Verimli göğüs röntgeni yorumlamasına doğru: İlgili bölge çıkarımı, ağırlık başlangıçlandırma ve önişleme yöntemleri üzerine bir çalışma

Abstract

Accurate interpretation of chest X-rays (CXRs) using machine learning techniques plays a vital role in improving diagnostic practices within healthcare. This study advances CXR analysis by exploring various weight initialization and preprocessing methods with deep neural networks. Specifically, (i) an object detection model was trained to identify regions of interest (ROIs) within CXRs, enabling focused image analysis; (ii) weight initialization strategies, including partial fine-tuning, complete fine-tuning, and random initialization, were assessed using the EfficientNet-B1 model on both original and cropped images; and (iii) preprocessing techniques such as histogram equalization, bilateral filtering, Gaussian filtering, and Contrast Limited Adaptive Histogram Equalization (CLAHE) were applied to enhance image quality. Results show that ROI selection significantly improves model performance by focusing on relevant image areas (+4.8%). Pre-trained ImageNet weights with complete fine-tuning outperformed random initialization (+6.5%), demonstrating the advantages of transfer learning, particularly with limited datasets. Among preprocessing methods, the combination of CLAHE with Gaussian filtering achieved the highest validation accuracy (0.958), suggesting that advanced preprocessing methods substantially enhance model performance. These results underscore the importance of effective weight initialization and preprocessing in optimizing diagnostic accuracy and efficiency in CXR analysis.

Keywords

• Göğüs röntgeni
• Derin öğrenme
• Tıbbi görüntü analizi
• Ön işleme
• İlgili alan

The Role of Weight Initialization and Preprocessing Techniques in Analyzing Chest X-ray Images with Deep Neural Networks: A Comparative Study

Abstract

Accurate interpretation of chest X-rays (CXRs) using machine learning techniques plays a vital role in improving diagnostic practices within healthcare. This study advances CXR analysis by exploring various weight initialization and preprocessing methods with deep neural networks. Specifically, (i) an object detection model was trained to identify regions of interest (ROIs) within CXRs, enabling focused image analysis; (ii) weight initialization strategies, including partial fine-tuning, complete fine-tuning, and random initialization, were assessed using the EfficientNet-B1 model on both original and cropped images; and (iii) preprocessing techniques such as histogram equalization, bilateral filtering, Gaussian filtering, and Contrast Limited Adaptive Histogram Equalization (CLAHE) were applied to enhance image quality. Results show that ROI selection significantly improves model performance by focusing on relevant image areas (+4.8%). Pre-trained ImageNet weights with complete fine-tuning outperformed random initialization (+6.5%), demonstrating the advantages of transfer learning, particularly with limited datasets. Among preprocessing methods, the combination of CLAHE with Gaussian filtering achieved the highest validation accuracy (0.958), suggesting that advanced preprocessing methods substantially enhance model performance. These results underscore the importance of effective weight initialization and preprocessing in optimizing diagnostic accuracy and efficiency in CXR analysis.

Keywords

• Chest X-rays
• deep learning
• medical image analysis
• preprocessing

References

1. [1] WHO, “The top 10 causes of death.” Accessed: Aug. 03, 2024. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death
2. [2] WHO, “Chronic obstructive pulmonary disease (COPD).” Accessed: Aug. 03, 2024. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/chronic-obstructive-pulmonary-disease-(copd)
3. [3] WHO, “Lung cancer.” Accessed: Aug. 03, 2024. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/lung-cancer
4. [4] C. M. Jones et al., “Chest radiographs and machine learning – Past, present and future,” J. Med. Imaging Radiat. Oncol., vol. 65, no. 5, pp. 538–544, 2021, doi: 10.1111/1754-9485.13274.
5. [5] M. A. Thanoon, M. A. Zulkifley, M. A. A. Mohd Zainuri, and S. R. Abdani, “A Review of Deep Learning Techniques for Lung Cancer Screening and Diagnosis Based on CT Images,” Diagnostics, vol. 13, no. 16, 2023, doi: 10.3390/diagnostics13162617.
6. [6] C. Deniz, “Modern Computer Tomography with Artificial Intelligence and Deep Learning Applications,” Artif. Intell. Theory Appl., vol. 3, no. 2, pp. 123–136, 2023.
7. [7] R. Booij, R. P. J. Budde, M. L. Dijkshoorn, and M. van Straten, “Technological developments of X-ray computed tomography over half a century: User’s influence on protocol optimization,” Eur. J. Radiol., vol. 131, p. 109261, Oct. 2020, doi: 10.1016/j.ejrad.2020.109261.
8. [8] J. M. Jiang, L. Miao, X. Liang, Z. H. Liu, L. Zhang, and M. Li, “The Value of Deep Learning Image Reconstruction in Improving the Quality of Low-Dose Chest CT Images,” Diagnostics, vol. 12, no. 10, p. 2560, Oct. 2022, doi: 10.3390/diagnostics12102560.

9. [9] P. Celard, E. L. Iglesias, J. M. Sorribes-Fdez, R. Romero, A. S. Vieira, and L. Borrajo, “A survey on deep learning applied to medical images: from simple artificial neural networks to generative models,” Neural Comput. Appl., vol. 35, no. 3, pp. 2291–2323, 2023, doi: 10.1007/s00521-022-07953-4.
10. [10] X. Xie, J. Niu, X. Liu, Z. Chen, S. Tang, and S. Yu, “A survey on incorporating domain knowledge into deep learning for medical image analysis,” Med. Image Anal., vol. 69, 2021, doi: 10.1016/j.media.2021.101985.
11. [11] E. Çallı, E. Sogancioglu, B. van Ginneken, K. G. van Leeuwen, and K. Murphy, “Deep learning for chest X-ray analysis: A survey,” Med. Image Anal., vol. 72, 2021, doi: 10.1016/j.media.2021.102125.
12. [12] D. Meedeniya, H. Kumarasinghe, S. Kolonne, C. Fernando, I. D. la T. Díez, and G. Marques, “Chest X-ray analysis empowered with deep learning: A systematic review,” Appl. Soft Comput., vol. 126, 2022, doi: 10.1016/j.asoc.2022.109319.
13. [13] F. Shamshad et al., “Transformers in medical imaging: A survey,” in Medical Image Analysis, 2023. doi: 10.1016/j.media.2023.102802.
14. [14] A. Dosovitskiy et al., “an Image Is Worth 16X16 Words: Transformers for Image Recognition At Scale,” in ICLR 2021 - 9th International Conference on Learning Representations, 2021.
15. [15] X. Wu et al., “CheXNet: Combing Transformer and CNN for Thorax Disease Diagnosis from Chest X-ray Images,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2024, pp. 73–84. doi: 10.1007/978-981-99-8558-6_7.
16. [16] C. C. Ukwuoma et al., “A hybrid explainable ensemble transformer encoder for pneumonia identification from chest X-ray images,” J. Adv. Res., vol. 48, pp. 191–211, 2023, doi: 10.1016/j.jare.2022.08.021.
17. [17] O. F. Soylemez and B. Ergen, “Facial Landmark Based Region of Interest Localization for Deep Facial Expression Recognition,” Teh. Vjesn., vol. 29, no. 1, pp. 38–44, 2022, doi: 10.17559/TV-20200423145443.
18. [18] R. G, “A Study to Find Facts Behind Preprocessing on Deep Learning Algorithms,” J. Innov. Image Process., vol. 3, no. 1, pp. 66–74, 2021, doi: 10.36548/jiip.2021.1.006.
19. [19] M. V. Narkhede, P. P. Bartakke, and M. S. Sutaone, “A review on weight initialization strategies for neural networks,” Artif. Intell. Rev., vol. 55, no. 1, pp. 291–322, 2022, doi: 10.1007/s10462-021-10033-z.
20. [20] U. Avni, H. Greenspan, E. Konen, M. Sharon, and J. Goldberger, “X-ray categorization and retrieval on the organ and pathology level, using patch-based visual words,” IEEE Trans. Med. Imaging, vol. 30, no. 3, pp. 733–746, 2011, doi: 10.1109/TMI.2010.2095026.
21. [21] M. R. Zare, W. C. Seng, and A. Mueen, “Automatic classification of medical X-ray images,” Malaysian J. Comput. Sci., vol. 26, no. 1, pp. 9–22, 2013, doi: 10.22452/mjcs.vol26no1.2.
22. [22] N. R. S. Parveen and M. M. Sathik, “Detection of Pneumonia in chest X-ray images,” J. Xray. Sci. Technol., vol. 19, no. 4, pp. 423–428, 2011, doi: 10.3233/XST-2011-0304.
23. [23] L. Houam, A. Hafiane, A. Boukrouche, E. Lespessailles, and R. Jennane, “Texture characterization using local binary pattern and wavelets. Application to bone radiographs,” 2012 3rd Int. Conf. Image Process. Theory, Tools Appl. IPTA 2012, pp. 371–376, 2012, doi: 10.1109/IPTA.2012.6469546.
24. [24] V. Kumar and C. Science, “Neural Network Based Approach for Detection of Abnormal Regions of Lung Cancer in X-Ray Image,” vol. 1, no. 5, pp. 1–7, 2012.
25. [25] S. Jaeger et al., “Automatic tuberculosis screening using chest radiographs,” IEEE Trans. Med. Imaging, vol. 33, no. 2, pp. 233–245, Feb. 2014, doi: 10.1109/TMI.2013.2284099.
26. [26] H. Arimura, S. Katsuragawa, Q. Li, T. Ishida, and K. Doi, “Development of a computerized method for identifying the posteroanterior and lateral views of chest radiographs by use of a template matching technique,” Med. Phys., vol. 29, no. 7, pp. 1556–1561, 2002, doi: 10.1118/1.1487426.
27. [27] K. C. Santosh et al., “Rotation detection in chest radiographs based on generalized line histogram of rib-orientations,” Proc. - IEEE Symp. Comput. Med. Syst., pp. 138–142, 2014, doi: 10.1109/CBMS.2014.56.
28. [28] Ö. F. Söylemez, “Forward selection-based ensemble of deep neural networks for melanoma classification in dermoscopy images,” Int. J. Imaging Syst. Technol., vol. 33, no. 6, pp. 1929–1943, 2023, doi: 10.1002/ima.22912.
29. [29] Q. Wang et al., “Automated segmentation and diagnosis of pneumothorax on chest X-rays with fully convolutional multi-scale ScSE-DenseNet: a retrospective study,” BMC Med. Inform. Decis. Mak., vol. 20, 2020, doi: 10.1186/s12911-020-01325-5.
30. [30] M. Owais, M. Arsalan, T. Mahmood, Y. H. Kim, and K. R. Park, “Comprehensive computer-aided decision support framework to diagnose tuberculosis from chest X-Ray images: Data mining study,” JMIR Med. Informatics, vol. 8, no. 12, 2020, doi: 10.2196/21790.
31. [31] S. Rajaraman, S. Sornapudi, P. O. Alderson, L. R. Folio, and S. K. Antani, “Analyzing inter-reader variability affecting deep ensemble learning for COVID-19 detection in chest radiographs.,” PLoS One, vol. 15, no. 11, p. e0242301, 2020, [Online]. Available: https://pubmed.ncbi.nlm.nih.gov/33180877/
32. [32] X. Wang et al., “Potential of deep learning in assessing pneumoconiosis depicted on digital chest radiography,” Occup. Environ. Med., vol. 77, no. 9, pp. 597–602, 2020, doi: 10.1136/oemed-2019-106386.
33. [33] G. I. Okolo, S. Katsigiannis, and N. Ramzan, “IEViT: An enhanced vision transformer architecture for chest X-ray image classification,” Comput. Methods Programs Biomed., vol. 226, 2022, doi: 10.1016/j.cmpb.2022.107141.
34. [34] T. Chen et al., “A vision transformer machine learning model for COVID-19 diagnosis using chest X-ray images,” Healthc. Anal., vol. 5, 2024, doi: 10.1016/j.health.2024.100332.
35. [35] D. Kermany, K. Zhang, M. Goldbaum, and others, “Labeled optical coherence tomography (oct) and chest x-ray images for classification,” Mendeley data, vol. 2, no. 2, p. 651, 2018.
36. [36] D. S. Kermany et al., “Identifying Medical Diagnoses and Treatable Diseases by Image-Based Deep Learning,” Cell, vol. 172, no. 5, pp. 1122-1131.e9, 2018, doi: 10.1016/j.cell.2018.02.010.
37. [37] X. Glorot and Y. Bengio, “Understanding the difficulty of training deep feedforward neural networks,” J. Mach. Learn. Res., vol. 9, pp. 249–256, 2010.
38. [38] K. He, X. Zhang, S. Ren, and J. Sun, “Delving deep into rectifiers: Surpassing human-level performance on imagenet classification,” in Proceedings of the IEEE International Conference on Computer Vision, 2015, pp. 1026–1034. doi: 10.1109/ICCV.2015.123.
39. [39] C.-Y. Wang, I.-H. Yeh, and H.-Y. M. Liao, “YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information,” 2024, [Online]. Available: http://arxiv.org/abs/2402.13616
40. [40] I. Loshchilov and F. Hutter, “SGDR: Stochastic gradient descent with warm restarts,” in 5th International Conference on Learning Representations, ICLR 2017, 2017. doi: 10.48550/arXiv.1608.03983.
41. [41] H. Sharma, J. S. Jain, P. Bansal, and S. Gupta, “Feature extraction and classification of chest X-ray images using CNN to detect pneumonia,” in Proceedings of the Confluence 2020 - 10th International Conference on Cloud Computing, Data Science and Engineering, 2020, pp. 227–231. doi: 10.1109/Confluence47617.2020.9057809.
42. [42] M. S. Al Reshan et al., “Detection of Pneumonia from Chest X-ray Images Utilizing MobileNet Model,” Healthcare, vol. 11, no. 11, p. 1561, May 2023, doi: 10.3390/healthcare11111561.
43. [43] X. Wang, Y. Peng, L. Lu, Z. Lu, M. Bagheri, and R. M. Summers, “ChestX-ray8: Hospital-scale chest X-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases,” Proc. - 30th IEEE Conf. Comput. Vis. Pattern Recognition, CVPR 2017, vol. 2017-Janua, pp. 3462–3471, 2017, doi: 10.1109/CVPR.2017.369.
44. [44] V. Chouhan et al., “A novel transfer learning based approach for pneumonia detection in chest X-ray images,” Appl. Sci., vol. 10, no. 2, 2020, doi: 10.3390/app10020559.
45. [45] M. Gazda, J. Plavka, J. Gazda, and P. Drotar, “Self-Supervised Deep Convolutional Neural Network for Chest X-Ray Classification,” IEEE Access, vol. 9, pp. 151972–151982, 2021, doi: 10.1109/ACCESS.2021.3125324.