子宮內膜癌是女性生殖系統中常見的惡性腫瘤之一，且其發病率在過去幾十年中呈現上升的趨勢。而如何將子宮內膜癌在組織病理學上有效分出其分級以及在分子分型的分類上來辨識出其微衛星不穩定 (MSI) 狀態就成為了重要的課題。
本研究使用了癌症基因族譜 (TCGA) 與台灣台北國防醫學中心三軍總醫院所提供被診斷為子宮內膜癌之患者的全景域玻片影像 (WSI)。其中在 TCGA 資料集內，共有 529 位子宮內膜癌患者，並利用Polymerase Chain Reaction(PCR) 和 Next­Generation Sequencing(NGS) 兩種技術來進行 MSI 狀態的標註。在台灣台北國防醫學中心三軍總醫院資料集內，共有 225 位子宮內膜癌患者，並利用免疫組織化學 (Immunohistochemistry,IHC) 之四種 MMR 酶 (MLH1,MSH2,MSH6,PMS2)的缺失數量來進行 MSI 狀態的標註。並依照患者的組織學分級進行組別劃分，分別為判斷組織學分級的 ((G1,G2)，(G3,Serous carcinoma))，以及判斷 MSI 狀態的(G1,G2)、(G3)、(G1,G2,G3)。
本研究中藉由蘇木精­伊紅染色 (Hematoxylin and eosin, H&E) 影像，結合深度學習提出一個用於子宮內膜癌組織學分級和 MSI 狀態的預測系統。並與其他先進的深度學習方法進行比較，其中包括:CLAM,TOAD 和 MIL RNN。而在 NGS 技
術所標註的 TCGA 資料集有良好的結果。在 G1,G2 組別中，Accuarcy 為 0.887、Precision 為 0.875、F1­score 為 0.925、Sensitivity 為 0.980、MSS 為 0.823。在 G3組別中，Accuarcy 為 0.836、Precision 為 0.813、F1­score 為 0.871、Sensitivity 為 0.938、MSS 為 0.814。證明了透過此方法能輔助預測患者在 NGS 之標註下的 MSI狀態。

Endometrial cancer is one of the common malignant tumors in the female reproductive system, and its incidence has shown an upward trend in the past few decades. How to effectively classify endometrial cancer into histopathological grades and identify its microsatellite instability (MSI) status in terms of molecular typing has become an important issue.
This study used The Cancer Genome Atlas Program (TCGA) and the National Defense Medical Center, Tri­Service General Hospital in Taipei, Taiwan, with whole slide images (WSI) of patients diagnosed with endometrial cancer. Among them, in the TCGA
data set, there are 529 patients with endometrial cancer, and the two technologies of Polymerase Chain Reaction (PCR) and Next­Generation Sequencing (NGS) are used to mark the MSI status. In Taiwan National Defense Medical Center, Tri­Service General Hospital dataset, there are 225 patients with endometrial cancer, and the deletion of four MMR enzymes (MLH1, MSH2, MSH6, PMS2) in immunohistochemistry (IHC) was used to determine the MSI status. According to the histological grade of the patients, the groups are divided into ((G1, G2), (G3, Serous carcinoma)) for judging the histological grade, and (G1, G2), (G3), (G1,G2,G3) for judging the MSI status.
In this study, hematoxylin­eosin (H&E) staining images combined with deep learning proposed a prediction system for endometrial cancer histological grading and MSI status. And compared with other advanced deep learning methods, including: CLAM, TOAD and MIL RNN. However, the TCGA dataset marked by NGS technology has good results. In the G1 and G2 groups, Accuarcy is 0.887, Precision is 0.875, F1­score is 0.925, Sensitivity is 0.980, and MSS is 0.823. In the G3 group, Accuarcy was 0.836, Precision was 0.813, F1­score was 0.871, Sensitivity was 0.938, and MSS was 0.814. It is proved that this method can assist in predicting the MSI status of patients under NGS labeling.

[1] L. Rebecca, D. Kimberly, E. Hannah, and A. Jemal, “Cancer statistics,” CA Cancer
J Clin, vol. 72, pp. 7–33, 2022.
[2] N. Colombo, C. Creutzberg, F. Amant, T. Bosse, A. González­Martín, J. Ledermann, C. Marth, R. Nout, D. Querleu, M. Mirza, C. Sessa, M. Abal, O. Altundag,
F. Amant, A. van Leeuwenhoek, S. Banerjee, T. Bosse, A. Casado, L. de Agustín,
D. Cibula, N. Colombo, C. Creutzberg, J.­M. del Campo, G. Emons, F. Goffin,
A. González­Martín, S. Greggi, C. Haie­Meder, D. Katsaros, V. Kesic, C. Kurzeder,
S. Lax, F. Lécuru, J. L. T. Levy, D. Lorusso, J. Mäenpää, C. Marth, X. MatiasGuiu, P. Morice, H. Nijman, R. Nout, M. Powell, D. Querleu, M. Mirza, N. Reed,
A. Rodolakis, H. Salvesen, J. Sehouli, C. Sessa, A. Taylor, A. Westermann, and
A. Zeimet, “Esmo­esgo­estro consensus conference on endometrial cancer: diagnosis, treatment and follow­up,” Annals of Oncology, vol. 27, pp. 16–41, 2016.
[3] A. Jamieson, T. Bosse, and J. N. McAlpine, “The emerging role of molecular pathology in directing the systemic treatment of endometrial cancer,” Therapeutic advances in medical oncology, vol. 13, p. 17588359211035959, 2021.
[4] C. G. A. R. Network, C. Kandoth, N. Schultz, A. D. Cherniack, R. Akbani, Y. Liu,
H. Shen, A. G. Robertson, I. Pashtan, R. Shen, C. C. Benz, C. Yau, P. W. Laird,
L. Ding, W. Zhang, G. B. Mills, R. Kucherlapati, E. R. Mardis, and D. A. Levine,
“Integrated genomic characterization of endometrial carcinoma,” Nature, vol. 497,
no. 7447, pp. 67–73, 2013.
[5] S. Kommoss, M. K. McConechy, F. Kommoss, S. Leung, A. Bunz, J. Magrill,
H. Britton, F. Kommoss, F. Grevenkamp, A. Karnezis, W. Yang, A. Lum, B. Krämer,
F. Taran, A. Staebler, S. Lax, S. Y. Brucker, D. G. Huntsman, C. B. Gilks, J. N. M.
McAlpine, and A. Talhouk, “Final validation of the promise molecular classifier for
endometrial carcinoma in a large population­based case series,” Annals of oncology : official journal of the European Society for Medical Oncology, vol. 29, no. 5,
pp. 1180–1188, 2018.
[6] A. Raffone, A. Travaglino, O. Gabrielli, M. Micheli, V. Zuccalà, G. Bitonti, C. Camastra, V. Gargiulo, L. Insabato, and F. Zullo, “Clinical features of promise groups
49
identify different phenotypes of patients with endometrial cancer,” Archives of gynecology and obstetrics, vol. 303, no. 6, pp. 1393–1400, 2021.
[7] T. Bosse, M. Powell, E. Crosbie, A. Leary, J. Kroep, K. Han, J. Mcalpine,
N. Horeweg, S. D. Boer, M. D. Bruyn, R. Nout, V. Smit, H. Nijman, N. Singh,
H. Mackay, R. Edmondson, L. Mileshkin, D. Church, H. Kitchener, and
C. Creutzberg, “595 implementation of collaborative translational research (transportec) findings in an international endometrial cancer clinical trials program
(rainbo),” vol. 31, no. Suppl 3, pp. A108–A109, 2021.
[8] M. K. Talhouk, A. andMcConechy, S. Leung, H. H. Li­Chang, J. S. Kwon, N. Melnyk, W. Yang, J. Senz, N. Boyd, A. N. Karnezis, D. G. Huntsman, C. B. Gilks, and
J. N. McAlpine, “A clinically applicable molecular­based classification for endometrial cancers,” British journal of cancer, vol. 113, no. 2, pp. 299–310, 2015.
[9] A. Raffone, A. Travaglino, M. Mascolo, L. Carbone, M. Guida, L. Insabato, and
F. Zullo, “Tcga molecular groups of endometrial cancer: Pooled data about prognosis,” Gynecologic oncology, vol. 155, no. 2, pp. 374–383, 2019.
[10] N. Goel, A. Yadav, and B. M. Singh, “Medical image processing: A review,” in
2016 Second International Innovative Applications of Computational Intelligence
on Power, Energy and Controls with their Impact on Humanity (CIPECH), 2016.
[11] U. Bhattacharjya, K. K. Sarma, J. P. Medhi, B. K. Choudhury, and G. Barman,
“Automated diagnosis of covid­19 using radiological modalities and artificial intelligence functionalities: A retrospective study based on chest hrct database,” Biomed
Signal Process Control, vol. 80, p. 104297, 2023.
[12] C. Jin, H. Yu, J. Ke, P. Ding, Y. Yi, X. Jiang, X. Duan, J. Tang, D. T. Chang, X. Wu,
F. Gao, and R. Li, “Predicting treatment response from longitudinal images using
multi­task deep learning,” Nature communications, vol. 12, no. 1, p. 1851, 2021.
[13] D. Shen, G. Wu, and H.­I. Suk, “Deep learning in medical image analysis,” Annual
Review of Biomedical Engineering, vol. 19, no. 1, pp. 221–248, 2017.
[14] J. Weese and C. Lorenz, “Four challenges in medical image analysis from an industrial perspective,” Medical image analysis, vol. 33, pp. 44–49, 2016.
50
[15] K. Li, Z. Qian, Y. Han, E. I.­C. Chang, B. Wei, M. Lai, J. Liao, Y. Fan, and Y. Xu,
“Weakly supervised histopathology image segmentation with self­attention,” Medical Image Analysis, vol. 86, p. 102791, 2023.
[16] G. Campanella, M. G. Hanna, L. Geneslaw, A. Miraflor, V. W. K. Silva, K. J. Busam,
E. Brogi, V. E. Reuter, D. S. Klimstra, and T. J. Fuchs, “Clinical­grade computational
pathology using weakly supervised deep learning on whole slide images,” Nature
medicine, vol. 25, no. 8, pp. 1301–1309, 2019.
[17] Y. F. Li, L. Z. Guo, and Z. H. Zhou, “Towards safe weakly supervised learning,”
IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 43, no. 1,
pp. 334–346, 2021.
[18] J. Kim, H. Kim, C. Kim, J. Lee, K. Kim, Y. Park, H. Kim, S. Ki, Y. Kim, and W. Kim,
“Weakly­supervised deep learning for ultrasound diagnosis of breast cancer,” Sci
Rep, vol. 11, no. 24382, 2021.
[19] M. Z. Liu, C. Swintelski, S. Sun, M. Siddique, E. Desperito, S. Jambawalikar, and
R. Ha, “Weakly supervised deep learning approach to breast mri assessment,” Academic Radiology, vol. 29, pp. S166–S172, 2022.
[20] M. Lu, D. Williamson, T. Chen, R. Chen, M. Barbieri, and F. Mahmood, “Dataefficient and weakly supervised computational pathology on whole­slide images,”
Nat Biomed Eng, vol. 5, pp. 555–570, 2021.
[21] M. Y. Lu, T. Y. Chen, D. F. K. Williamson, M. Zhao, M. Shady, J. Lipkova, and
F. Mahmood, “Ai­based pathology predicts origins for cancers of unknown primary,”
Nature, vol. 594, pp. 106–110, 2021.
[22] Y. J. Lin, T. K. Chao, M. A. Khalil, Y. C. Lee, D. Z. Hong, J. J. Wu, and C. W.
Wang, “Deep learning fast screening approach on cytological whole slides for thyroid
cancer diagnosis,” Cancers, vol. 13, no. 15, p. 3891, 2021.
[23] E. Shelhamer, J. Long, and T. Darrell, “Fully convolutional networks for semantic segmentation,” IEEE transactions on pattern analysis and machine intelligence,
vol. 39, no. 4, pp. 640–651, 2017.
51
[24] X. Wang, R. Zhang, T. Kong, L. Li, and C. Shen, “Solov2: Dynamic and fast instance segmentation,” Advances in Neural information processing systems, vol. 33,
pp. 17721–17732, 2020.
[25] S. Yadav and S. Jadhav, “Deep convolutional neural network based medical image
classification for disease diagnosis,” J Big Data, vol. 6, no. 113, 2019.
[26] H. A. Qadir, Y. Shin, J. Solhusvik, J. Bergsland, L. Aabakken, and I. Balasingham,
“Toward real­time polyp detection using fully cnns for 2d gaussian shapes prediction,” Medical Image Analysis, vol. 68, p. 101897, 2021.
[27] S. Niyas, S. Pawan, M. Anand Kumar, and J. Rajan, “Medical image segmentation with 3d convolutional neural networks: A survey,” Neurocomputing, vol. 493,
pp. 397–413, 2022.
[28] J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for semantic
segmentation,” in 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3431–3440, 2015.
[29] O. Ronneberger, P. Fischer, and T. Brox, “U­net: Convolutional networks for
biomedical image segmentation,” 2015.
[30] V. Badrinarayanan, A. Handa, and R. Cipolla, “Segnet: A deep convolutional
encoder­decoder architecture for robust semantic pixel­wise labelling,” 2015.
[31] K. He, G. Gkioxari, P. Dollár, and R. Girshick, “Mask r­cnn,” 2018.
[32] X. Li, S. Wang, X. Wei, J. Zhu, R. Yu, M. Zhao, M. Yu, Z. Liu, and S. Liu, “Fully convolutional networks for ultrasound image segmentation of thyroid nodules,” in 2018
IEEE 20th International Conference on High Performance Computing and Communications; IEEE 16th International Conference on Smart City; IEEE 4th International Conference on Data Science and Systems (HPCC/SmartCity/DSS), pp. 886–
890, 2018.
[33] C. W. Wang, K. Y. Lin, Y. J. Lin, M. A. Khalil, K. L. Chu, and T. K. Chao, “A soft label deep learning to assist breast cancer target therapy and thyroid cancer diagnosis,”
Cancers, vol. 14, no. 21, 2022.
52
[34] M. A. Khalil, Y. C. Lee, H. C. Lien, Y. M. Jeng, and C. W. Wang, “Fast segmentation
of metastatic foci in h&e whole­slide images for breast cancer diagnosis,” Diagnostics, vol. 12, no. 4, p. 990, 2022.
[35] C.­W. Wang, Y.­C. Lee, Y.­J. Lin, C.­C. Chang, A.­K.­O. Sai, C.­H. Wang, and T.­K.
Chao, “Interpretable attention­based deep learning ensemble for personalized ovarian cancer treatment without manual annotations,” Computerized Medical Imaging
and Graphics, vol. 107, p. 102233, 2023.
[36] C.­W. Wang, Y.­C. Lee, Y.­J. Lin, C.­C. Chang, A.­K.­O. Sai, C.­H. Wang, and T.­
K. Chao, “Ensemble biomarkers for guiding anti­angiogenesis therapy for ovarian
cancer using deep learning,” Clinical and translational medicine, vol. 13, no. 1,
p. e1162, 2023.
[37] C. W. Wang, C. C. Chang, Y. C. Lee, Y. J. Lin, S. C. Lo, P. C. Hsu, Y. A. Liou, C. H.
Wang, and T. K. Chao, “Weakly supervised deep learning for prediction of treatment
effectiveness on ovarian cancer from histopathology images,” Computerized Medical Imaging and Graphics, vol. 99, p. 102093, 2022.
[38] C. W. Wang, Y. C. Lee, C. C. Chang, Y. J. Lin, Y. A. Liou, P. C. Hsu, C. C. Chang,
A. K. O. Sai, C. H. Wang, and T. K. Chao, “A weakly supervised deep learning
method for guiding ovarian cancer treatment and identifying an effective biomarker,”
Cancers, vol. 14, no. 7, p. 1651, 2022.
[39] C. W. Wang, Y. A. Liou, Y. J. Lin, C. C. Chang, P. H. Chu, Y. C. Lee, C. H. Wang,
and T. K. Chao, “Artificial intelligence­assisted fast screening cervical high grade
squamous intraepithelial lesion and squamous cell carcinoma diagnosis and treatment planning,” Scientific Reports, vol. 11, no. 1, pp. 1–14, 2021.
[40] F. Isensee, P. F. Jaeger, S. A. A. Kohl, J. Petersen, and K. H. Maier­Hein, “nnu­net:
a self­configuring method for deep learning­based biomedical image segmentation,”
Nature Methods, vol. 18, pp. 203–211, 2021.
[41] Z. Yan, Y. Su, H. Sun, H. Yu, W. Ma, H. Chi, H. Cao, and Q. Chang, “Segnetbased left ventricular mri segmentation for the diagnosis of cardiac hypertrophy and
myocardial infarction,” Computer Methods and Programs in Biomedicine, vol. 227,
p. 107197, 2022.
53
[42] Y. Ding, Q. Yang, Y. Wang, D. Chen, Z. Qin, and J. Zhang, “Mallesnet: A multiobject assistance based network for brachial plexus segmentation in ultrasound images,” Medical Image Analysis, vol. 80, p. 102511, 2022.
[43] J. Lao, Y. Chen, Z.­C. Li, Q. Li, J. Zhang, J. Liu, and G. Zhai, “A deep learning­based
radiomics model for prediction of survival in glioblastoma multiforme,” Scientific
Reports, vol. 7, p. 10353, 2017.
[44] R. Miotto, L. Li, B. A. Kidd, and J. T. Dudley, “Deep patient: An unsupervised
representation to predict the future of patients from the electronic health records,”
Scientific Reports, vol. 6, p. 26094, 2016.
[45] P. Mobadersany, S. Yousefi, M. Amgad, D. A. Gutman, J. S. Barnholtz­Sloan, J. E. V.
Vega, D. J. Brat, , and L. A. D. Cooper, “Predicting cancer outcomes from histology
and genomics using convolutional networks,” PNAS, vol. 115, no. 13, 2018.
[46] A. Esteva, B. Kuprel, R. A. Novoa, J. Ko, S. M. Swetter, H. M. Blau, and S. Thrun,
“Dermatologist­level classification of skin cancer with deep neural networks,” Nature, vol. 542, pp. 115–118, 2017.
[47] B. Ehteshami Bejnordi, M. Veta, P. Johannes van Diest, B. van Ginneken, N. Karssemeijer, G. Litjens, J. A. W. M. van der Laak, , and the CAMELYON16 Consortium,
“Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node
Metastases in Women With Breast Cancer,” JAMA, vol. 318, no. 22, pp. 2199–2210,
2017.
[48] R. Joyseeree, S. Otálora, H. Müller, and A. Depeursinge, “Fusing learned representations from riesz filters and deep cnn for lung tissue classification,” Medical Image
Analysis, vol. 56, pp. 172–183, 2019.
[49] S. Yadav and S. Jadhav, “Deep convolutional neural network based medical image
classification for disease diagnosis,” J Big Data, vol. 6, p. 113, 2019.
[50] P. Rajpurkar, J. Irvin, R. L. Ball, K. Zhu, B. Yang, H. Mehta, T. Duan, D. Ding,
A. Bagul, Langlotz, C. P., B. N. Patel, K. W. Yeom, K. Shpanskaya, F. G. Blankenberg, J. Seekins, T. J. Amrhein, D. A. Mong, S. S. Halabi, E. J. Zucker, A. Y. Ng,
and M. P. Lungren, “Deep learning for chest radiograph diagnosis: A retrospective
comparison of the chexnext algorithm to practicing radiologists,” PLoS Med, vol. 15,
p. e1002686, 2018.
54
[51] G. Zhao, Q. Feng, C. Chen, Z. Zhou, and Y. Yu, “Diagnose like a radiologist:
Hybrid neuro­probabilistic reasoning for attribute­based medical image diagnosis,”
IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 11,
pp. 7400–7416, 2022.
[52] K.­H. Yu, G. J. Berry, D. L. Rubin, C. Ré, R. B. Altman, and M. Snyder, “Association of omics features with histopathology patterns in lung adenocarcinoma,” Cell
Systems, vol. 5, no. 6, pp. 620–627.e3, 2017.
[53] J. L. Katzman, U. Shaham, A. Cloninger, J. Bates, T. Jiang, and Y. Kluger, “Deepsurv: personalized treatment recommender system using a cox proportional hazards
deep neural network,” BMC Medical Research Methodology, vol. 18, p. 24, 2018.
[54] E. Wulczyn, D. F. Steiner, M. Moran, M. Plass, R. Reihs, F. Tan, Flament­Auvigne,
Isabelle, T. Brown, P. Regitnig, P.­H. C. Chen, N. Hegde, A. Sadhwani, R. MacDonald, B. Ayalew, G. S. Corrado, L. H. Peng, D. Tse, H. Müller, Z. Xu, M. C. Liu,
YunStumpe, K. Zatloukal, and C. H. Mermel, “Interpretable survival prediction for
colorectal cancer using deep learning,” npj Digital Medicine, vol. 4, p. 71, 2021.
[55] Z. Li, J. Zhang, T. Tan, X. Teng, X. Sun, H. Zhao, L. Liu, Y. Xiao, B. Lee, Y. Li,
Q. Zhang, S. Sun, Y. Zheng, J. Yan, N. Li, Y. Hong, J. Ko, H. Jung, Y. Liu, Y.­c. Chen,
C. w. Wang, V. Yurovskiy, P. Maevskikh, V. Khanagha, Y. Jiang, L. Yu, Z. Liu, D. Li,
P. J. Schüffler, Q. Yu, H. Chen, Y. Tang, and G. Litjens, “Deep learning methods for
lung cancer segmentation in whole­slide histopathology images­the acdc@lunghp
challenge 2019,” IEEE J Biomed Health Inform, vol. 25, pp. 429–440, 2021.
[56] E. Shelhamer, J. Long, and T. Darrell, “Fully convolutional networks for semantic
segmentation,” IEEE Transactions on Pattern Analysis and Machine Intelligence,
vol. 39, no. 4, pp. 640–651, 2017.
[57] C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, “Rethinking the inception architecture for computer vision,” in Proceedings of the IEEE conference on
computer vision and pattern recognition, pp. 2818–2826, 2016.