TY - GEN
T1 - Development of state transition model and speech recognition module for training of neurological examination
AU - Sugamiya, Y.
AU - Matsunaga, K.
AU - Wang, C.
AU - Tokunaga, S.
AU - Kawasaki, S.
AU - Ishii, H.
AU - Nakae, Y.
AU - Katayama, T.
AU - Takanishi, A.
N1 - Publisher Copyright:
© 2014 IEEE.
PY - 2014/4/20
Y1 - 2014/4/20
N2 - This study aims to develop a state transition model and speech recognition module for application to a whole-body patient simulator for scenario based training (SBT) of neurological examination procedures. These procedures are very important for the early identification of neurological system disorders. In neurological examinations, the doctor selects procedures by situation of patients, interacts with the patient and performs a series of medical procedures to judge the site of nerve disorders and lesions. SBT is one type of training that doctor performs to be based on scenario. Scenario is patient situation and examination procedure. SBT is used for the training of such examinations. SBT is often performed using simulated patients (SPs); however, the number of SPs is not sufficient and SPs cannot adequately reproduce diseases. Therefore, we integrated a state transition model with a commercial speech recognition software and developed a dialogue system for SBT. First, we customized the grammatical rules and the different words that the software can recognize. By doing so, the recognition rate improved to about 90%. Second, we developed a probability model for the state transition model. In neurological examinations, patient pause is limited and an order of procedures is generally decided. We developed a state transition model including probability model with the customized speech recognition module.
AB - This study aims to develop a state transition model and speech recognition module for application to a whole-body patient simulator for scenario based training (SBT) of neurological examination procedures. These procedures are very important for the early identification of neurological system disorders. In neurological examinations, the doctor selects procedures by situation of patients, interacts with the patient and performs a series of medical procedures to judge the site of nerve disorders and lesions. SBT is one type of training that doctor performs to be based on scenario. Scenario is patient situation and examination procedure. SBT is used for the training of such examinations. SBT is often performed using simulated patients (SPs); however, the number of SPs is not sufficient and SPs cannot adequately reproduce diseases. Therefore, we integrated a state transition model with a commercial speech recognition software and developed a dialogue system for SBT. First, we customized the grammatical rules and the different words that the software can recognize. By doing so, the recognition rate improved to about 90%. Second, we developed a probability model for the state transition model. In neurological examinations, patient pause is limited and an order of procedures is generally decided. We developed a state transition model including probability model with the customized speech recognition module.
UR - http://www.scopus.com/inward/record.url?scp=84949927861&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84949927861&partnerID=8YFLogxK
U2 - 10.1109/ROBIO.2014.7090483
DO - 10.1109/ROBIO.2014.7090483
M3 - Conference contribution
AN - SCOPUS:84949927861
T3 - 2014 IEEE International Conference on Robotics and Biomimetics, IEEE ROBIO 2014
SP - 1124
EP - 1129
BT - 2014 IEEE International Conference on Robotics and Biomimetics, IEEE ROBIO 2014
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2014 IEEE International Conference on Robotics and Biomimetics, IEEE ROBIO 2014
Y2 - 5 December 2014 through 10 December 2014
ER -