J4 ›› 2012, Vol. 9 ›› Issue (3): 343-352.doi: 10.1016/S1672-6529(11)60128-5

• 论文 • 上一篇    下一篇

Plantar ROI Characterization during the Stance Phase of Gait Based on a Low-cost Pressure Acquisition Platform

Zhanyong Mei, Guoru Zhao, Qingsong Zhu, Lei Wang   

  1. Institute of Biomedical and Health Engineering, Shenzhen Institutes of Advanced Technology,
    Chinese Academy of Sciences, Shenzhen 518055, P. R. China
  • 出版日期:2012-09-30
  • 通讯作者: Lei Wang E-mail:Wang.lei@siat.ac.cn

Plantar ROI Characterization during the Stance Phase of Gait Based on a Low-cost Pressure Acquisition Platform

Zhanyong Mei, Guoru Zhao, Qingsong Zhu, Lei Wang   

  1. Institute of Biomedical and Health Engineering, Shenzhen Institutes of Advanced Technology,
    Chinese Academy of Sciences, Shenzhen 518055, P. R. China
  • Online:2012-09-30
  • Contact: Lei Wang E-mail:Wang.lei@siat.ac.cn

摘要:

Plantar Region of Interest (ROI) detection is important for the early diagnosis and treatment of morphologic defects of the foot and foot bionic research. Conventional methods have employed complex procedures and expensive instruments which prohibit their widespread use in healthcare. In this paper an automatic plantar ROIs detection method using a customized low-cost pressure acquisition device is proposed. Plantar pressure data and 3D motion capture data were collected from 28 subjects (14 healthy subjects and 14 subjects with hallux valgus). The maximal inter-frame difference during the stance phase was calculated. Consequently, the ROIs were defined by the first-order difference in combination with prior anatomic knowledge. The anatomic locations were determined by the maximal inter-frame difference and second maximal inter-frame difference, which nearly coincided. Our system can achieve average recognition accuracies of 92.90%, 89.30%, 89.30%, 92.90%, 92.90%, and 89.30% for plantar ROIs hallux and metatarsi I–V, respectively, as compared with the annotations using the 3D motion capture system. The maximal difference of metatarsus heads II-V, and the impulse of the medial and lateral heel features made a significant contribution to the classification of hallux valgus and healthy subjects with ? 80% sensitivity and specificity. Furthermore, the plantar pressure acquisition system is portable and convenient to use, thus can be used in home- or community-based healthcare applications.

关键词: dynamic plantar pressure, ROIs, inter-frame difference, foot biomechanics, Fisher discrimination

Abstract:

Plantar Region of Interest (ROI) detection is important for the early diagnosis and treatment of morphologic defects of the foot and foot bionic research. Conventional methods have employed complex procedures and expensive instruments which prohibit their widespread use in healthcare. In this paper an automatic plantar ROIs detection method using a customized low-cost pressure acquisition device is proposed. Plantar pressure data and 3D motion capture data were collected from 28 subjects (14 healthy subjects and 14 subjects with hallux valgus). The maximal inter-frame difference during the stance phase was calculated. Consequently, the ROIs were defined by the first-order difference in combination with prior anatomic knowledge. The anatomic locations were determined by the maximal inter-frame difference and second maximal inter-frame difference, which nearly coincided. Our system can achieve average recognition accuracies of 92.90%, 89.30%, 89.30%, 92.90%, 92.90%, and 89.30% for plantar ROIs hallux and metatarsi I–V, respectively, as compared with the annotations using the 3D motion capture system. The maximal difference of metatarsus heads II-V, and the impulse of the medial and lateral heel features made a significant contribution to the classification of hallux valgus and healthy subjects with ? 80% sensitivity and specificity. Furthermore, the plantar pressure acquisition system is portable and convenient to use, thus can be used in home- or community-based healthcare applications.

Key words: dynamic plantar pressure, ROIs, inter-frame difference, foot biomechanics, Fisher discrimination