Avaliação de Dor em Expressão Facial Neonatal por meio de Redes Neurais Profundas

• Lucas F. Buzuti Centro Universitário FEI
• Carlos E. Thomaz Centro Universitário FEI

DOI: https://doi.org/10.5753/sbcas.2021.16099

Abstract

Neonatal pain assessment might suffer variation among health professionals, leading to late intervention and flimsy treatment of pain in several occasions. In this context, the goal of this dissertation was to assess quantitatively and qualitatively models of Convolutional Neural Networks in the neonatal pain classification task using face images of two distinct databases (an international, named COPE, and other national, named UNIFESP). Our quantitative results showed the top performance of N-CNN on neonatal pain classification, with average accuracy of 87.2% and 78.7% for the databases COPE and UNIFESP, respectively. However, the quantitative analysis indicated that all neural models evaluated, including the N-CNN model, might learn artifacts from the imagens and not proper discriminating information, suggesting further research to apply such models in clinical practice.

References

Anand, K. J. and Carr, D. B. (1989). The neuroanatomy, neurophysiology, and neurochemistry of pain, stress, and analgesia in newborns and children. Pediatric Clinics of North America, 36(4):795–822.

Bhutta, A. T. and Anand, K. (2002). Vulnerability of the developing brain: neuronal mechanisms. Clinics in perinatology, 29(3):357–372.

Brahnam, S., Chuang, C.-F., Shih, F. Y., and Slack, M. R. (2006). Machine recognition and representation of neonatal facial displays of acute pain. Articial intelligence in medicine, 36(3):211–222.

Brummelte, S., Grunau, R. E., Chau, V., Poskitt, K. J., Brant, R., Vinall, J., Gover, A., Synnes, A. R., and Miller, S. P. (2012). Procedural pain and brain development in premature newborns. Annals of neurology, 71(3):385–396.

Deng, J., Guo, J., Zhou, Y., Yu, J., Kotsia, I., and Zafeiriou, S. (2019). Retinaface: Single stage dense face localisation in the wild. arXiv preprint arXiv:1905.00641.

Grunau, R. E., Tu, M. T., Whiteld, M. F., Oberlander, T. F., Weinberg, J., Yu, W., Thiessen, P., Gosse, G., and Scheifele, D. (2010). Cortisol, behavior, and heart rate reactivity to immunization pain at 4 months corrected age in infants born very preterm. The Clinical journal of pain, 26(8):698.

Grunau, R. E., Weinberg, J., and Whiteld, M. F. (2004). Neonatal procedural pain and preterm infant cortisol response to novelty at 8 months. Pediatrics, 114(1):e77–e84.

He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image reIn Proceedings of the IEEE conference on computer vision and pattern cognition. recognition, pages 770–778.

Heiderich, T. M., Leslie, A. T. F. S., and Guinsburg, R. (2015). Neonatal procedural pain can be assessed by computer software that has good sensitivity and specicity to detect facial movements. Acta Paediatrica, 104(2):e63–e69.

Selvaraju, R. R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., and Batra, D. (2017). Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE international conference on computer vision, pages 618–626.

Vinall, J., Miller, S. P., Chau, V., Brummelte, S., Synnes, A. R., and Grunau, R. E. (2012). Neonatal pain in relation to postnatal growth in infants born very preterm. Pain, 153(7):1374–1381.

Walker, S. M. (2017). Translational studies identify long-term impact of prior neonatal pain experience. Pain, 158:S29–S42.

Zamzmi, G., Paul, R., Goldgof, D., Kasturi, R., and Sun, Y. (2019). Pain assessment from facial expression: Neonatal convolutional neural network (n-cnn). In 2019 International Joint Conference on Neural Networks (IJCNN), pages 1–7. IEEE.