Conclusion

Conclusion

We propose an AI-based labeler for label extraction from the unstructured medical reports and compare its performance to the CheXpert Labeler. We propose a novel model architecture CARES which performs multi-label classification and report generation for a given chest x-ray image and compare its performance with the SOTA. We propose a metric RFQI for measuring the accuracy of the reports generated using AI models. For the chest x-ray report generation, the key objective is to identify the correct radiological finding. Since the other metrics (BLEU-4, CIDEr, etc.) give equal weightage to each word in the sentence, RFQI tends to be more clinically relevant since it tries to match only the radiological finding and not the entire sentence.

Refrences

[1] Irvin, J., Rajpurkar, P., Ko, M., Yu, Y., Ciurea-Ilcus, S., Chute, C., Marklund, H., Haghgoo, B., Ball, R., Shpanskaya, K., Seekins, J., Mong, D., Halabi, S., Sandberg, J., Jones, R., Larson, D., Langlotz, C., Patel, B., Lungren, M. and Ng, A., 2020. Chexpert: A Large Chest Radiograph Dataset With Uncertainty Labels And Expert Comparison. [online] arXiv.org. Available at: https://arxiv.org/abs/1901.07031

[2] Smit, A., Jain, S., Rajpurkar, P., Pareek, A., Ng, A. and Lungren, M., 2020. Chexbert: Combining Automatic Labelers And Expert Annotations For Accurate Radiology Report Labeling Using BERT. [online] arXiv.org. Available at: https://arxiv.org/abs/2004.09167

[3] McDermott, M., Hsu, T., Weng, W., Ghassemi, M. and Szolovits, P., 2020. Chexpert++: Approximating The Chexpert Labeler For Speed,Differentiability, And Probabilistic Output. [online] arXiv.org. Available at: https://arxiv.org/abs/2006.15229

[4] GitHub. 2020. Jfhealthcare/Chexpert. [online] Available at: https://github.com/jfhealthcare/Chexpert.

[5] Pham, H., Le, T., Tran, D., Ngo, D. and Nguyen, H., 2020. Interpreting Chest X-Rays Via Cnns That Exploit Hierarchical Disease Dependencies And Uncertainty Labels. [online] arXiv.org. Available at: https://arxiv.org/abs/1911.06475.

[6] Xu, K., Ba, J., Kiros, R., Cho, K., Courville, A., Salakhutdinov, R., Zemel, R. and Bengio, Y., 2020. Show, Attend And Tell: Neural Image Caption Generation With Visual Attention. [online] arXiv.org. Available at: https://arxiv.org/abs/1502.03044

[7] Jing, B.; Xie, P.; Xing, E.; and Xing, E. 2017. On the automatic generation of medical imaging reports. arXiv preprint arXiv:1711.08195.

[8] Wang, X.; Peng, Y.; Lu, L.; Lu, Z.; and Summers, R. M. 2018. Tienet: Text-image embedding network for common thorax disease classification and reporting in chest x-rays. In Proceedings of the IEEE conference on CVPR, 9049–9058.

[9] Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. arXiv:1810.04805 [cs], October 2018. URL http://arxiv.org/abs/1810.04805. arXiv: 1810.04805.

[10] Putha, P., Tadepalli, M., Reddy, B., Raj, T., Chiramal, J., Govil, S., Sinha, N., KS, M., Reddivari, S., Jagirdar, A., Rao, P. and Warier, P., 2020. Can Artificial Intelligence Reliably Report Chest X-Rays?: Radiologist Validation Of An Algorithm Trained On 2.3 Million X-Rays. [online] arXiv.org. Available at: https://arxiv.org/abs/1807.07455.

[11] Y. Zhang, X. Wang, Z. Xu, Q. Yu, A. Yuille, and D. Xu, “When radiology report generation meets knowledge graph,” arXiv preprint arXiv:2002.08277, 2020.

[12] Li, C. Y., Liang, X., Hu, Z., & Xing, E. P. (2019). Knowledge-driven encode, retrieve, paraphrase for medical image report generation. In The Thirty-Third AAAI Conference on Artificial Intelligence, AAAI 2019, The Thirty-First Innovative Applications of Artificial Intelligence Conference, IAAI 2019, The Ninth AAAI Symposium on Educational Advances in Artificial Intelligence, EAAI 2019, Honolulu, Hawaii, USA, January 27 - February 1, 2019, pp. 6666–6673. AAAI Press.