A Machine Learning Applied Diagnosis Method for Subcutaneous Cyst by Ultrasonography.

Area of Science:

• Medical imaging
• Machine learning
• Diagnostic systems

Background:

• Ultrasound imaging is vital for disease detection but requires extensive expertise.
• Computer-aided diagnosis systems aim to assist clinicians and reduce workload.
• Deep learning shows promise for enhancing ultrasonic computer-aided diagnosis.

Purpose of the Study:

• To develop and evaluate machine learning methods for diagnosing subcutaneous cysts.
• To compare the performance of different algorithms in ultrasound-based cyst diagnosis.

Main Methods:

• Extracted clinical features and ultrasonography images from 132 patients.
• Applied and compared four machine learning algorithms: decision tree (DT), support vector machine (SVM), K-nearest neighbors (KNN), and neural networks (NN).
• Utilized a 70% training and 30% testing data split.

Main Results:

• Support vector machine (SVM) demonstrated the highest accuracy, ranging from 91.7% to 100%.
• SVM outperformed other tested algorithms in diagnosing subcutaneous cysts via ultrasonography.

Conclusions:

• SVM is a highly accurate method for diagnosing subcutaneous cysts using ultrasound.
• This machine learning approach provides a valuable reference for clinical applications in subcutaneous cyst ultrasonography.