استفاده از شبکه¬های عصبی کانولوشن جهت تعیین اندازه تومور و متاستاز غدد لنفاوی در تصاویر تشدید مغناطیسی بیماران سرطان کولون
محورهای موضوعی : مهندسی پزشکی
محمدرضا هدیه زاده
1
*
,
مهدی یوسفی
2
1 - گروه مهندسی پزشکی، واحد دزفول، دانشگاه آزاد اسلامی، دزفول، ایران.
2 - گروه مهندسی برق، واحد نجف آباد، دانشگاه آزاد اسلامی، نجف آباد، ایران.
کلید واژه: سرطان کولون, VGG-16, تصویربرداری تشدید مغناطیسی, اندازه تومور, متاستاز به غدد لنفاوی,
چکیده مقاله :
سرطان روده بزرگ سومین سرطان رایج ازکل سرطانها در سال 2020 است. انکولوژیست در تعیین درمان مناسب ناگزیر از دانستن مرحله تومور بوده که متداولترین روش سیستم مرحلهبندی "تومور، گره، متاستاز" است. لذا در این مطالعه سعی بر آن است تا به کمک تصاویر رزونانس مغناطیسی، اندازه تومور و میزان گسترش به غدد لنفاوی تخمین زده شود. دادههای این مطالعه از پرتال TCIA جمعآوری شده و در اولین گام پردازشی، پس از ارتقاء کیفیت تصویر، ناحیهبندی جهت جداسازی ناحیه تومور از کل تصویر انجام گرفت. در این مطالعه جهت استخراج ویژگی و طبقهبندی توأمان، از شبکه عصبی کانولوشن VGG-16 و برای اعتبار سنجی از روش اعتبار سنجی متقابل ده لایه استفاده شده است. نتایج این مطالعه حاکی از صحت 2/94% است که نسبت به روشهای پاتولوژیک و توموگرافی کامپیوتری با صحتهای 91% و 77%، کارایی بهتری را نشان داده است. لازم به ذکر است که کارایی الگوریتم پیاده شده در تفکیک طبقه یک از سایر طبقات بیشتر بوده است اما اختلاف معناداری بین مقادیر میانگین پارامترها در سه طبقه دیده نشده است.
Colon cancer is the third most common cancer from all cancers in 2020. In determining the appropriate treatment, the oncologist has to know the stage of the tumor, which is the most common method of the “Tumour, Node, Metastasis” staging system. Therefore, in this study, an attempt is made to estimate the size of the tumor and the extent of its spread to the lymph nodes with the help of magnetic resonance images. The data of this study was collected from the TCIA portal and in the first processing step, after improving the image quality, segmentation was done to separate the tumor region from the whole image. In this study, in order to extract and classify the features, the convolutional neural network VGG-16 was used and for the validation, the 10-fold method was applied. The results of this study indicate an accuracy of 94.2%, which is better than pathological and CT methods with accuracy of 91% and 77%. It should be noted that the effectiveness of the implemented algorithm in separating the 1st class was higher than the other class. Still, no significant difference was seen between the average values of the parameters in the three classes.
[1] V. Mahadevan, “Anatomy of the caecum, appendix and colon,” Surg., vol. 38, no. 1, pp. 1–6, 2020. https://doi.org/10.1016/j.mpsur.2019.10.017
[2] P. Boyle and J. S. Langman, “ABC of colorectal cancer: Epidemiology,” BMJ Br. Med. J., vol. 321,
no. 7264, p. 805, 2000. https://doi.org/10.1136/sbmj.0012452
[3] M. Araghi et al., “Global trends in colorectal cancer mortality: projections to the year 2035,” Int. J. cancer, vol. 144, no. 12, pp. 2992–3000, 2019. https://doi.org/10.1002/ijc.32055
[4] M. S. Hossain et al., “Colorectal cancer: a review of carcinogenesis, global epidemiology, current challenges, risk factors, preventive and treatment strategies,” Cancers (Basel)., vol. 14, no. 7, p. 1732, 2022. https://doi.org/10.3390/cancers14071732
[5] Y. Xi and P. Xu, “Global colorectal cancer burden in 2020 and projections to 2040,” Transl. Oncol., vol. 14, no. 10, p. 101174, 2021. https://doi.org/10.1016/j.tranon.2021.101174
[6] M. L. H. Ong and J. B. Schofield, “Assessment of lymph node involvement in colorectal cancer,” World J. Gastrointest. Surg., vol. 8, no. 3, p. 179, 2016. https://doi.org/ 10.4240/wjgs.v8.i3.179
[7] D. Tamandl, T. Mang, and A. Ba-Ssalamah, “Imaging of colorectal cancer–the clue to individualized treatment,” Innov. Surg. Sci., vol. 3, no. 1, pp. 3–15, 2018. https://doi.org/10.1515/iss-2017-0049
[8] R. Vikram and R. B. Iyer, “PET/CT imaging in the diagnosis, staging, and follow-up of colorectal cancer,” Cancer Imaging, vol. 8, no. Spec Iss A, p. S46, 2008. https://doi.org/10.1102/1470-7330.2008.9009
[9] M. Tharwat, N. A. Sakr, S. El-Sappagh, H. Soliman, K.-S. Kwak, and M. Elmogy, “Colon cancer diagnosis based on machine learning and deep learning: Modalities and analysis techniques,” Sensors, vol. 22, no. 23, p. 9250, 2022. https://doi.org/10.3390/s22239250
[10] L. Buk Cardoso et al., “Machine learning for predicting survival of colorectal cancer patients,” Sci. Rep., vol. 13, no. 1, p. 8874, 2023. https://doi.org/10.1038/s41598-023-35649-9
[11] B. Burnett et al., “Machine Learning in Colorectal Cancer Risk Prediction from Routinely Collected Data: A Review,” Diagnostics, vol. 13, no. 2, p. 301, 2023. https://doi.org/10.3390/diagnostics13020301
[12] H. Abdul Rahman, M. A. Ottom, and I. D. Dinov, “Machine learning-based colorectal cancer prediction using global dietary data,” BMC Cancer, vol. 23, no. 1, p. 144, 2023 https://doi.org/10.1186/s12885-023-10587-x
[13] C. B. Finn et al., “Development of a Machine Learning Model to Identify Colorectal Cancer Stage in Medicare Claims,” JCO Clin. Cancer Informatics, vol. 7, p. e2300003, 2023. https://doi.org/10.1200/CCI.23.00003
[14] H. M. Bülbül, G. Burakgazi, and U. Kesimal, “Preoperative assessment of grade, T stage, and lymph node involvement: machine learning-based CT texture analysis in colon cancer,” Jpn. J. Radiol., pp. 1–8, 2023. https://doi.org/10.1007/s11604-023-01502-2
[15] R. Awan et al., “Glandular morphometrics for objective grading of colorectal adenocarcinom
histology images,” Sci. Rep., vol. 7, no. 1, p. 16852, 2017. https://doi.org/10.1038/s41598-017-16516-w
[16] N. Ito, H. Kawahira, H. Nakashima, M. Uesato, H. Miyauchi, and H. Matsubara, “Endoscopic diagnostic support system for cT1b colorectal cancer using deep learning,” Oncology, vol. 96, no. 1, pp. 44–50, 2018. https://doi.org/10.1159/000491636
[17] J. Jian et al., “Fully convolutional networks (FCNs)-based segmentation method for colorectal tumors on T2-weighted magnetic resonance images,” Australas. Phys. Eng. Sci. Med., vol. 41, pp. 393–401, 2018. https://doi.org/10.1007/s13246-018-0636-9
[18] J. Kim et al., “Rectal cancer: Toward fully automatic discrimination of T2 and T3 rectal cancers using deep convolutional neural network,” Int. J. Imaging Syst. Technol., vol. 29, no. 3, pp. 247–259, 2019. https://doi.org/10.1002/ima.22311
[19] J. Li, P. Wang, Y. Zhou, H. Liang, and K. Luan, “Different machine learning and deep learning methods for the classification of colorectal cancer lymph node metastasis images,” Front. Bioeng. Biotechnol., vol. 8, p. 620257, 2021. https://doi.org/10.3389/fbioe.2020.620257
[20] M. Yousefi and M. Hedyehzadeh, “Recurrence Detection of Non-small cell Lung Cancer (NSCLC) Tumors Using Imaging and Clinical Features,” J. Heal. Biomed. Informatics, vol. 9, no. 1, pp. 40–49, 2022.
[21] M. Hedyehzadeh, K. Maghooli, M. MomenGharibvand, and S. Pistorius, “A Comparison of the Efficiency of Using a Deep CNN Approach with Other Common Regression Methods for the Prediction of EGFR Expression in Glioblastoma Patients,” J. Digit. Imaging, 2020, https://doi.org/10.1007/s10278-019-00290-4.
[22] M. Hedyehzadeh, K. Maghooli, and M. MomenGharibvand, “Glioma grade detection using grasshopper optimization algorithm‐optimized machine learning methods: The Cancer Imaging Archive study,” Int. J. Imaging Syst. Technol., vol. 31, no. 3, pp. 1670–1677, 2021. https://doi.org/10.1002/ima.22536
[23] M. Hedyehzadeh, S. Y. D. Nezhad, and N. Safdarian, “EVALUATION OF CONVENTIONAL MACHINE LEARNING METHODS FOR BRAIN TUMOUR TYPE CLASSIFICATION,” COMPTES RENDUS L Acad. Bulg. DES Sci., vol. 73, no. 6, pp. 856–865, 2020. https://doi.org/10.7546/CRABS.2020.06.14
[24] M. Hedyehzadeh, J. Pu, S. Leilizadeh, S. Gezer, C. Dresser, and C. A. Beeche, “A Comparison of Deep and Conventional Regression Methods for MRI-Based Estimation of Survival Time in GBM Patients,” Available SSRN 4001112. https://doi.org/10.2139/ssrn.4001112
[25] A. Achak and M. Hedyehzadeh, “Determining the Differentiation of Benign and Malignant NME Lesions in Contrast-Enhanced Spectral Mammography Images Based on Convolutional Neural Networks,” J. Med. Biol. Eng., vol. 43, no. 5, pp. 585–595, 2023. https://doi.org/10.1007/s40846-02300814-y
[26] M. Hedyehzadeh and M. Yousefi, “Application of Machine Learning Methods to Predict the Survival Rate of Glioblastoma Patients Using MR Images,” J. Heal. Biomed. Informatics, vol. 11, no. 1, pp. 1–13, 2024. https://doi.org/10.34172/jhbmi.2024.09
[27] A. Achak and M. Hedyehzadeh, “Assessing the Efficiency of Deep Learning Methods in Estimating the Malignancy of Bi-Rads 4 Breast Lesions Using Contrast-enhanced Spectral Mammography Images,” Qom Univ. Med. Sci. J., vol. 17, no. 1, p. 0, 2023. https://doi.org/10.32598/qums.17.2756.1
[28] K. Clark et al., “The Cancer Imaging Archive (TCIA): maintaining and operating a public information repository,” J. Digit. Imaging, vol. 26, pp. 1045–1057, 2013. https://doi.org/10.1007/s10278-013-9622-7
[29] M. Ito and F. Ino, “An Automated Method for Generating Training Sets for Deep Learning based Image Registration.,” in BIOIMAGING, 2018, pp. 140–147. https://doi.org/10.5220/0006634501400147