Name
Title | ||
|---|---|---|
Human burn depth and grafting prognosis using ResNeXt topology based deep learning network |
Abstract | ||
|---|---|---|
The human burn diagnosis is becoming essential. Early diagnosis of a burn can save human life. Therefore, it is essential to develop a fast, robust, and efficient computer-based system for the diagnosis of burn. In recent days, the deep Convolutional Neural Network (CNN) model is set as a benchmark for medical image diagnosis. We have investigated ResNeXt, VGG16 and AlexNet on our human burn image dataset. We found that these model performance is not optimal. Therefore, in this work, we have designed a new model called BNeXt. The BNeXt CNN model kernel and convolutional layers are designed in such a way that it can discriminate the human burn efficiently. In previous studies, the burn has been classified into two broad categories that are graft and non-graft. The proposed study first determines the degree of a burn and later it classifies a burn into graft and non-graft. The manual burn diagnosis is time-consuming and accuracy varies from 75 to 80% by an expert doctor. The proposed model is efficient with an accuracy of 97.17% for burn degree and 99.67% for the grafting and non-grafting determination based on the depth of a burn. This model can be deployed on the cloud or the local system for the fast screening of burn patients. |
Year | DOI | Venue |
|---|---|---|
2022 | 10.1007/s11042-022-12555-2 | Multimedia Tools and Applications |
Keywords | DocType | Volume |
Deep learning, Classification, Burn, Diagnosis, Graft, CNN | Journal | 81 |
Issue | ISSN | Citations |
13 | 1380-7501 | 0 |
PageRank | References | Authors |
0.34 | 5 | 3 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| D. P. Yadav | 1 | 0 | 0.34 |
| Anand Singh Jalal | 2 | 138 | 28.45 |
| Ved Prakash | 3 | 0 | 0.34 |