Thomas Neumuth1, Christian Meissner. 1. Universität Leipzig, Innovation Center Computer Assisted Surgery (ICCAS), Semmelweisstr. 14, 04103, Leipzig, Germany. thomas.neumuth@medizin.uni-leipzig.de
Abstract
PURPOSE: Automatic online recognition of surgical instruments is required to monitor instrument use for surgical process modeling. A system was developed and tested using available technologies. METHODS: A recognition system was developed using RFID technology to identify surgical activities. Information fusion for online recognition of surgical process models was conceived as a layer model to abstract information from specific sensor technologies. Redundant, complementary, and cooperative sensor signal fusion was used in the layer model to increase the surgical instrument recognition rate. Several different information fusion strategies were evaluated for situation recognition abilities in a mock-up environment based on simulations of surgical processes. RESULTS: This information fusion system was able to reliably detect, identify, and localize surgical instruments in an interventional suite. A combination of information fusion strategies was able to achieve a correct classification rate of 97% and was as effective as observer-based acquisition methods. CONCLUSION: Different information fusion strategies for the recognition of surgical instruments were evaluated, showing that redundant, complementary, and cooperative information fusion is feasible for recognition of surgical work steps. A combination of sensor- and observer-based modeling strategies provides the most robust solution for surgical process models.
PURPOSE: Automatic online recognition of surgical instruments is required to monitor instrument use for surgical process modeling. A system was developed and tested using available technologies. METHODS: A recognition system was developed using RFID technology to identify surgical activities. Information fusion for online recognition of surgical process models was conceived as a layer model to abstract information from specific sensor technologies. Redundant, complementary, and cooperative sensor signal fusion was used in the layer model to increase the surgical instrument recognition rate. Several different information fusion strategies were evaluated for situation recognition abilities in a mock-up environment based on simulations of surgical processes. RESULTS: This information fusion system was able to reliably detect, identify, and localize surgical instruments in an interventional suite. A combination of information fusion strategies was able to achieve a correct classification rate of 97% and was as effective as observer-based acquisition methods. CONCLUSION: Different information fusion strategies for the recognition of surgical instruments were evaluated, showing that redundant, complementary, and cooperative information fusion is feasible for recognition of surgical work steps. A combination of sensor- and observer-based modeling strategies provides the most robust solution for surgical process models.
Authors: Remko van der Togt; Erik Jan van Lieshout; Reinout Hensbroek; E Beinat; J M Binnekade; P J M Bakker Journal: JAMA Date: 2008-06-25 Impact factor: 56.272
Authors: Thomas Neumuth; Pierre Jannin; Juliane Schlomberg; Jürgen Meixensberger; Peter Wiedemann; Oliver Burgert Journal: Int J Comput Assist Radiol Surg Date: 2010-06-06 Impact factor: 2.924
Authors: Philipp Liebmann; Jürgen Meixensberger; Peter Wiedemann; Thomas Neumuth Journal: Int J Comput Assist Radiol Surg Date: 2013-03-07 Impact factor: 2.924
Authors: Annetje C P Guédon; M Paalvast; F C Meeuwsen; D M J Tax; A P van Dijke; L S G L Wauben; M van der Elst; J Dankelman; J J van den Dobbelsteen Journal: J Med Syst Date: 2016-10-14 Impact factor: 4.460