Artificial intelligence useful in accurately predicting survival in larynx squamous cell carcinoma

The review found that

• The average concordance indices of survival period predictions were 0.747 ± 0.009 from COX-PH, 0.596 ± 0.015 from RSF, 0.893 ± 0.017 from DNN regression, and 0.859 ± 0.018 from DNN multi-classification.
• The average mortality was 0.682 ± 0.055 with DNN regression and 0.841 ± 0.020 with DNN multi-classification.
• Using a DNN with multi-classification showed significantly better prediction performance than a DNN model with only T/N staging, shown to be an appropriate method for survival prediction.

From the findings collected, the authors concluded that AI technology helps to improve the predictive performance of survival models in LSCC and is more accurate in survival prediction than conventional models. Therefore, AI seems to be concretely detecting the spread of various cancer in the body, predicting survival more accurately and improving oncologic outcomes.

Reference:

Choi, N., Kim, J., Yi, H. et al. The use of artificial intelligence models to predict survival in patients with laryngeal squamous cell carcinoma. Sci Rep 13, 9734 (2023).

https://doi.org/10.1038/s41598-023-35627-1