In this paper, we present two methods to utilize Latent Dirichlet Allocation and Hidden Markov Model to automatically detect surgical workflow phases based on codebook which is built by quantizing the extracted optical flow vectors from the recorded videos of surgical processes. To detect the current phase at a given time point of an operation, some recorded training data with correct phase label need to be learned by a topic model. All documents which are actually short clips divided from the recorded videos are presented as mixtures over learned latent topics. These presentations are then quantized as observed values of a Hidden Markov Model (HMM). The major difference between two proposed methods is that while the first method quantizes all topic-based presentations based on k-means, the second method does this based on multivariate Gaussian mixture model. A Left to Right HMM is appropriate for this work because there is no switching the order between surgical phases.
CITATION STYLE
Tran, D. T., Sakurai, R., & Lee, J. H. (2016). An improvement of surgical phase detection using latent dirichlet allocation and hidden markov model. In Smart Innovation, Systems and Technologies (Vol. 45, pp. 249–261). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-319-23024-5_23
Mendeley helps you to discover research relevant for your work.