由於世界各地的老年人口不斷增加，心臟衰竭（HF）是一個全球性的流行病。胸部X 光(CXR) 檢查是最常見的非侵入式檢查，可以提供病人心臟的主要信息。胸部X 光檢查可以提供有關病人心臟狀況的關鍵訊息。

本研究的目的是評估在全球範圍內，用一種簡單而經濟的方法，提供有關病人心臟狀況的關鍵訊息。本研究旨在評估人工智能（AI）在從CXR 診斷HF 方面的可能性。

我們提出專門為CXR 圖像建立的前處理系統流程，改善心臟衰竭模型訓練的成效。並分析不同大小的數據集、不同模型結構對效能的影響。數據集包括1,840 張HF 的CXR 圖像和2,100 張非HF 的圖像。準確率、精確度、召回率和F1 得分也相互比較，以獲得最適合從CXR 圖像識別HF 的模型。

在這些模型中，與其他模型（VGG16、EfficientNetB3、ResNet18、MnasNet、Inception-V3 和Resnext50）相比，Inception-ResNet-v2 是診斷CXR 圖像中HF 的最佳模型（準確度91.0%，精確度94.9%，召回率87.9%，F1 分數91.2%）。

在心血管疾病中人工智慧是一個在HF 的管理中的新興應用。我們提供了最合適的CNN 模型，用於從CXR 圖像中識別HF。我們的發現可以提供適當的診斷和治療HF。

Heart failure (HF) is a global epidemic due to the increasing number of elderly people around the world. A chest x-ray(CXR) is the most common non-invasive test that can provide key information about a patient's heart. The test can provide key information about the patient's heart condition. The purpose of this study is to evaluate a simple and cost-effective method to provide key information about a patient's heart condition on a global scale. The purpose of this study is to evaluate the possibility of artificial intelligence (AI) in diagnosing HF from CXR.

We proposed a preprocessing system process built specifically for chest X-ray images
to improve the performance of HF model training. We also analyzed the impact of different size datasets and different model structures on the performance. The dataset consists of 1,840 CXR images of HF and 2,100 images of non-HF. Accuracy, precision, recall, F1 score and Class activation Map(Class Activation Map) were also compared with each other to obtain the most suitable model for recognizing HF from CXR images.

Among these models, Inception-ResNet-v2 was the best model for diagnosing HF in CXR images compared with other models (VGG16, EfficientNetB3, ResNet18, MnasNet, Inception-V3, and Resnext50) (91.0% accuracy, 94.9% precision , 87.9% recall, and
91.2% F1 score).

An emerging application of artificial intelligence in cardiovascular disease is in the management of HF. We provide the most suitable CNN model for identifying HF cation from CXR images. Our findings can provide appropriate diagnosis and treatment of HF.

[1] W. H. Organization, “The top 10 causes of death,” 2020.
[2] T. Matsumoto, S. Kodera, H. Shinohara, H. Ieki, T. Yamaguchi, Y. Higashikuni, A. Kiyosue, K. Ito, J. Ando, E. Takimoto, et al., “Diagnosing heart failure from chest x-ray images using deep learning,” International Heart Journal, vol. 61, no. 4, pp. 781–786, 2020.
[3] D. Pan, P. Pellicori, K. Dobbs, J. Bulemfu, I. Sokoreli, A. Urbinati, O. Brown, S. Sze, A. S. Rigby, S. Kazmi, et al., “Prognostic value of the chest x-ray in patients hospitalised for heart failure,” Clinical Research in Cardiology, vol. 110, no. 11, pp. 1743–1756, 2021.
[4] A. Hosny, C. Parmar, J. Quackenbush, L. H. Schwartz, and H. J. Aerts, “Artificial intelligence in radiology,” Nature Reviews Cancer, vol. 18, no. 8, pp. 500–510, 2018.
[5] N. Nasrullah, J. Sang, M. S. Alam, M. Mateen, B. Cai, and H. Hu, “Automated lung nodule detection and classification using deep learning combined with multiple strategies,” Sensors, vol. 19, no. 17, p. 3722, 2019.
[6] 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 journal of gastroenterology, vol. 25, no. 6, p. 672, 2019.
[7] T. Magadza and S. Viriri, “Deep learning for brain tumor segmentation: a survey of state-of-the-art,” Journal of Imaging, vol. 7, no. 2, p. 19, 2021.
[8] C. Hassan, M. Spadaccini, A. Iannone, R. Maselli, M. Jovani, V. T. Chandrasekar, G. Antonelli, H. Yu, M. Areia, M. Dinis-Ribeiro, et al., “Performance of artificial intelligence in colonoscopy for adenoma and polyp detection: a systematic review and meta-analysis,” Gastrointestinal endoscopy, vol. 93, no. 1, pp. 77–85, 2021.
[9] A. L. Bui, T. B. Horwich, and G. C. Fonarow, “Epidemiology and risk profile of heart failure,” Nature Reviews Cardiology, vol. 8, no. 1, pp. 30–41, 2011.
[10] A. Javeed, S. U. Khan, L. Ali, S. Ali, Y. Imrana, and A. Rahman, “Machine learning-based automated diagnostic systems developed for heart failure prediction using different types of data modalities: a systematic review and future directions,” Computational and Mathematical Methods in Medicine, vol. 2022, 2022.
[11] A. Groenewegen, F. H. Rutten, A. Mosterd, and A. W. Hoes, “Epidemiology of heart failure,” European journal of heart failure, vol. 22, no. 8, pp. 1342–1356, 2020.
[12] T. Koga and K. Fujimoto, “Images in clinical medicine. kerley’s a, b, and c lines,” The New England journal of medicine, vol. 360, no. 15, p. 1539, 2009.
[13] Y. Udovychenko, A. Popov, and I. Chaikovsky, “Multistage classification of current density distribution maps of various heart states based on correlation analysis and k-nn algorithm,” Frontiers in Medical Technology, vol. 3, 2021.
[14] R. Arsanjani, D. Dey, T. Khachatryan, A. Shalev, S. W. Hayes, M. Fish, R. Nakanishi, G. Germano, D. S. Berman, and P. Slomka, “Prediction of revascularization after myocardial perfusion spect by machine learning in a large population,” Journal of Nuclear Cardiology, vol. 22, no. 5, pp. 877–884, 2015.
[15] L. Šajn and M. Kukar, “Image processing and machine learning for fully automated probabilistic evaluation of medical images,” Computer methods and programs in biomedicine, vol. 104, no. 3, pp. e75–e86, 2011.
[16] T. Matsumoto, S. Kodera, H. Shinohara, H. Ieki, T. Yamaguchi, Y. Higashikuni, A. Kiyosue, K. Ito, J. Ando, E. Takimoto, et al., “Diagnosing heart failure from chest x-ray images using deep learning,” International Heart Journal, vol. 61, no. 4, pp. 781–786, 2020.
[17] Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proceedings of the IEEE, vol. 86, no. 11, pp. 2278–2324, 1998.
[18] K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” arXiv preprint arXiv:1409.1556, 2014.
[19] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” pp. 770–778, 2016.
[20] S. Xie, R. Girshick, P. Dollár, Z. Tu, and K. He, “Aggregated residual transformations for deep neural networks,” pp. 1492–1500, 2017.
[21] C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, and A. Rabinovich, “Going deeper with convolutions,” pp. 1–9, 2015.
[22] C. Szegedy, S. Ioffe, V. Vanhoucke, and A. A. Alemi, “Inception-v4, inception-resnet and the impact of residual connections on learning,” 2017.
[23] M. Tan, B. Chen, R. Pang, V. Vasudevan, M. Sandler, A. Howard, and Q. V. Le, “Mnasnet: Platformaware neural architecture search for mobile,” pp. 2820–2828, 2019.
[24] M. Tan and Q. Le, “Efficientnet: Rethinking model scaling for convolutional neural networks,” pp. 6105–6114, 2019.
[25] S. M. Pizer, E. P. Amburn, J. D. Austin, R. Cromartie, A. Geselowitz, T. Greer, B. ter Haar Romeny, J. B. Zimmerman, and K. Zuiderveld, “Adaptive histogram equalization and its variations,” Computer vision, graphics, and image processing, vol. 39, no. 3, pp. 355–368, 1987.
[26] “Covid-19 radiography database,” 2020.
[27] T. A. McDonagh, M. Metra, M. Adamo, R. S. Gardner, A. Baumbach, M. Böhm, H. Burri, J. Butler, J. Čelutkienė, O. Chioncel, et al., “2021 esc guidelines for the diagnosis and treatment of acute and chronic heart failure: Developed by the task force for the diagnosis and treatment of acute and chronic heart failure of the european society of cardiology (esc) with the special contribution of the heart failure association (hfa) of the esc,” European heart journal, vol. 42, no. 36, pp. 3599–3726, 2021.
[28] B. Zhou, A. Khosla, A. Lapedriza, A. Oliva, and A. Torralba, “Learning deep features for discriminative localization,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 2921–2929, 2016.
[29] R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Grad-cam: Visual explanations from deep networks via gradient-based localization,” in Proceedings of the IEEE international conference on computer vision, pp. 618–626, 2017.