Screening and Diagnosis of Chronic Pharyngitis Based on Deep Learning

Area of Science:

• Artificial Intelligence in Medicine
• Otolaryngology
• Speech Pathology

Background:

• Chronic pharyngitis presents diagnostic challenges due to symptom overlap with conditions like chronic tonsillitis.
• Current diagnostic methods can be costly and prone to misdiagnosis, necessitating improved approaches.
• Speech disorders are a recognized symptom in chronic pharyngitis patients.

Purpose of the Study:

• To develop and evaluate a novel, cost-effective, and rapid diagnostic method for chronic pharyngitis.
• To leverage artificial intelligence, specifically convolutional neural networks, for disease diagnosis.
• To utilize speech disorder symptoms as a basis for AI-driven diagnosis.

Main Methods:

• Voice data from patients was converted into speech spectrograms to extract characteristic information.
• A deep convolutional neural network (CNN) architecture was designed and implemented for diagnostic classification.
• Parameter optimization experiments were conducted to enhance the CNN's recognition efficiency.

Main Results:

• The developed CNN model demonstrated a high recognition rate for chronic pharyngitis.
• The AI-based approach showed a significant diagnostic effect, indicating its potential clinical utility.
• Speech spectrogram analysis proved effective for computer-aided diagnosis of chronic pharyngitis.

Conclusions:

• Convolutional neural networks offer a promising tool for the accurate and efficient diagnosis of chronic pharyngitis.
• AI-powered analysis of speech disorders presents a viable alternative to traditional diagnostic methods.
• This research supports the development of accessible and rapid diagnostic solutions for chronic pharyngitis.