A novel technique for analysis of postural information with wearable devices

These days, as many jobs involve sitting behind desks and working with computers for extended periods, more and more people are suffering from back problems. Maintenance of an appropriate posture may prevent future back problems. There are various medical methods for studying postures abnormalities of the back but most of these methods are limited to be utilized in diagnostics and follow-up of treatment and not used in a continuous or in a preventive manner. Therefore, designing and developing methods for measuring, analyzing and reporting of posture information, aimed for prevention of future back problems is of fundamental interest. In this work, a proof-of-concept system, including five accelerometer sensor units is presented. Additionally, an index, which we call spine inclination index (SII), is introduced and used for converting the raw data to meaningful presentable information. Initial evaluation includes measurements with six subjects. Subjects were asked to mimic accentuated kyphotic, straight and accentuated lordotic postures while sitting. Our results show that the designed device and SII index are able to distinguish between different postures very well. In addition, since this device measures the inclination angle of different spinal postures, its output can be directly compared with other widely used methods.