Data-derived models for segmentation with application to surgical assessment and training

Balakrishnan Varadarajan; Carol Reiley; Henry Lin; Sanjeev Khudanpur; Gregory Hager

doi:10.1007/978-3-642-04268-3_53

Data-derived models for segmentation with application to surgical assessment and training

Balakrishnan Varadarajan, Carol Reiley, Henry Lin, Sanjeev Khudanpur, Gregory Hager

Whiting School of Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

54 Scopus citations

Abstract

This paper addresses automatic skill assessment in robotic minimally invasive surgery. Hidden Markov models (HMMs) are developed for individual surgical gestures (or surgemes) that comprise a typical bench-top surgical training task. It is known that such HMMs can be used to recognize and segment surgemes in previously unseen trials [1]. Here, the topology of each surgeme HMM is designed in a data-driven manner, mixing trials from multiple surgeons with varying skill levels, resulting in HMM states that model skill-specific sub-gestures. The sequence of HMM states visited while performing a surgeme are therefore indicative of the surgeon's skill level. This expectation is confirmed by the average edit distance between the state-level "transcripts" of the same surgeme performed by two surgeons with different expertise levels. Some surgemes are further shown to be more indicative of skill than others.

Original language	English (US)
Title of host publication	Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009 - 12th International Conference, Proceedings
Pages	426-434
Number of pages	9
Edition	PART 1
DOIs	https://doi.org/10.1007/978-3-642-04268-3_53
State	Published - 2009
Event	12th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2009 - London, United Kingdom Duration: Sep 20 2009 → Sep 24 2009

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 1
Volume	5761 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	12th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2009
Country/Territory	United Kingdom
City	London
Period	9/20/09 → 9/24/09

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-642-04268-3_53

Cite this

Varadarajan, B., Reiley, C., Lin, H., Khudanpur, S., & Hager, G. (2009). Data-derived models for segmentation with application to surgical assessment and training. In Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009 - 12th International Conference, Proceedings (PART 1 ed., pp. 426-434). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5761 LNCS, No. PART 1). https://doi.org/10.1007/978-3-642-04268-3_53

Data-derived models for segmentation with application to surgical assessment and training. / Varadarajan, Balakrishnan; Reiley, Carol; Lin, Henry et al.
Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009 - 12th International Conference, Proceedings. PART 1. ed. 2009. p. 426-434 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5761 LNCS, No. PART 1).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Varadarajan, B, Reiley, C, Lin, H, Khudanpur, S & Hager, G 2009, Data-derived models for segmentation with application to surgical assessment and training. in Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009 - 12th International Conference, Proceedings. PART 1 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 1, vol. 5761 LNCS, pp. 426-434, 12th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2009, London, United Kingdom, 9/20/09. https://doi.org/10.1007/978-3-642-04268-3_53

Varadarajan B, Reiley C, Lin H, Khudanpur S, Hager G. Data-derived models for segmentation with application to surgical assessment and training. In Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009 - 12th International Conference, Proceedings. PART 1 ed. 2009. p. 426-434. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1). doi: 10.1007/978-3-642-04268-3_53

Varadarajan, Balakrishnan ; Reiley, Carol ; Lin, Henry et al. / Data-derived models for segmentation with application to surgical assessment and training. Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009 - 12th International Conference, Proceedings. PART 1. ed. 2009. pp. 426-434 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1).

@inproceedings{f6db4a4a9caf49f4b27140bf0c524692,

title = "Data-derived models for segmentation with application to surgical assessment and training",

abstract = "This paper addresses automatic skill assessment in robotic minimally invasive surgery. Hidden Markov models (HMMs) are developed for individual surgical gestures (or surgemes) that comprise a typical bench-top surgical training task. It is known that such HMMs can be used to recognize and segment surgemes in previously unseen trials [1]. Here, the topology of each surgeme HMM is designed in a data-driven manner, mixing trials from multiple surgeons with varying skill levels, resulting in HMM states that model skill-specific sub-gestures. The sequence of HMM states visited while performing a surgeme are therefore indicative of the surgeon's skill level. This expectation is confirmed by the average edit distance between the state-level {"}transcripts{"} of the same surgeme performed by two surgeons with different expertise levels. Some surgemes are further shown to be more indicative of skill than others.",

author = "Balakrishnan Varadarajan and Carol Reiley and Henry Lin and Sanjeev Khudanpur and Gregory Hager",

note = "Copyright: Copyright 2010 Elsevier B.V., All rights reserved.; 12th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2009 ; Conference date: 20-09-2009 Through 24-09-2009",

year = "2009",

doi = "10.1007/978-3-642-04268-3_53",

language = "English (US)",

isbn = "3642042678",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

number = "PART 1",

pages = "426--434",

booktitle = "Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009 - 12th International Conference, Proceedings",

edition = "PART 1",

}

TY - GEN

T1 - Data-derived models for segmentation with application to surgical assessment and training

AU - Varadarajan, Balakrishnan

AU - Reiley, Carol

AU - Lin, Henry

AU - Khudanpur, Sanjeev

AU - Hager, Gregory

PY - 2009

Y1 - 2009

N2 - This paper addresses automatic skill assessment in robotic minimally invasive surgery. Hidden Markov models (HMMs) are developed for individual surgical gestures (or surgemes) that comprise a typical bench-top surgical training task. It is known that such HMMs can be used to recognize and segment surgemes in previously unseen trials [1]. Here, the topology of each surgeme HMM is designed in a data-driven manner, mixing trials from multiple surgeons with varying skill levels, resulting in HMM states that model skill-specific sub-gestures. The sequence of HMM states visited while performing a surgeme are therefore indicative of the surgeon's skill level. This expectation is confirmed by the average edit distance between the state-level "transcripts" of the same surgeme performed by two surgeons with different expertise levels. Some surgemes are further shown to be more indicative of skill than others.

AB - This paper addresses automatic skill assessment in robotic minimally invasive surgery. Hidden Markov models (HMMs) are developed for individual surgical gestures (or surgemes) that comprise a typical bench-top surgical training task. It is known that such HMMs can be used to recognize and segment surgemes in previously unseen trials [1]. Here, the topology of each surgeme HMM is designed in a data-driven manner, mixing trials from multiple surgeons with varying skill levels, resulting in HMM states that model skill-specific sub-gestures. The sequence of HMM states visited while performing a surgeme are therefore indicative of the surgeon's skill level. This expectation is confirmed by the average edit distance between the state-level "transcripts" of the same surgeme performed by two surgeons with different expertise levels. Some surgemes are further shown to be more indicative of skill than others.

UR - http://www.scopus.com/inward/record.url?scp=84862993919&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84862993919&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-04268-3_53

DO - 10.1007/978-3-642-04268-3_53

M3 - Conference contribution

C2 - 20426016

AN - SCOPUS:84862993919

SN - 3642042678

SN - 9783642042676

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 426

EP - 434

BT - Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009 - 12th International Conference, Proceedings

T2 - 12th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2009

Y2 - 20 September 2009 through 24 September 2009

ER -

Data-derived models for segmentation with application to surgical assessment and training

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this