 |
|
|
|
以作者查詢圖書館館藏 、以作者查詢臺灣博碩士 、以作者查詢全國書目 、勘誤回報 、線上人數：69 、訪客IP：3.12.34.96
姓名 江郅錡(Chih-Chi Chiang) 查詢紙本館藏 畢業系所 數學系 論文名稱 機器學習在中醫舌下脈絡分類的應用
(Application of Machine Learning in the Classification of Sublingual Vein in Traditional Chinese Medicine)相關論文 檔案 [Endnote RIS 格式]
[相關文章]
[文章引用]
[完整記錄]
[館藏目錄]
[檢視]
- 本電子論文使用權限為同意立即開放。
- 已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
- 請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。
摘要(中) 舌診為中醫辨證論治的重要指標，是一種簡便，非侵入性之檢查。舌下脈絡(Sublingualvein, SV) 為中醫判斷瘀阻與血瘀證的象徵。目前已經有許多研究顯示SV 瘀阻程度與其疾病嚴重程度正相關，但是SV 的診斷常常因為不同醫師間一些的經驗、色彩感知、心態等主觀因素，造成不同的判讀結果。本研究的目標希望發展機器學習為基礎的電腦輔助系統，協助診斷病人的舌下脈絡瘀阻程度。我們考慮二元分類問題，亦即輕微和嚴重兩級。我們測試許多監督式機器學習模型，包括支援向量機(SVM)、K 鄰近法、決策樹、RidgeClassifier 等等，最後篩選出13 個模型進行實驗。為了提
升準確率與節省訓練時間，本研究使用了兩種提取特徵的技術，一是利用主成分分析(Principle Component Analysis, PCA) 結合逆向切片迴歸(Sliced Inverse Regression,SIR) 進行提取，或是利用卷積神經網路(Convolutional Neural Network, CNN）提取。經過數值實驗發現，使用原始照片且灰階影像時，準確率平均後只有60.5% 左右，在去除雜訊，並且使用紅綠藍彩色模式三通道都保留後，機器學習的平均準確率可以
到達81%，而13 個模型之中，最好的模型SVM_linear 準確率可以到85.5%。另外，PCA+SIR 降維方法，可以在準確率維持的情況下，使得訓練時間只需原本的1/35 左右，達到節省運算成本的效果，最後，我們也使用的卷積神經網路，進行特徵提取，13 個模型中的RidgeClassifier 準確率更是可以到達87.5%。最後，再配合混淆矩陣及接收者操作特徵曲線（receiver operating characteristic curve， ROC 曲線），我們就可以提供醫生，科學化的分辨舌下脈絡嚴重度的輔助工具。摘要(英) Tongue diagnosis is an important indicator of traditional Chinese medicine (TCM) syndrome differentiation and treatment. It is a simple and non-invasive examination. Sublingual vein (SV) is a symbol of TCM to judge sublingual blood stasis . There have been many theses showing that the degree of SV stasis is positively correlated with the severity of the disease. However, the diagnosis of SV is often due to subjective factors such as experience,color perception, and mentality among different physicians, resulting in different interpretation results. The goal of this research is to develop a computer-assisted system
based on machine learning to assist in diagnosing the degree of SV stasis in patients. We consider the problem of binary classification, mild and severe. We test many supervised
machine learning models, including support vector machines (SVM), K-nearest neighbor,decision trees, RidgeClassifier, etc. In order to improve accuracy and save training time,
this research uses two techniques to extract features. One is to use Principle Component Analysis (PCA) combined with Sliced Inverse Regression (SIR) for extraction. Another
way is using Convolutional Neural Network (CNN) for extraction. Experiments results have found that when using original photos and grayscale images, the average accuracy
is only about 60.5%. After removing the noise, and the red, green, and blue three channels are retained, the average accuracy of machine learning can reach 81%.Among the
thirteen models, the best model SVM_linear has an accuracy of 85.5%. In addition, we using PCA+SIR, with the accuracy maintained, the training time is only about 1/35 of
the original, achieving the goal of saving computing costs. Finally, We also use the CNN for feature extraction, and RidgeClassifier of the thirteen models has highest accuracy
87.5%. All of the above, with the combination of the confusion matrix and receiver operating characteristic curve (ROC curve), we can provide doctors a scientific system to
diagnose the severity of SV stasis .關鍵字(中) ★ 舌下脈絡
★ 主成分分析+逆向切片回歸
★ 機器學習關鍵字(英) ★ Sublingual vein
★ PCA+SIR
★ Machine learning論文目次 致謝. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
1 緒論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 方法. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1 機器學習發展與種類. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 機器學習演算法. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 Support Vector Machine (SVM) family . . . . . . . . . . . . . . . . 6
2.2.2 Linear model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.3 Nearest Neighbors family . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.4 Decision Tree family . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.5 其他的演算法及演算法的選擇. . . . . . . . . . . . . . . . . . . . . 11
2.3 Convolutional Neural Network(CNN) . . . . . . . . . . . . . . . . . . . . . 12
2.3.1 卷積神經網路介紹. . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3.2 卷積層. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.3.3 池化層. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3.4 平坦層. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3.5 全連接層. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4 特徵提取. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4.1 統計值和比例. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4.2 降維. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4.3 使用CNN 進行特徵提取. . . . . . . . . . . . . . . . . . . . . . . 20
3 實驗及討論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1 資料來源. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2 實驗流程. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.3 照片的預處理. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3.1 感興趣的區域. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3.2 照片的標準化. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.3.3 照片轉換成機器學習的資料. . . . . . . . . . . . . . . . . . . . . . 26
3.4 不同預處理的比較. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.5 模型的選擇. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.6 不同特徵提取方式的比較. . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.6.1 降維. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.6.2 CNN 提取特徵的比較. . . . . . . . . . . . . . . . . . . . . . . . . 31
3.7 ROC 曲線. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.8 卷積神經網路訓練結果. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4 結論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42參考文獻 [1] L.-C. Lo, Y.-F. Chen, W.-J. Chen, T.-L. Cheng, and J.Y. Chiang. The study on
the agreement between automatic tongue diagnosis system and traditional Chinese
medicine practitioners. Evid. Based Complementary Altern. Med., 2012:9 pages, 2012.
[2] Y.-S. Huang, M.-C. Sun, P.-C. Hsu, Y.-L. Chen, J.Y. Chiang, and L.-C. Lo. The
relationship between ischemic stroke patients with and without retroflex tongue: a
retrospective study. Evid. Based Complementary Altern. Med., 2017:7 pages, 2017.
[3] L.-C. Lo, T.-L. Cheng, Y.-J. Chen, S. Natsagdorj, and J.Y. Chiang. TCM tongue
diagnosis index of early-stage breast cancer. Complement. Ther. Med., 23:705–713,
2015.
[4] L.-C. Lo, T.-L. Cheng, J.Y. Chiang, and N. Damdinsuren. Breast cancer index: a
perspective on tongue diagnosis in traditional Chinese medicine. J. Tradit. Complement.
Med., 3:194–203, 2013.
[5] T.-C. Wu, K.-L. Wu, W.-L. Hu, J.-M. Sheen, C.-N. Lu, J.Y. Chiang, and Y.-C.
Hung. Tongue diagnosis indices for upper gastrointestinal disorders: Protocol for a
cross-sectional, case-controlled observational study. Medicine, 97:e9607, 2018.
[6] T.-C. Wu, C.-N. Lu, W.-L. Hu, K.-L. Wu, J.Y. Chiang, J.-M. Sheen, and Y.-C. Hung.
Tongue diagnosis indices for gastroesophageal reflux disease: A cross-sectional, casecontrolled
observational study. Medicine, 99:e20471, 2020.
[7] P.-C. Hsu, Y.-C. Huang, J.Y. Chiang, H.-H. Chang, P.-Y. Liao, and L.-C. Lo. The
association between arterial stiffness and tongue manifestations of blood stasis in
patients with type 2 diabetes. BMC Complement Altern. Med., 16:1–7, 2016.
[8] P.-C. Hsu, H.-K. Wu, Y.-C. Huang, H.-H. Chang, T.-C. Lee, Y.-P. Chen, J.Y. Chiang,
and L.-C. Lo. The tongue features associated with type 2 diabetes mellitus. Medicine,
98, 2019.
[9] L.-C. Lo, C.H. Chen, J.Y. Chiang, T.-L. Cheng, H.-J. Lin, and H.-H. Chang. Tongue
diagnosis of traditional Chinese medicine for rheumatoid arthritis. Afr. J. Tradit.
Complement. Altern. Med., 10:360–369, 2013.
[10] Y.-C. Hung and W.-L. Hu. Tongue inspection and diagnosis: Past, present, and
future. Jpn J Med, 3:2, 2020.
[11] L. Hedström, M. Albrektsson, and H. Bergh. Is there a connection between sublingual
varices and hypertension? BMC Oral Health, 15:1–6, 2015.
[12] B. N. Vijay Raghawa Rao. Clinical Examinations in Cardiology-E-Book. Elsevier
Health Sciences, 2007.
[13] Q. Chen, Z. Xu, and Y. Chai. General situation on the study of inspection of
sublingual veins. J Tradit Chin Med, 24:133–136, 2004.
[14] N. Akkaya, D. Ölmez, and G. Özkan. Evaluation of the factors associated with
sublingual varices: a descriptive clinical study. Folia Morphol, 78:325–330, 2019.
[15] N.T. Duarte, A. de Oliveira Godoy, J. da Rocha Tenório, N.S. Andrade, J.B. Franco,
M. Pérez-Sayáns, and K.L. Ortega. Prevalence of sublingual varices in patients with
cirrhosis and the correlation with nitrogen compounds. J. oral med., oral surg., oral
pathol. oral radiol. (Online), 129:39–44, 2020.
[16] 符小玉and 孫克偉. 肝纖維化分期與舌下脈絡積分的關係. 中西醫结合肝病雜誌,
18:12–14, 2008.
[17] W.-Z. Deng, Q.-B. Lang, and C.-Q. Ling. Relationship between abnormal characteristics
of sublingual collateral and portal vein hemodynamic changes in patients with
primary hepatic carcinoma. Chin J Integr Med, 14:28–32, 2008.
[18] J Zhao, L.Y. Guo, J.M. Yang, and J.W. Jia. Sublingual vein parameters, AFP,
AFP-L3, and GP73 in patients with hepatocellular carcinoma. Genet. Mol. Res.,
14:7062–7067, 2015.
[19] Z. Yan, M. Yu, K. Wang, and N. Li. Sublingual vein segmentation from near infrared
sublingual images. Comput. Aided Des., 20:1569–1574, 2008.
[20] L.-Q. Zhou, J.-Y. Wang, S.-Y. Yu, G.-G. Wu, Q. Wei, Y.-B. Deng, X.-L. Wu, X.-W.
Cui, and C.F. Dietrich. Artificial intelligence in medical imaging of the liver. World
J. Gastroenterol., 25:672, 2019.
[21] S.M. McKinney, M. Sieniek, V. Godbole, J. Godwin, N. Antropova, H. Ashrafian,
T. Back, M. Chesus, G.S. Corrado, A. Darzi, et al. International evaluation of an AI
system for breast cancer screening. Nature, 577:89–94, 2020.
[22] Q. Li and Z. Liu. Tongue color analysis and discrimination based on hyperspectral
images. Comput Med Imaging Graph, 33:217–221, 2009.
[23] Q. Li, Y. Wang, H. Liu, Z. Sun, and Z. Liu. Tongue fissure extraction and classification
using hyperspectral imaging technology. Appl Opt, 49:2006–2013, 2010.
[24] Q. Li, Y. Wang, H. Liu, Y. Guan, and L. Xu. Sublingual vein extraction algorithm
based on hyperspectral tongue imaging technology. Comput Med Imaging Graph,
35:179–185, 2011.
[25] Y. Tang, Y. Sun, J.Y. Chiang, and X. Li. Research on multiple-instance learning for
tongue coating classification. IEEE Access, 9:66361–66370, 2021.
[26] L. Zhi, D. Zhang, J.-Q. Yan, Q.-L. Li, and Q.-L. Tang. Classification of hyperspectral
medical tongue images for tongue diagnosis. Comput Med Imaging Graph, 31:672–
678, 2007.
[27] J. Hou, H.-Y. Su, B. Yan, H. Zheng, Z.-L. Sun, and X.-C. Cai. Classification of
tongue color based on CNN. In 2017 IEEE 2nd International Conference on Big
Data Analysis (ICBDA), pages 725–729. IEEE, 2017.
[28] Y. Jia, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. Girshick, S. Guadarrama,
and T. Darrell. Caffe: Convolutional architecture for fast feature embedding. In
Proceedings of the 22nd ACM international conference on Multimedia, pages 675–
678, 2014.
[29] X. Li, Y. Tang, and Y. Sun. Tongue coating classification based on multiple-instance
learning and deep features. In International Conference on Neural Information Processing,
pages 504–511. Springer, 2019.
[30] B. Schölkopf, A.J. Smola, R.C. Williamson, and P.L. Bartlett. New support vector
algorithms. Neural Comput., 12:1207–1245, 2000.
[31] M. Takeichi and T. Sato. Computerized color analysis of” xue yu”(blood stasis) in
the sublingual vein using a new technology. Am. J. Chin. Med., 25:213–219, 1997.
[32] K.-C. Li. Sliced inverse regression for dimension reduction. J. Amer. Statist. Assoc.,
86:316–327, 1991.
[33] S. Wold, K. Esbensen, and P. Geladi. Principal component analysis. Chemom. Intell.
Lab. Syst., 2:37–52, 1987.
[34] Y. Tu, Y. S. Hung, Li Hu, and Z. Zhang. PCA-SIR: A new nonlinear supervised
dimension reduction method with application to pain prediction from EEG. In 2015
7th International IEEE/EMBS Conference on Neural Engineering (NER), pages
1004–1007. IEEE, 2015.指導教授 黃楓南(Feng-Nan Hwang) 審核日期 2021-9-2 推文 plurk
funp
live
udn
HD
myshare
netvibes
friend
youpush
delicious
baidu
網路書籤 Google bookmarks
del.icio.us
hemidemi
facebook
twitter
google
reddit