05.
Results and discussion: Significant difference in the acceleration data was found among subjects with chronic stroke and healthy adults under four assessment postures: SWS with E/C (p = 0.048), FTS with E/O (p = 0.027), FTS with E/C (p = 0.000), and STS with E/C (p = 0.048). Furthermore, according to the gyroscope data, there were significant differences in how the two groups performed the postures. The results demonstrate that a smartphone with a built-in accelerometer and gyroscope can be used to classify balance performances between healthy adults and subjects with chronic stroke.
Conclusion: This study shows that smartphones are feasible to assess balance for subjects with chronic stroke.

• 出版日期2018-7