Derin Sinir Ağlarını Kullanan Göğüs Röntgenleri ile Otomatik Tüberküloz Sınıflandırması Örnek Çalışma: Nijerya Halk Sağlığı

Year 2024, Volume: 19 Issue: 1, 55 - 64, 28.03.2024

Muhammad Zaharaddeen Abubakar Mustafa Kaya Mustafa Eriş Mohammed Mansur Abubakar Serkan Karakuş Khalid Jibril Sani

https://doi.org/10.55525/tjst.1222836

Abstract

Bulaşıcı bir akciğer rahatsızlığı olan tüberküloz, önde gelen küresel ölüm faktörü olarak karşımıza çıkıyor. Nijerya'da halk sağlığı üzerindeki önemli etkisi, kapsamlı müdahale stratejilerini gerektirmektedir. Bu hastalığın tespit edilmesi, önlenmesi ve tedavi edilmesi hâlâ zorunludur. Tanı araçları arasında göğüs röntgeni (CXR) görüntüleri çok önemli bir role sahiptir. Derin öğrenmedeki son gelişmeler, tıbbi görüntü analizini önemli ölçüde iyileştirdi. Bu araştırmada, sağlam modeller oluşturmak için kamuya açık ve tescilli CXR görüntü veri kümelerinden yararlandık. Önceden eğitilmiş derin sinir ağlarından yararlanarak tüberküloz tespitini geliştirmeyi hedefledik. Etkileyici bir şekilde, deneylerimiz dikkate değer sonuçlar verdi. Özellikle, ilgili açık ve özel veri setlerinde %98 ve %86'lık f1 puanlarına ulaşıldı. Bu sonuçlar, CXR görüntülerinden tüberkülozun etkili bir şekilde tanımlanmasında derin sinir ağlarının gücünün altını çiziyor. Çalışma, bu teknolojinin hastalığın yaygın etkisiyle mücadelede umut vaat ettiğini belirliyor.

Keywords

Derin öğrenme, Tüberküloz, ResNet, MobileNet, Veri arttırma

Automated Tuberculosis Classification with Chest X-Rays Using Deep Neural Networks -Case Study: Nigerian Public Health

Abstract

Tuberculosis, a contagious lung ailment, stands as a prominent global mortality factor. Its significant impact on public health in Nigeria necessitates comprehensive intervention strategies. Detecting, preventing, and treating this disease remains imperative. Chest X-ray (CXR) images hold a pivotal role among diagnostic tools. Recent strides in deep learning have notably improved medical image analysis. In this research, we harnessed publicly available and proprietary CXR image datasets to construct robust models. Leveraging pre-trained deep neural networks, we aimed to enhance tuberculosis detection. Impressively, our experimentation yielded remarkable outcomes. Notably, f1-scores of 98% and 86% were attained on the respective public and private datasets. These results underscore the potency of deep neural networks in effectively identifying tuberculosis from CXR images. The study emphasizes the promise of this technology in combating the disease's spread and impact.

Keywords

Deep Learning, Tuberclosis, ResNet, MobileNet, Data Augmentation

References

• Desmon S. “Taking a Deep Dive into Why Nigeria’s TB Rates are So High,” 2018. https://ccp.jhu.edu/2018/10/22/nigeria-tb-rates-high-tuberculosis/ (Last accessed Aug. 29, 2023).
• WHO. “The End TB Strategy Global strategy and targets for tuberculosis prevention, care and control after 2015,” 2015, https://apps.who.int/iris/rest/bitstreams/1271371/retrieve. (Last accessed Aug. 29, 2023).
• Alcantra MF, Liu C, Liu B, Brunette M, Zhang N, Sun T, Zhang P, Chen Q, Li Y, Albarracin CM, Peinado J, Garavito ES, Garcia LL, Curioso WH. “Improving tuberculosis diagnostics using deep learning and mobile health technologies among resource-poor communities in Perú,” Smart Heal., vol. 1–2, no. March, pp. 66–76, 2017, doi: 10.1016/j.smhl.2017.04.003.
• WHO. Global Tuberculosis report 2016, https://apps.who.int/iris/bitstream/handle/10665/250441/9789241565394-eng.pdf?sequence=1 (Last accessed Aug. 29, 2023)
• Harries AD et al.. “Deaths from tuberculosis in sub-Saharan African countries with a high prevalence of HIV-1,” Lancet Infect. Dis., vol. 357, no. 9267, pp. 1519–1523, 2001, doi: 10.1016/S0140-6736(00)04639-0.
• CDC A. “Tuberculosis,” Tuberculosis in Africa. https://africacdc.org/disease/tuberculosis/#:~:text=TB, Africa CDC, (Last accessed Aug. 29, 2023).
• WHO. “World Tuberculosis Day,” 2018. https://www.afro.who.int/health-topics/tuberculosis-tb. (Last accessed Aug. 29, 2023).
• Vassall A and Gidado M. “Post-2015 Development Agenda: Nigeria Perspectives; Tuberculosis,” pp. 1–40, 2015.
• Chukwu O and Okoeguale B. “FIRST National TB Prevalence Survey 2012, Nigeria,” Nigeria, 2012. . Available: https://www.who.int/tb/publications/NigeriaReport_WEB_NEW.pdf. (Last accessed Aug. 29, 2023).
• Odume B, Falkun V, Chukwuogo O, Ogbudebe C, Useni S, Nwokoye N, Aniwada E, Olusola Faleye B, Okekearu I, Nongo D, Odusote T, and Lawanson A. Impact of COVID-19 on TB active case finding in Nigeria, Public Health Action. 2020 Dec 21; 10(4): 157–162. doi: 10.5588/pha.20.0037
• WHO. “Chest Radiography in Tuberculosis,” WHO Libr. Cat. Data, p. 44, 2016, . Available: http://www.who.int/about/licensing/copyright_form%0Ahttp://www.who.int/about/licensing/copyright_form). (Last accessed Aug. 29, 2023).
• Wang S and Summers RM. “Machine learning and radiology,” Med. Image Anal., vol. 16, no. 5, pp. 933–951, 2012, doi: 10.1016/j.media.2012.02.005.
• Lopes UK and Valiati JF. “Pre-trained convolutional neural networks as feature extractors for tuberculosis detection,” Comput. Biol. Med., vol. 89, pp. 135–143, 2017, doi: 10.1016/j.compbiomed.2017.08.001.
• Chen B, Zhang L, Chen H, Liang K, and Chen X. “A novel extended Kalman filter with support vector machine based method for the automatic diagnosis and segmentation of brain tumors”, Comput. Methods Programs Biomed. 2021 Mar;200:105797. doi: 10.1016/j.cmpb.2020.105797. PMID: 3331787
• Houssein EH, Emam MM, Ali AA, and Suganthan PN. “Deep and machine learning techniques for medical imaging-based breast cancer: A comprehensive review,” Expert Syst. Appl., no. April, p. 114161, 2020, doi: 10.1016/j.eswa.2020.114161.
• Ozturk T, Talo M, Yildirim EA, Baloglu UB, Yildirim O, and Rajendra Acharya U. “Automated detection of COVID-19 cases using deep neural networks with X-ray images,” Comput. Biol. Med., vol. 121, no. April, p. 103792, 2020, doi: 10.1016/j.compbiomed.2020.103792.
• Olsen CR, Mentz RJ, Anstrom KJ, Page D, and Patel PA. “Clinical applications of machine learning in the diagnosis, classification, and prediction of heart failure: Machine learning in heart failure,” Am. Heart J., vol. 229, pp. 1–17, 2020, doi: 10.1016/j.ahj.2020.07.009.
• Chaki J, Thillai Ganesh S, Cidham SK, and Ananda Theertan S. “Machine learning and artificial intelligence based Diabetes Mellitus detection and self-management: A systematic review,” J. King Saud Univ. - Comput. Inf. Sci., no. xxxx, 2020, doi: 10.1016/j.jksuci.2020.06.013.
• Rajpurkar P et al.. “Deep learning for chest radiograph diagnosis: A retrospective comparison of the CheXNeXt algorithm to practicing radiologists,” PLoS Med., vol. 15, no. 11, pp. 1–17, 2018, doi: 10.1371/journal.pmed.1002686.
• Pasa F, Golkov V, Pfeiffer F, Cremers D, and Pfeiffer D. “Efficient Deep Network Architectures for Fast Chest X-Ray Tuberculosis Screening and Visualization,” Sci. Rep., vol. 9, no. 1, pp. 2–10, 2019, doi: 10.1038/s41598-019-42557-4.
• Krizhevsky A, Sutskever I, and Hinton GE. “ImageNet Classification with Deep Convolutional Neural Networks,” Commun. ACM, vol. 60, no. 6, pp. 84–90, June 2017. doi = 10.1145/3065386
• Xiong Y, Ba X, Hou A, Zhang K, Chen L, and Li T. “Automatic detection of mycobacterium tuberculosis using artificial intelligence,” J. Thorac. Dis., vol. 10, no. 3, pp. 1936–1940, 2018, doi: 10.21037/jtd.2018.01.91.
• Hwang EJ et al.. “Development and Validation of a Deep Learning-based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs,” Clin. Infect. Dis., vol. 69, no. 5, pp. 739–747, 2019, doi: 10.1093/cid/ciy967.
• Rajpurkar P et al.. “CheXaid_ deep learning assistance for physician diagnosis of tuberculosis using chest x-rays in patients with HIV,” npj Digital Med., vol. 3, no. 1, pp. 1–8, 2020, doi: 10.1038/s41746-020-00322-2.
• Rahman T et al.. “Reliable tuberculosis detection using chest x-ray with deep learning, segmentation, and visualization,” IEEE Access., August, 2020, doi: 10.1109/access.2020.3031384.
• Hwang S, Kim HE, Jeong J, and Kim HJ. “A novel approach for tuberculosis screening based on deep convolutional neural networks,” Med. Imaging 2016 Comput. Diagnosis, vol. 9785, p. 97852W, 2016, doi: 10.1117/12.2216198.
• Hooda R, Mittal A, and Sofat S. “Automated TB classification using an ensemble of deep architectures,” Multimed. Tools Appl., vol. 78, no. 22, pp. 31515–31532, 2019, doi: 10.1007/s11042-019-07984-5.
• Hijazi MHA, Yang LQ, Alfred R, Mahdin H, and Yaakob R. “Ensemble deep learning for tuberculosis detection,” Indones. J. Electr. Eng. Comput. Sci., vol. 17, no. 2, pp. 1014–1020, 2019, doi: 10.11591/ijeecs.v17.i2.pp1014-1020.
• Hwa SKT, Hijazi MHA, Bade A, Yaakob R, and Jeffree MS. “Ensemble deep learning for tuberculosis detection using chest X-ray and canny edge detected images,” Int. J. Artif. Intell., vol. 8, no. 4, pp. 429–435, 2019, doi: 10.11591/ijai.v8.i4.pp429-435.
• Lakhani P and Sundaram B. “Deep Learning at Chest Radiography: Automated Classification of Pulmonary Tuberculosis by Using Convolutional Neural Networks.” Radiology, vol. 284, no. 2, pp. 574–582, Aug., 2017. doi: 10.1148/radiol.2017162326. Epub 2017 Apr 24. PMID: 28436741.