Reinforcement Learning Based Vocal Fold Localization in Preoperative Neck CT for Injection Laryngoplasty

被引:1
作者
Al, Walid Abdullah [1 ]
Cha, Wonjae [2 ]
Yun, Il Dong [1 ]
机构
[1] Hankuk Univ Foreign Studies, Div Comp Engn, Yongin 17035, South Korea
[2] Seoul Natl Univ, Bundang Hosp, Coll Med, Dept Otorhinolaryngol Head & Neck Surg, Seongnam 13620, South Korea
来源
APPLIED SCIENCES-BASEL | 2023年 / 13卷 / 01期
基金
新加坡国家研究基金会;
关键词
injection laryngoplasty; neck CT; vocal fold localization; deep learning; reinforcement learning; mirror environment;
D O I
10.3390/app13010262
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
Transcutaneous injection laryngoplasty is a well-known procedure for treating a paralyzed vocal fold by injecting augmentation material to it. Hence, vocal fold localization plays a vital role in the preoperative planning, as the fold location is required to determine the optimal injection route. In this communication, we propose a mirror environment based reinforcement learning (RL) algorithm for localizing the right and left vocal folds in preoperative neck CT. RL-based methods commonly showed noteworthy outcomes in general anatomic landmark localization problems in recent years. However, such methods suggest training individual agents for localizing each fold, although the right and left vocal folds are located in close proximity and have high feature-similarity. Utilizing the lateral symmetry between the right and left vocal folds, the proposed mirror environment allows for a single agent for localizing both folds by treating the left fold as a flipped version of the right fold. Thus, localization of both folds can be trained using a single training session that utilizes the inter-fold correlation and avoids redundant feature learning. Experiments with 120 CT volumes showed improved localization performance and training efficiency of the proposed method compared with the standard RL method.
引用
收藏
页数:9
相关论文
共 14 条
  • [1] A Study of incidence and etiopathology of vocal cord paralysis
    Shafkat Ahmad
    Ajaz Muzamil
    Mohd Lateef
    [J]. Indian Journal of Otolaryngology and Head and Neck Surgery, 2002, 54 (4) : 294 - 296
  • [2] Partial Policy-Based Reinforcement Learning for Anatomical Landmark Localization in 3D Medical Images
    Al, Walid Abdullah
    Yun, Il Dong
    [J]. IEEE TRANSACTIONS ON MEDICAL IMAGING, 2020, 39 (04) : 1245 - 1255
  • [3] Evaluating reinforcement learning agents for anatomical landmark detection
    Alansary, Amir
    Oktay, Ozan
    Li, Yuanwei
    Le Folgoc, Loic
    Hou, Benjamin
    Vaillant, Ghislain
    Kamnitsas, Konstantinos
    Vlontzos, Athanasios
    Glocker, Ben
    Kainz, Bernhard
    Rueckert, Daniel
    [J]. MEDICAL IMAGE ANALYSIS, 2019, 53 : 156 - 164
  • [4] Percutaneous Injection Laryngoplasty
    Chhetri, Dinesh K.
    Jamal, Nausheen
    [J]. LARYNGOSCOPE, 2014, 124 (03) : 742 - 745
  • [5] Ghesu FC, 2016, International conference on medical image computing and computer-assisted intervention, P229
  • [6] Towards intelligent robust detection of anatomical structures in incomplete volumetric data
    Ghesu, Florin C.
    Georgescu, Bogdan
    Grbic, Sasa
    Maier, Andreas
    Hornegger, Joachim
    Comaniciu, Dorin
    [J]. MEDICAL IMAGE ANALYSIS, 2018, 48 : 203 - 213
  • [7] The 3D Printing of the Paralyzed Vocal Fold: Added Value in Injection Laryngoplasty
    Hamdan, Abdul-Latif
    Haddad, Ghassan
    Haydar, Ali
    Hamade, Ramsey
    [J]. JOURNAL OF VOICE, 2018, 32 (04) : 499 - 501
  • [8] An Open-Source Three-Dimensionally Printed Laryngeal Model for Injection Laryngoplasty Training
    Lee, Mark
    Ang, Chelston
    Andreadis, Katerina
    Shin, James
    Rameau, Anais
    [J]. LARYNGOSCOPE, 2021, 131 (03) : E890 - E895
  • [9] LJPvd M., 2008, J Mach Learn Res, V9, P9
  • [10] A Deep Regression Architecture with Two-Stage Re-initialization for High Performance Facial Landmark Detection
    Lv, Jiangjing
    Shao, Xiaohu
    Xing, Junliang
    Cheng, Cheng
    Zhou, Xi
    [J]. 30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017), 2017, : 3691 - 3700