Spatio-Temporal Registration of Plantar Pressure Image Sequences

Spatio-Temporal Registration of Plantar Pressure Image Sequences Francisco P. M. Oliveira, Carolina S. D. Tábuas, Pedro N. S. Gomes, João Manuel R. S. Tavares Instituto de Engenharia Mecânica e Gestão Industrial, Faculdade de Engenharia, Universidade do Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal, emails: {francisco.oliveira, meb10015, bio10070, tavares}@fe.up.pt The analysis of plantar pressure data provides valuable information on the role of the foot and ankle during gait and other activities. The information gathered can be used, foe example, to define suitable rehabilitation programs through alterations of footwear, foot orthoses, exercise programs and weight-bearing restrictions. It can also assist the diagnosis and rehabilitation of impairments associated with musculoskeletal, integumentary and neurological disorders. Particularly, it has a vital role in the assessment and prevention of ulceration of patients with diabetes and peripheral neuropathy. Generally, the plantar pressure data can be converted into a discrete rectangular array at time instant (static image) or over a period of time (image sequence). It is common accepted that 3 to 5 walking trials enhances the reliability of the dynamic pressure measurements. Hence, the automated spatio-temporal registration of image sequence trials is worthwhile to attain representative plantar pressure image sequences and support comparisons on a pixel-by-pixel basis. Here, the use of automated computational methods in the spatio-temporal registration of plantar pressure image sequences is presented and discussed. It will be shown that the spatiotemporal transformations that best align the pairs of sequences are obtained by minimizing the mean squared error (MSE) among the plantar pressure values. Also, as only intra-subject registration is involved in this work, a rigid geometric model reveals to be efficient in the spatial registration, and linear and cubic B-splines based transformations suitable for the temporal registration. Using synthetic control transformations applied on real image sequences acquired according to a frame rate of 25 Hz (40 ms period), the best mean temporal error achieved was around 0.2 milliseconds (ms) and the spatial error around 0.08 mm, which is several times inferior to the spatial resolution of the acquisition device used. Acknowledgements The first author would like to thank “Fundação Calouste Gulbenkian”, in Portugal, for his PhD grant. This work was partially done in the scope of the research project PTDC/EEACRO/103320/2008, financially supported by “Fundação para a Ciência e a Tecnologia”, in Portugal. References Hughes J, Pratt L, Linge K, Clarke P, Klenerman L (1991) The reliability of pressure measurements: the EMED F system. Clin Biomech 6(1):14-18. Oliveira F P M,Tavares J M R S (2012) Enhanced spatio-temporal alignment of plantar pressure image sequences using B-splines. Med Biol Eng Comput DOI: 10.1007/s11517-012-0988-3 (in press).