Exploratory Analysis of Deep Learning Model for Non Invasive Classification of Pulmonary Hypertension Based On Chest X-Ray Images
Abstract
Pulmonary Hypertension (PH) is a progressive condition in which early detection is essential for improving outcomes. Chest X-rays (CXR) may contain patterns associated with PH. This study evaluated automated PH detection from CXRs. A retrospective private dataset of 1,354 exams (1,138 PH) was analyzed. Multiple CNNs and Transformer-based architectures were trained and tested. CNNs achieved AUROC values from 0.71 to 0.76, outperforming Transformers (AUROC ≤ 0.60). Performance was mainly driven by PH cases with mPAP > 25 mmHg, while sensitivity decreased in normal and mildly elevated ranges. These results support the feasibility of non-invasive PH screening, although larger and more balanced datasets are needed for clinical validation.References
Becerra-Muñoz, V. M., Gómez Sáenz, J. T., and Escribano Subías, P. (2024). The importance of data in pulmonary arterial hypertension: From international registries to machine learning. Medicina Clínica (English Edition), 162(12):591–598.
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., and Houlsby, N. (2021). An image is worth 16x16 words: Transformers for image recognition at scale. In International Conference on Learning Representations, pages 1–21.
Fadilah, A., Putri, V. Y. S., Puling, I. M. D. R., and Willyanto, S. E. (2024). Assessing the precision of machine learning for diagnosing pulmonary arterial hypertension: a systematic review and meta-analysis of diagnostic accuracy studies. Frontiers in Cardiovascular Medicine, 11:1–17.
Ferreira, M. I. P., Oliveira, H. G., Krug, B. C., Gonçalves, C. B. T., Amaral, K. M., Mosca, M., Picon, P. D., M., R. R., and Schneiders, R. E. (2014). Hipertensão arterial pulmonar - protocolo clínico e diretrizes terapêuticas.
He, K., Zhang, X., Ren, S., and Sun, J. (2015). Deep residual learning for image recognition. arXiv, pages 1–12.
Huang, G., Liu, Z., Van Der Maaten, L., and Weinberger, K. Q. (2017). Densely connected convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4700–4708.
Humbert, M., Kovacs, G., Hoeper, M. M., Badagliacca, R., Berger, R. M. F., Brida, M., Carlsen, J., Coats, A. J. S., Escribano-Subias, P., Ferrari, P., Ferreira, D. S., Ghofrani, H. A., Giannakoulas, G., Kiely, D. G., Mayer, E., Meszaros, G., Nagavci, B., Olsson, K. M., Pepke-Zaba, J., Quint, J. K., Rådegran, G., Simonneau, G., Sitbon, O., Tonia, T., Toshner, M., Vachiery, J. L., Vonk Noordegraaf, A., Delcroix, M., Rosenkranz, S., and Group, E. S. D. (2022). 2022 esc/ers guidelines for the diagnosis and treatment of pulmonary hypertension: Developed by the task force for the diagnosis and treatment of pulmonary hypertension of the european society of cardiology (esc) and the european respiratory society (ers). endorsed by the international society for heart and lung transplantation (ishlt) and the european reference network on rare respiratory diseases (ern-lung). European Heart Journal, 43(38):3618–3731.
Imai, S., Sakao, S., Nagata, J., Naito, A., Sekine, A., Sugiura, T., Shigeta, A., Nishiyama, A., Yokota, H., Shimizu, N., Sugawara, T., Nomi, T., Honda, S., Ogaki, K., Tanabe, N., Baba, T., and Suzuki, T. (2024). Artificial intelligence-based model for predicting pulmonary arterial hypertension on chest x-ray images. BMC Pulmonary Medicine, 24:101.
Khou, V., Anderson, J. J., Strange, G., Corrigan, C., Collins, N., Celermajer, D. S., Dwyer, N., Feenstra, J., Horrigan, M., Keating, D., Kotlyar, E., Lavender, M., McWilliams, T. J., Steele, P., Weintraub, R., Whitford, H., Whyte, K., Williams, T. J., Wrobel, J. P., Keogh, A., and Lau, E. M. (2020). Diagnostic delay in pulmonary arterial hypertension: Insights from the australian and new zealand pulmonary hypertension registry. Respirology, 25:863–871.
Kusunose, K., Hirata, Y., Tsuji, T., Kotoku, J., and Sata, M. (2020). Deep learning to predict elevated pulmonary artery pressure in patients with suspected pulmonary hypertension using standard chest x ray. Scientific Reports, 10:19311.
Li, Z., Luo, G., Ji, Z., Wang, S., and Pan, S. (2024). Explanatory deep learning to predict elevated pulmonary artery pressure in children with ventricular septal defects using standard chest x-rays: a novel approach. Frontiers in Cardiovascular Medicine, Volume 11 - 2024.
Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., and Guo, B. (2021). Swin transformer: Hierarchical vision transformer using shifted windows. arXiv, pages 1–14.
Liu, Z., Mao, H., Wu, C.-Y., Feichtenhofer, C., Darrell, T., and Xie, S. (2022). A convnet for the 2020s. arXiv, pages 1–15.
Luna-López, R., Ruiz Martín, A., and Escribano Subías, P. (2022). Pulmonary arterial hypertension. Medicina Clínica (English Edition), 158(12):622–629.
Magalhães Junior, H. M. and Safatle, L. P. (2023). Hipertensão pulmonar - protocolo clínico e diretrizes terapêuticas.
McLaughlin, V. V., Shillington, A., and Rich, S. (2002). Survival in primary pulmonary hypertension. Circulation, 106(12):1477–1482.
Peña, E., Dennie, C., Veinot, J., and Muñiz, S. H. (2012). Pulmonary hypertension: How the radiologist can help. RadioGraphics, 32(1):9–32.
Ribeiro, E., Cardenas, D. A. C., Dias, F. M., Krieger, J. E., and Gutierrez, M. A. (2024). Explainable artificial intelligence in deep learning–based detection of aortic elongation on chest x-ray images. European Heart Journal - Digital Health, 5(5):524–534.
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., and Rabinovich, A. (2014). Going deeper with convolutions. arXiv, pages 1–12.
Tan, M. and Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning, pages 6105–6114. PMLR.
Zou, X.-L., Ren, Y., Feng, D.-Y., He, X.-Q., Guo, Y.-F., Yang, H.-L., Li, X., Fang, J., Li, Q., Ye, J.-J., Han, L.-Q., and Zhang, T.-T. (2020). A promising approach for screening pulmonary hypertension based on frontal chest radiographs using deep learning: A retrospective study. PLOS ONE, 15(7):1–13.
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., and Houlsby, N. (2021). An image is worth 16x16 words: Transformers for image recognition at scale. In International Conference on Learning Representations, pages 1–21.
Fadilah, A., Putri, V. Y. S., Puling, I. M. D. R., and Willyanto, S. E. (2024). Assessing the precision of machine learning for diagnosing pulmonary arterial hypertension: a systematic review and meta-analysis of diagnostic accuracy studies. Frontiers in Cardiovascular Medicine, 11:1–17.
Ferreira, M. I. P., Oliveira, H. G., Krug, B. C., Gonçalves, C. B. T., Amaral, K. M., Mosca, M., Picon, P. D., M., R. R., and Schneiders, R. E. (2014). Hipertensão arterial pulmonar - protocolo clínico e diretrizes terapêuticas.
He, K., Zhang, X., Ren, S., and Sun, J. (2015). Deep residual learning for image recognition. arXiv, pages 1–12.
Huang, G., Liu, Z., Van Der Maaten, L., and Weinberger, K. Q. (2017). Densely connected convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4700–4708.
Humbert, M., Kovacs, G., Hoeper, M. M., Badagliacca, R., Berger, R. M. F., Brida, M., Carlsen, J., Coats, A. J. S., Escribano-Subias, P., Ferrari, P., Ferreira, D. S., Ghofrani, H. A., Giannakoulas, G., Kiely, D. G., Mayer, E., Meszaros, G., Nagavci, B., Olsson, K. M., Pepke-Zaba, J., Quint, J. K., Rådegran, G., Simonneau, G., Sitbon, O., Tonia, T., Toshner, M., Vachiery, J. L., Vonk Noordegraaf, A., Delcroix, M., Rosenkranz, S., and Group, E. S. D. (2022). 2022 esc/ers guidelines for the diagnosis and treatment of pulmonary hypertension: Developed by the task force for the diagnosis and treatment of pulmonary hypertension of the european society of cardiology (esc) and the european respiratory society (ers). endorsed by the international society for heart and lung transplantation (ishlt) and the european reference network on rare respiratory diseases (ern-lung). European Heart Journal, 43(38):3618–3731.
Imai, S., Sakao, S., Nagata, J., Naito, A., Sekine, A., Sugiura, T., Shigeta, A., Nishiyama, A., Yokota, H., Shimizu, N., Sugawara, T., Nomi, T., Honda, S., Ogaki, K., Tanabe, N., Baba, T., and Suzuki, T. (2024). Artificial intelligence-based model for predicting pulmonary arterial hypertension on chest x-ray images. BMC Pulmonary Medicine, 24:101.
Khou, V., Anderson, J. J., Strange, G., Corrigan, C., Collins, N., Celermajer, D. S., Dwyer, N., Feenstra, J., Horrigan, M., Keating, D., Kotlyar, E., Lavender, M., McWilliams, T. J., Steele, P., Weintraub, R., Whitford, H., Whyte, K., Williams, T. J., Wrobel, J. P., Keogh, A., and Lau, E. M. (2020). Diagnostic delay in pulmonary arterial hypertension: Insights from the australian and new zealand pulmonary hypertension registry. Respirology, 25:863–871.
Kusunose, K., Hirata, Y., Tsuji, T., Kotoku, J., and Sata, M. (2020). Deep learning to predict elevated pulmonary artery pressure in patients with suspected pulmonary hypertension using standard chest x ray. Scientific Reports, 10:19311.
Li, Z., Luo, G., Ji, Z., Wang, S., and Pan, S. (2024). Explanatory deep learning to predict elevated pulmonary artery pressure in children with ventricular septal defects using standard chest x-rays: a novel approach. Frontiers in Cardiovascular Medicine, Volume 11 - 2024.
Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., and Guo, B. (2021). Swin transformer: Hierarchical vision transformer using shifted windows. arXiv, pages 1–14.
Liu, Z., Mao, H., Wu, C.-Y., Feichtenhofer, C., Darrell, T., and Xie, S. (2022). A convnet for the 2020s. arXiv, pages 1–15.
Luna-López, R., Ruiz Martín, A., and Escribano Subías, P. (2022). Pulmonary arterial hypertension. Medicina Clínica (English Edition), 158(12):622–629.
Magalhães Junior, H. M. and Safatle, L. P. (2023). Hipertensão pulmonar - protocolo clínico e diretrizes terapêuticas.
McLaughlin, V. V., Shillington, A., and Rich, S. (2002). Survival in primary pulmonary hypertension. Circulation, 106(12):1477–1482.
Peña, E., Dennie, C., Veinot, J., and Muñiz, S. H. (2012). Pulmonary hypertension: How the radiologist can help. RadioGraphics, 32(1):9–32.
Ribeiro, E., Cardenas, D. A. C., Dias, F. M., Krieger, J. E., and Gutierrez, M. A. (2024). Explainable artificial intelligence in deep learning–based detection of aortic elongation on chest x-ray images. European Heart Journal - Digital Health, 5(5):524–534.
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., and Rabinovich, A. (2014). Going deeper with convolutions. arXiv, pages 1–12.
Tan, M. and Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning, pages 6105–6114. PMLR.
Zou, X.-L., Ren, Y., Feng, D.-Y., He, X.-Q., Guo, Y.-F., Yang, H.-L., Li, X., Fang, J., Li, Q., Ye, J.-J., Han, L.-Q., and Zhang, T.-T. (2020). A promising approach for screening pulmonary hypertension based on frontal chest radiographs using deep learning: A retrospective study. PLOS ONE, 15(7):1–13.
Published
2026-06-01
How to Cite
RIBEIRO, Estela; CARVALHO, Gabriella G.; CARDENAS, Diego A. C.; SOUZA, Rogério de; GUTIERREZ, Marco A..
Exploratory Analysis of Deep Learning Model for Non Invasive Classification of Pulmonary Hypertension Based On Chest X-Ray Images. In: BRAZILIAN SYMPOSIUM ON COMPUTING APPLIED TO HEALTH (SBCAS), 26. , 2026, Ouro Preto/MG.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 241-252.
ISSN 2763-8952.
DOI: https://doi.org/10.5753/sbcas.2026.20671.
