• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Article

Identification of Contact Angle and Heterogenity Detection in Tactile Images

Alexandrov D., Nersisyan S.

Introduction. Automated analysis of tactile images registered by specialized medical tools is a novel domain, which results promptly find their applications in clinical practice. Medical Tactile Endosurgical Complex (MTEC) is currently the only commercially available device for intraoperative instrumental mechanoreceptoric palpation. One of the main challenges related to processing data generated by MTEC is heterogeneity detection in tactile images. This problem is highly important because it is a key step of localization of visually undetectable pathologies using instrumental palpation. Objectives. One of the main difficulties related to the problem of heterogeneity detection is a possibility to vary contact angle between mechanoreceptor and sample during tactile examination, so the aim of the research was to develop a method for automated contact angle identification and detection of heterogeneity in tactile images registered by MTEC. Methods. The proposed method of a tactile press contact angle estimation is based on classification with a specifically designed feature space. For heterogeneity detection we developed two different approaches. The first one is based on separation of heterogeneity detector into several components corresponding to similar contact angles. The second approach uses standard classification approach with contact angle as a high weighted element of the feature space. Results. Validation on a set of samples modeling normal tissues and pathologies showed high accuracy for contact angle identification. Both methods of heterogenity detection provided approximately the same accuracy clearly outperforming previously available methods in case of significant deviations of a contact angle from zero. Conclusion. The methods developed provide an accurat solution for problems of contact angle identification and detection of heterogeneity even in case of significant contact angle deviations, and such deviations are unavoidable in clinical practice, especially in minimally-invasive surgery.