RESULTS OF THE USE OF ARTIFICIAL INTELLIGENCE IN DIAGNOSTICS OF PARONASAL SINUS DISEASES
Keywords:
cone-beam computed tomography; artificial intelligence; Teachable Machine; convolutional neural networks; diseases of the paranasal sinusesAbstract
In modern otorhinolaryngology, radiological methods, in particular, cone-beam computed tomography, are the gold standard for diagnosing diseases of the paranasal sinuses. Materials and methods. The study is based on a dataset consisting of 2155 X-ray images in the coronal projection, extracted from 3D cone-beam computed tomography scans. Results. First test. The results of the first stage showed limited classification accuracy. The average accuracy was 55%, which is insufficient for clinical use, as the model often confused similar types of sinusitis (e.g., purulent maxillary sinusitis and maxillary ethmoiditis) due to insufficient data. Conclusion. The study showed that simple and accessible artificial intelligence tools can be effective for solving specific diagnostic tasks.
References
Litjens, G., Kooi, T., Bejnordi, N. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., ... & Sánchez, C. I. (2017). A survey on deep learning in medical image analysis. Medical image analysis, 42, 60-88.
Topol, E. J. (2019). High-performance medicine: the convergence of human and artificial intelligence. Nature medicine, 25(1), 44-56.
Liu, J., Li, Y., Wang, X., & Chen, G. (2021). Deep learning for diagnosis and prognosis of head and neck cancer. Journal of Medical Imaging, 8(3), 031301.
