PMID- 37801923
OWN - NLM
STAT- Publisher
LR  - 20231006
IS  - 1879-0534 (Electronic)
IS  - 0010-4825 (Linking)
VI  - 166
DP  - 2023 Sep 29
TI  - A multi-stage transfer learning strategy for diagnosing a class of rare laryngeal 
      movement disorders.
PG  - 107534
LID - S0010-4825(23)00999-X [pii]
LID - 10.1016/j.compbiomed.2023.107534 [doi]
AB  - BACKGROUND: It remains hard to directly apply deep learning-based methods to 
      assist diagnosing essential tremor of voice (ETV) and abductor and adductor 
      spasmodic dysphonia (ABSD and ADSD). One of the main challenges is that, as a 
      class of rare laryngeal movement disorders (LMDs), there are limited available 
      databases to be investigated. Another worthy explored research question is which 
      above sub-disorder benefits most from diagnosis based on sustained phonations. 
      The question is from the fact that sustained phonations can help detect 
      pathological voice from healthy voice. METHOD: A transfer learning strategy is 
      developed for LMD diagnosis with limited data, which consists of three 
      fundamental parts. (1) An extra vocally healthy database from the International 
      Dialects of English Archive (IDEA) is employed to pre-train a convolutional 
      autoencoder. (2) The transferred proportion of the pre-trained encoder is 
      explored. And its impact on LMD diagnosis is also evaluated, yielding a two-stage 
      transfer model. (3) A third stage is designed following the initial two stages to 
      embed information of pathological sustained phonation into the model. This stage 
      verifies the different effects of applying sustained phonation on diagnosing the 
      three sub-disorders, and helps boost the final diagnostic performance. RESULTS: 
      The analysis in this study is based on clinician-labeled LMD data obtained from 
      the Vanderbilt University Medical Center (VUMC). We find that diagnosing ETV 
      shows sensitivity to sustained phonation within the current database. Meanwhile, 
      the results show that the proposed multi-stage transfer learning strategy can 
      produce (1) accuracy of 65.3% on classifying normal and other three sub-disorders 
      all at once, (2) accuracy of 85.3% in differentiating normal, ABSD, and ETV, and 
      (3) accuracy of 77.7% for normal, ADSD and ETV. These findings demonstrate the 
      effectiveness of the proposed approach.
CI  - Copyright (c) 2023 Elsevier Ltd. All rights reserved.
FAU - Yao, Yu
AU  - Yao Y
AD  - The College of Information Science and Engineering, Northeastern University, 
      Wenhua Road 3-11, Shenyang, 110819, Liaoning, PR China; The Department of 
      Electrical Engineering and Computer Science, Vanderbilt University, 2301 
      Vanderbilt Place, Nashville, 37235, TN, USA. Electronic address: 
      yu.ethan.yao@gmail.com.
FAU - Powell, Maria
AU  - Powell M
AD  - Vanderbilt University Medical Center, 1211 Medical Center Drive, Nashville, 
      37232, TN, USA.
FAU - White, Jules
AU  - White J
AD  - The Department of Electrical Engineering and Computer Science, Vanderbilt 
      University, 2301 Vanderbilt Place, Nashville, 37235, TN, USA.
FAU - Feng, Jian
AU  - Feng J
AD  - The College of Information Science and Engineering, Northeastern University, 
      Wenhua Road 3-11, Shenyang, 110819, Liaoning, PR China.
FAU - Fu, Quchen
AU  - Fu Q
AD  - The Department of Electrical Engineering and Computer Science, Vanderbilt 
      University, 2301 Vanderbilt Place, Nashville, 37235, TN, USA.
FAU - Zhang, Peng
AU  - Zhang P
AD  - The Department of Electrical Engineering and Computer Science, Vanderbilt 
      University, 2301 Vanderbilt Place, Nashville, 37235, TN, USA.
FAU - Schmidt, Douglas C
AU  - Schmidt DC
AD  - The Department of Electrical Engineering and Computer Science, Vanderbilt 
      University, 2301 Vanderbilt Place, Nashville, 37235, TN, USA.
LA  - eng
PT  - Journal Article
DEP - 20230929
PL  - United States
TA  - Comput Biol Med
JT  - Computers in biology and medicine
JID - 1250250
SB  - IM
OTO - NOTNLM
OT  - Convolutional autoencoder
OT  - Laryngeal movement disorders
OT  - Limited data
OT  - Multi-stage transfer
OT  - Sustained phonation
COIS- Declaration of competing interest The authors declare that they have no known 
      competing financial interests or personal relationships that could have appeared 
      to influence the work reported in this paper.
EDAT- 2023/10/07 00:42
MHDA- 2023/10/07 00:42
CRDT- 2023/10/06 18:08
PHST- 2023/06/05 00:00 [received]
PHST- 2023/08/28 00:00 [revised]
PHST- 2023/09/27 00:00 [accepted]
PHST- 2023/10/07 00:42 [medline]
PHST- 2023/10/07 00:42 [pubmed]
PHST- 2023/10/06 18:08 [entrez]
AID - S0010-4825(23)00999-X [pii]
AID - 10.1016/j.compbiomed.2023.107534 [doi]
PST - aheadofprint
SO  - Comput Biol Med. 2023 Sep 29;166:107534. doi: 10.1016/j.compbiomed.2023.107534.