Pharyngitis is a common condition affecting the respiratory system, often caused by viral or bacterial infections. It is associated with symptoms such as sore throat, redness, and swelling, significantly impacting individuals' quality of life. Given the similarity of symptoms among different types of pharyngitis, accurate and rapid identification of the infection type is critical, as misdiagnosis can lead to inappropriate treatment and more severe consequences.
This research aims to develop an efficient model for the automatic detection of pharyngitis types using mobile imaging technology and machine learning algorithms. The primary goal is to accurately distinguish between viral and bacterial pharyngitis, addressing the limitations of traditional diagnostic methods that require in-person examinations and costly, time-consuming tests.
To enhance the accuracy and speed of diagnosis, image processing and deep learning techniques are employed. The study utilizes the PGUPharyngitis dataset, comprising 742 high-quality images of inflamed throats captured using mobile phones. These images are analyzed and classified using neural network models and advanced machine learning techniques.
The results demonstrate that deep learning algorithms and image processing techniques are effective in detecting and categorizing various types of pharyngitis, delivering precise outcomes. Furthermore, leveraging mobile images as a diagnostic tool offers advantages such as high speed and reduced reliance on invasive testing.
This study highlights the innovative application of mobile imaging in diagnosing pharyngitis. The findings suggest that integrating mobile technology with machine learning can provide a more efficient method for disease diagnosis, ultimately improving the quality of healthcare in this domain.
Keywords:automatic diagnosis, deep learning, neural networks, machine learning, pharyngitis, bacterial, viral, mobile imaging