請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/87609完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 傅立成 | zh_TW |
| dc.contributor.advisor | Li-Chen Fu | en |
| dc.contributor.author | 林祐霆 | zh_TW |
| dc.contributor.author | Yu Ting Lin | en |
| dc.date.accessioned | 2023-06-20T16:20:48Z | - |
| dc.date.available | 2023-11-09 | - |
| dc.date.copyright | 2023-06-20 | - |
| dc.date.issued | 2023 | - |
| dc.date.submitted | 2023-02-14 | - |
| dc.identifier.citation | [1] SercanÖ. Arik and Tomas Pfister. Tabnet: Attentive interpretable tabular learning. Proceedings of the AAAI Conference on Artificial Intelligence, 35(8):6679–6687, May2021.
[2] Woo Suk Hong, Adrian Daniel Haimovich, and R.Andrew Taylor. Predicting hospital admission at emergency department triage using machine learning. PLOSONE, 13(7):e0201016, July2018. [3] A. Guttmann, M.J. Schull, M.J. Vermeulen, and T.A. Stukel. Association between waiting times and short term mortality and hospital admission after departure from emergency department: population based cohort study from ontario, canada. BMJ, 342(jun011):d2983–d2983, June2011. [4] C.-J. Ng, Z.-S. Yen, J. C.-H. Tsai, L. C. Chen, S. J. Lin, Y. Y. Sang, and J.-C. Chen and. Validation of the taiwan triage and acuity scale: a new computerised five-level triage system. Emergency Medicine Journal, 28(12):1026–1031, November 2010. [5] Michael J.Bullard, Bernard Unger, Julie Spence, and Eric Grafstein and. Revisions to the canadian emergency department triage and acuity scale (CTAS) adult guidelines. CJEM, 10(02):136–142, March2008. [6] P. Tanabe. Refining emergency severity index triage criteria. Academic Emergency Medicine, 12(6):497-501, June2005. [7] Sae Won Choi, Taehoon Ko, Ki Jeong Hong, and Kyung Hwan Kim. Machine learning-based prediction of korean triage and acuity scale level in emergency department patients. Health care informatics research, 25(4):305–312, 2019. [8] Richard C. Wuerz, Leslie W. Milne, David R. Eitel, Debbie Travers, and Nicki Gilboy. Reliability and validity of a new five-level triage instrument. Academic Emergency Medicine, 7(3):236–242, March 2000. [9] Paula Tanabe, Rick Gimbel, Paul R. Yarnold, Demetrios N.Kyriacou, and James G. Adams. Reliability and validity of scores on the emergency severity index version 3. Academic Emergency Medicine, 11(1):59–65, January 2004. [10] Richard C. Wuerz, Debbie Travers, Nicki Gilboy, David R.Eitel, Alex Rosenau, and Ramine Yazhari. Implementation and refinement of the emergency severity index. Academic Emergency Medicine, 8(2):170–176, February 2001. [11] Tianqi Chen and Carlos Guestrin. Xgboost. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and DataMining, Aug 2016. [12] Woo Suk Hong, Adrian Daniel Haimovich, and R Andrew Taylor. Predicting hospital admission at emergency department triage using machine learning. PloS one, 13(7):e0201016, 2018. [13] Jamie Miles, Janette Turner, Richard Jacques, Julia Williams, and Suzanne Mason. Using machine learning risk prediction models to triage the acuity of undifferentiated patients entering the emergency care system: a systematic review. Diagnostic and prognostic research, 4(1):1–12, 2020. [14] Jeffrey Pennington, Richard Socher, and Christopher D Manning. Glove: Global vectors for word representation. In Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP), pages1532–1543, 2014. [15] Tomas Mikolov, Kai Chen, Greg Corrado, and Jeffrey Dean. Efficient estimation of word representations in vector space, 2013. [16] Matthew E. Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer. Deep contextualized word representations, 2018. [17] Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre-training of deep bidirectional transformers for language understanding, 2019. [18] Tim Salimans Alec Radford, Karthik Narasimhan and Ilya Sutskever. Improving language understanding by generative pre-training. In Technical report, OpenAI, 2018. [19] Liudmila Prokhorenkova, Gleb Gusev, Aleksandr Vorobev, Anna Veronika Dorogush, and Andrey Gulin. Catboost: unbiased boosting with categorical features. Advances in neural information processing systems, 31, 2018. [20] GuolinKe, QiMeng, Thomas Finley, Taifeng Wang, Wei Chen, Weidong Ma, Qiwei Ye, and Tie-Yan Liu. Lightgbm: A highly efficient gradient boosting decision tree. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R.Garnett, editors, Advances in Neural Information Processing Systems, volume 30. Curran Associates, Inc., 2017. [21] Sergei Popov, Stanislav Morozov, and Artem Babenko. Neural oblivious decision ensembles for deep learning on tabular data. arXiv preprint arXiv:1909.06312, 2019. [22] Xin Huang, Ashish Khetan, Milan Cvitkovic, and Zohar Karnin. Tabtransformer: Tabular data modeling using contextual embeddings. arXiv preprint arXiv:2012.06678, 2020. [23] Oleksandr Ivanov, Lisa Wolf, Deena Brecher, Erica Lewis, Kevin Masek, Kyla Montgomery, Yurii Andrieiev, Moss McLaughlin, Stephen Liu, Robert Dunne, et al. Improving ed emergency severity index acuity assignment using machine learning and clinical natural language processing. Journal of Emergency Nursing, 47(2):265–278, 2021. [24] Yecheng Liu, Jiandong Gao, Jihai Liu, Joseph Harold Walline, Xiaoying Liu, Ting Zhang, Yunyang Wu, Ji Wu, Huadong Zhu, and Weiguo Zhu. Development and validation of a practical machine-learning triage algorithm for the detection of patients in need of critical care in the emergency department. Scientific reports, 11(1):1–9, 2021. [25] YannN. Dauphin, Angela Fan, Michael Auli, and David Grangier. Language modeling with gated convolutional networks, 2017. [26] André F. T.Martins and Ramón Fernandez Astudillo. From softmax to sparsemax: A sparse model of attention and multi-label classification, 2016. [27] Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. Attention is all you need, 2017. [28] Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. Neural machine translation by jointly learning to align and translate, 2016. [29] Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, June 2016. [30] Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E. Hinton. Layer normalization, 2016. [31] Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner. Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11):2278–2324, 1998. [32] Yves Grandvalet and Yoshua Bengio. Semi-supervised learning by entropy minimization. Advances in neural information processing systems, 17, 2004. [33] Leo Breiman. Random forests. Machine learning, 45(1):5–32, 2001. [34] Marti A. Hearst, Susan T Dumais, Edgar Osuna, John Platt, and Bernhard Scholkopf. Support vector machines. IEEE Intelligent Systems and their applications, 13(4):18–28, 1998. [35] Yoshihiko Raita, Tadahiro Goto, Mohammad Kamal Faridi, David F. M. Brown, Carlos A. Camargo, and Kohei Hasegawa. Emergency department triage prediction of clinical outcomes using machine learning models. Critical Care, 23(1), February 2019. [36] Li-Hung Yao, Ka-Chun Leung, Jheng-Huang Hong, Chu-Lin Tsai, and Li-Chen Fu. A system for predicting hospital admission at emergency department based on electronic health record using convolution neural network. In 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC). IEEE, October 2020. [37] Ka-Chun Leung, Yu-Ting Lin, De-Yang Hong, Chu-Lin Tsai, and Li-Chen Fu. A novel interpretable deep learning-based system for triage prediction in the emergency department: A prospective study. In 2021 IEEE International Conference on Systems, Man, and Cybernetics (SMC). IEEE, October 2021. [38] Jesse Vig. Bertviz: A tool for visualizing multihead self-attention in the bert model. In ICLR Workshop: Debugging Machine Learning Models, 2019. [39] Ganesh Jawahar, Benoît Sagot, and Djamé Seddah. What does bert learn about the structure of language? In ACL 2019-57th Annual Meeting of the Association for Computational Linguistics, 2019. | - |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/87609 | - |
| dc.description.abstract | 急診檢傷站是所有病患來到急診的第一站，此站設立的目的是希望透過對病患基本的生命徵象的量測與對人對自身病狀的描述，由護理人員評估狀況，並對病人進行分級，在給予最合適的醫療救護的同時，也能有效地運用有效的醫療資源。儘管目前已經有許多國家都有自己的一套急診檢傷系統，但是卻仍然面對很多限制。舉例來說， 在給予急診分級時，分級的結果會十分仰賴診療的護理人員的個人經驗與知識，在大多時候的狀況，常常會有錯誤分級的情況發生。這樣的情況不僅會對有緊急狀況的病人造成不好的影響，同時也可能增加整個醫療系統的負擔。這樣的問題在台灣也不例外，儘管台灣有TTAS檢傷分級系統，但是它並沒有足夠的能力對急診室的病人去進行分級，絕大多數的病患會被分至三級。
為了要從根本上解決這個問題，我們的目標是希望能夠透過 AI 去訓練一個機器學習模型能夠有效地分析病人的病狀並進行分流。即使已經有許多機器學習的研究致力於解決急診室分流不佳造成擁塞的問題，但是大部分的研究僅僅使用最傳統的演算法，且缺乏對文字資訊處理的能力，因此在此實驗中，我們嘗試搭建一個多模態的模型，能夠同時處理病人生命徵象與文字資訊進行分析預測，期待透過多方面的資訊整合獲得一個具有分辨重症病患能力的系統。 | zh_TW |
| dc.description.abstract | Triage is the process of accurately assessing the patient's symptoms and providing them with proper clinical treatments in the emergency department (ED). While many countries have developed their triage process to stratify patient’s clinical severity and thus distribute medical resources, there is still some limitation of the current triage process. Since triage level is mainly performed by the experienced nurse based on the mix of subjective and objective criteria, mis-triage often happens in the ED. It can not only cause adverse effects on patients but also impose an undue burden on the healthcare delivery system. In Taiwan, although the five-level triage system, Taiwan Triage and Acuity Scale (TTAS), is adopted in the ED to help nurses to distinguish patients’ emergent conditions, it cannot prioritize patients appropriately. Most of the patients are classified as level 3. In order to address the problem, our study aims to design a deep learning prediction system that can help the physician to prioritize patients to the appropriate triage level. To our knowledge, most of the existing models are developed by traditional machine learning methods and only utilize the patient’s demographics and vital signs information. Even though the chief complaint is included in the study, it is usually treated as a categorical feature instead of a text-based feature. As a result, to grasp rich information from multi-modal features, we proposed the multi-modality model that can handle heterogeneous data by using a representation learning approach. | en |
| dc.description.provenance | Submitted by admin ntu (admin@lib.ntu.edu.tw) on 2023-06-20T16:20:48Z No. of bitstreams: 0 | en |
| dc.description.provenance | Made available in DSpace on 2023-06-20T16:20:48Z (GMT). No. of bitstreams: 0 | en |
| dc.description.tableofcontents | 中文摘要: i
Abstract: iii Contents: v List of figures: ix List of table: xi Chapter1 Introduction 1 1.1 Background 1 1.2 Motivation 2 1.3 Challenges 4 1.3.1 Utilizing multi-modality features efficiently is challenging 5 1.3.2 Explaining the prediction result of deep learning is difficult 5 1.3.3 High dependency on TTAS cannot prioritize patients appropriately 6 1.4 Related Works 6 1.4.1 Contextualized word embedding 7 1.4.2 Deep learning for tabular data 8 1.4.3 Current works in triage system 9 1.5 Objectives 10 1.5.1 To develop a model that can deal with multi-modality features 11 1.5.2 To construct a deep learning model with interpretability 11 1.5.3 To build a system that can imitate the process of physician diagnosis 12 1.6 Thesis Organization 13 Chapter2 Preliminaries 14 2.1 Imbalanced Data Distribution 14 2.1.1 Oversamlping 14 2.1.2 Undersampling 15 2.2 Multi-TaskLearning 16 2.2.1 Learning a shared representation 16 2.2.2 Take-weighting in loss function 17 2.3 TabNet 18 2.3.1 Gated Linear Unit 18 2.3.2 Attentive Transformer 18 2.3.3 Feature Transformer 19 2.3.4 The Overall Model Achitecture 20 2.3.5 Interpretability 20 2.3.6 Unsupervised Pre-training 21 2.4 Transformer 23 2.4.1 Self-Attention Mechanism 23 2.4.2 Multi-head Attention 25 2.4.3 Positional Encoding 25 2.4.4 The Overall Model Architecture 26 2.4.4.1 Encoder 27 2.4.4.2 Decoder 27 2.5 Bidirectional Encoder Representations from Transformers 28 2.5.1 Masked Language Modeling 29 2.5.2 Next Sentence Prediction 30 2.6 Convolutional Neural Network 32 2.6.1 Convolutional Layer 33 2.6.2 Pooling Layer 34 2.6.3 Fully connected Layer 35 Chapter3 Methodology: 36 3.1 System Overview 37 3.2 Data Preparation 38 3.2.1 The NTUH Retrospective Data 38 3.2.2 The final retrospective dataset enrolled in our program 38 3.2.3 The NTUH Prospective Dataset 39 3.2.4 Structural data and Text data in NTUH Pospective dataset 40 3.2.5 Data Augmentation 44 3.3 Pre-training Stage 45 3.3.1 Pre-train the Encoders 45 3.3.1.1 Vital Sign Encoder 46 3.3.1.2 MacBERT Encoder 49 3.4 Fine-tuning Stage 49 3.4.1 The Overall Model Architecture 50 3.4.2 The Input 50 3.4.3 The Encoders 51 3.4.3.1 Pre-trained Vital sign encoder 52 3.4.3.2 Pre-trained Language Model Encoder 52 3.4.4 The Classifiers 53 3.4.5 The Output 54 3.5 Loss Function 56 Chapter4 System Evaluation and Application 58 4.1 Training Settings 58 4.2 Evaluation Metrics 59 4.3 Data Characteristics 60 4.4 Experiments 62 4.4.1 Compared with other machine learning methods 62 4.4.1.1 Triage Level Prediction 62 4.4.1.2 Hospitalization Prediction 62 4.4.1.3 Length of Stay Prediction 63 4.4.2 Ablation Studies 64 4.4.2.1 Effectiveness of multi-modality 64 4.4.2.2 Effectiveness of multi-tasks training and data augmentation 65 4.4.2.3 Effectiveness of different language model 66 4.4.2.4 Effectiveness of different fusion method 66 4.4.2.5 Interpretability 67 4.5 System Application 68 Chapter5 Conlcusion 72 5.1 Summary 72 5.2 FutureWork 73 Reference: 74 | - |
| dc.language.iso | en | - |
| dc.title | 利用深度學習分析檢傷、住院與暫留時間之重症病患偵測系統 | zh_TW |
| dc.title | Interpretable Deep Learning System for Identifying Critical Patients by Predicting Triage Level, Hospitalization and Length of Stay: A Prospective Study | en |
| dc.type | Thesis | - |
| dc.date.schoolyear | 111-1 | - |
| dc.description.degree | 碩士 | - |
| dc.contributor.oralexamcommittee | 黃建華;蔡居霖;張智星;陳緼儂 | zh_TW |
| dc.contributor.oralexamcommittee | Jheng-Huang Hong;Chu-Lin Tsai;Jyh-Shing Jang;Yun-Nung Chen | en |
| dc.subject.keyword | 急診室,檢傷分級預測,住院預測,暫留時間預測,多模型整合,注意力網絡, | zh_TW |
| dc.subject.keyword | Emergency Department,Triage System,Hospital Admission,Length of Stay,Multi-modal integration,Transformer, | en |
| dc.relation.page | 79 | - |
| dc.identifier.doi | 10.6342/NTU202300277 | - |
| dc.rights.note | 未授權 | - |
| dc.date.accepted | 2023-02-15 | - |
| dc.contributor.author-college | 電機資訊學院 | - |
| dc.contributor.author-dept | 資訊網路與多媒體研究所 | - |
| 顯示於系所單位： | 資訊網路與多媒體研究所 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| ntu-111-1.pdf 目前未授權公開取用 | 4.22 MB | Adobe PDF |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。