Canine Thoracic Radiographs Classification Using Deep Learning Algorithms: An Investigation

  • Ashendra Kumar Saxena Professor, College of Computing Scinece and Information Technology, Teerthanker Mahaveer University, Moradabad, Uttar Pradesh, India
  • Divakara Sastry EV Professor, School of Agricultural Sciences, Jaipur National University, Jaipur, Rajasthan, India
  • Roopashree Assistant Professor, Department of Chemistry, School of Sciences, JAIN (Deemed-to-be University), Karnataka, India
Keywords: DenseNet-121, ResNet-50, Enhanced Layer wise deep neural Networks (EL-DNN), and canine thoracic radiographs (CTR)


Thoracic radiograph interpretation is a difficult and error-prone job for veterinarians. Even with recent developments in machine learning and computer vision, creating computer-aided diagnostic tools for radiographs is still a difficult and unresolved challenge, especially in veterinary medicine. This research aimed to develop a unique approach for categorizing canine thoracic radiographs (CTR) using Enhanced Layer wise deep neural Networks (EL-DNN). Thoracic radiographs of canine patients were collected retrospectively from 2010 to 2020. The radiograph data was split in half because it came from two distinct radiograph acquisition methods. The EL-DNNs' generalizability was evaluated using Data Set 2, whereas Data Set 1 was utilized for training and testing. We built and evaluated two alternative EL-DNNs, one using the ResNet-50 architecture and the other using the DenseNet-121.  The area under the Receiver Operator Curve (AUC) values over 0.8 were achieved by the ResNet-50-based EL-DNN for all included radiographic findings on Data Sets 1 and 2, except bronchial and interstitial patterns. The overall performance of the DenseNet-121 EL-DNN was inferior. The EL-DNN trained on ResNet-50 outperformed the other regarding generalization ability, demonstrating superior performance for the alveolar, megaesophagus, interstitial, and pneumothorax.


Rho, J., Shin, S. M., Jhang, K., Lee, G., Song, K. H., Shin, H., ... & Son, H. Y. (2023). Deep learning-based diagnosis of feline hypertrophic cardiomyopathy. Plos one, 18(2), e0280438.

Yoon, Y., Hwang, T., Choi, H., & Lee, H. (2019). Classification of radiographic lung pattern based on texture analysis and machine learning. Journal of Veterinary Science, 20(4).

Banzato, T., Wodzinski, M., Tauceri, F., Donà, C., Scavazza, F., Müller, H., & Zotti, A. (2021). An AI-based algorithm for the automatic classification of thoracic radiographs in cats. Frontiers in veterinary science, 8, 731936.

Pereira, A. I., Franco-Gonçalo, P., Leite, P., Ribeiro, A., Alves-Pimenta, M. S., Colaço, B., ... & Ginja, M. (2023). Artificial Intelligence in Veterinary Imaging: An Overview. Veterinary Sciences, 10(5), 320.

Celniak, W., Wodziński, M., Jurgas, A., Burti, S., Zotti, A., Atzori, M., ... & Banzato, T. (2023). Inter-species and inter-pathology self-supervised pre-training of deep learning models: a resource to improve the classification of veterinary thoracic radiographs.

Boissady, E., de La Comble, A., Zhu, X., & Hespel, A. M. (2020). Artificial intelligence evaluating primary thoracic lesions has an overall lower error rate compared to veterinarians or veterinarians in conjunction with the artificial intelligence. Veterinary Radiology & Ultrasound, 61(6), 619-627.

Basran, P. S., & Appleby, R. B. (2022). The unmet potential of artificial intelligence in veterinary medicine. American Journal of Veterinary Research, 83(5), 385-392.

Hennessey, E., DiFazio, M., Hennessey, R., & Cassel, N. (2022). Artificial intelligence in veterinary diagnostic imaging: A literature review. Veterinary Radiology & Ultrasound.

Ott, J., Bruyette, D., Arbuckle, C., Balsz, D., Hecht, S., Shubitz, L., & Baldi, P. (2021). Detecting pulmonary Coccidioidomycosis with deep convolutional neural networks. Machine Learning with Applications, 5, 100040.

Gupta, S., & Blankstein, R. (2021). Detecting Coronary Artery Calcium on Chest Radiographs: Can We Teach an Old Dog New Tricks? Radiology: Cardiothoracic Imaging, 3(3), e210123.

Adrien‐Maxence, H., Emilie, B., Alois, D. L. C., Michelle, A., Kate, A., Mylene, A., ... & Federica, M. (2022). Comparison of error rates between four pre trained DenseNet convolutional neural network models and 13 board‐certified veterinary radiologists when evaluating 15 labels of canine thoracic radiographs. Veterinary Radiology & Ultrasound, 63(4), 456-468.

Giannasi, C., Rushton, S., Rook, A., Steen, N. V. D., Venier, F., Ward, P., ... & Roberts, E. (2021). Canine thyroid carcinoma prognosis following the utilization of computed tomography-assisted staging. Veterinary Record, 189(1), no-no.

Pinto, I. I. R. (2019). Comparison of Heart Measurements in Thoracic Radiographs Before and After the Treatment of Pulmonary Edema in Dogs with Degenerative Mitral Valve Disease: A Retrospective Study of 18 Clinical Cases (Doctoral dissertation, Universidade de Lisboa (Portugal)).

Škor, O., Bicanová, L., Wolfesberger, B., Fuchs‐Baumgartinger, A., Ruetgen, B., Štěrbová, M., ... & Kleiter, M. (2021). Are B‐symptoms more reliable prognostic indicators than substage in canine nodal diffuse large B‐cell lymphoma. Veterinary and Comparative Oncology, 19(1), 201-208.

Sukut, S. L., D'Eon, M., Lawson, J., & Mayer, M. N. (2023). Providing comparison normal examples alongside pathologic thoracic radiographic cases can improve veterinary students’ ability to identify abnormal findings or diagnose disease. Veterinary Radiology & Ultrasound.

Burti, S., Osti, V. L., Zotti, A., & Banzato, T. (2020). Use of deep learning to detect cardiomegaly on thoracic radiographs in dogs. The Veterinary Journal, 262, 105505.

Li, S., Wang, Z., Visser, L. C., Wisner, E. R., & Cheng, H. (2020). Pilot study: application of artificial intelligence for detecting left atrial enlargement on canine thoracic radiographs. Veterinary radiology & ultrasound, 61(6), 611-618.

Tahghighi, P., Appleby, R. B., Norena, N., Ukwatta, E., & Komeili, A. (2023). Machine learning can appropriately classify the collimation of ventrodorsal and dorsoventral thoracic radiographic images of dogs and cats. American Journal of Veterinary Research, 1(aop), 1-8.

Dumortier, L., Guépin, F., Delignette-Muller, M. L., Boulocher, C., & Grenier, T. (2022). Deep learning in veterinary medicine, an approach based on CNN to detect pulmonary abnormalities from lateral thoracic radiographs in cats. Scientific Reports, 12(1), 11418.

Müller, T. R., Solano, M., & Tsunemi, M. H. (2022). Accuracy of artificial intelligence software for the detection of confirmed pleural effusion in thoracic radiographs in dogs. Veterinary Radiology & Ultrasound, 63(5), 573-579.

Jeong, Y., & Sung, J. (2022). An automated deep learning method and novel cardiac index to detect canine cardiomegaly from simple radiography. Scientific Reports, 12(1), 14494.

Hennessey, R. (2022). Using Machine Learning to Detect Hanging Errors in Canine Thoracic Radiographs.

Kim, E., Fischetti, A. J., Sreetharan, P., Weltman, J. G., & Fox, P. R. (2022). Comparison of artificial intelligence to the veterinary radiologist's diagnosis of canine cardiogenic pulmonary edema. Veterinary Radiology & Ultrasound, 63(3), 292-297.

Tonima, M. A., Esfahani, F., DeHart, A., & Zhang, Y. (2021, May). A lightweight combinational machine learning algorithm for sorting canine torso radiographs. In 2021 4th IEEE International Conference on Industrial Cyber-Physical Systems (ICPS) (pp. 347-352). IEEE.

Banzato, T., Wodzinski, M., Burti, S., Vettore, E., Muller, H., & Zotti, A. (2023). An AI-based algorithm for the automatic evaluation of image quality in canine thoracic radiographs.

How to Cite
Ashendra Kumar Saxena, Divakara Sastry EV, & Roopashree. (2023). Canine Thoracic Radiographs Classification Using Deep Learning Algorithms: An Investigation. Revista Electronica De Veterinaria, 24(2), 436 - 448. Retrieved from