As one of the main causes of casualties in the world, fires, burns and scalds result in serious social and economic burdens in the world, and the annual incidence of burns in China exceeds 1%. However, the accuracy of surgeons in judging burn depth is quite low at present, with less than 50% accuracy of inexperienced doctors and only 64% to 76% accuracy of experienced doctors. For these reasons, there is great potential according to the studies on burn diagnosis based on deep learning algorithm. Therefore, in this study, a variety of convolutional neural networks were used to compare the classification of burn degree, and the processed datasets were used for training. By testing a series of neural networks, the best result was the precision of 88.75% when using the pre-trained VGG16 network. This result shows that the deep learning algorithm has high accuracy and application potential in the field of burn diagnosis.

Bohan Wei. Comparative Study on the Application of Deep Learning Algorithm in Burn Depth Diagnosis. Academic Journal of Computing & Information Science (2024), Vol. 7, Issue 4: 110-117. https://doi.org/10.25236/AJCIS.2024.070415.

[1] Jiao C, Su K, Xie W, et al. Burn image segmentation based on mask regions with convolutional neural network deep learning framework: more accurate and more convenient[J/OL]. Burns Trauma, 2019, 7. DOI: 10.1186/s41038-018-0137-9.

[2] Ning W, Qi F, Wang J. Terahertz technology utilized to achieve the assessment of burn injuries[J/OL]. China Medical Devices, 2018, 33: 14-16. DOI: 10.3969/ j.issn.1674-1633.2018.07.003.

[3] Huang Y, Qiu L, Mei A L, et al. Meta-analysis on the diagnostic value of laser doppler imaging for burn depth [J/OL]. Chinese Journal of Burns, 2017, 33: 301 – 308[2023-10-14]. https: //pubmed. ncbi.nlm. nih.gov/28651422/. DOI: 10.3760/cma.j.issn.1009-2587.2017.05.009.

[4] Han X, Liu Y, He G, et al. Skin burn wound classification algorithm based onmulti-scale feature fusion[J/OL]. Electronic Measurement Technology, 2022, 45: 114 – 118. DOI: 10.19651/j.cnki. emt. 2209382.

[5] Yadav D P. A method for human burn diagnosis using machine learning and slic superpixels based segmentation [J/OL]. IOP Conference Series: Materials Science and Engineering, 2021, 1116: 012186. DOI: 10.1088/1757-899x/1116/1/012186.

[6] Yadav D P, Sharma A, Singh M, et al. Feature extraction based machine learning for human burn diagnosis from burn images[J/OL]. IEEE Journal of Translational Engineering in Health and Medicine, 2019, 7: 1-7. https://dx.doi.o rg/10.1109%2FJTEHM.2019.2923628. DOI: 10.1109/jtehm. 2019. 2923628.

[7] Rowland R A, Ponticorvo A, BALDADO M L, et al. Burn wound clas-sification model using spatial frequency-domain imaging and machine learning [J/OL]. Journal of Biomedical Optics, 2019, 24. DOI: 10.1117/1.jbo.24.5.056007.

[8] Rowland R A, Ponticorvo A, BALDADO M L, et al. A simple burn wound severity assessment classifier based on spatial frequency domain imaging (sfdi) and machine learning[J/OL]. International Society for Optics and Photonics, 2019, 10851: 1085109. DOI: 10.1117/12.2510670.