IAPSO-AIRS: A novel improved machine learning-based system for wart disease treatment

Area of Science:

• Dermatology
• Medical Informatics
• Artificial Intelligence

Background:

• Wart disease (WD), caused by human papillomavirus (HPV), presents diagnostic challenges, particularly in evaluating treatment response.
• Traditional machine learning methods struggle with WD classification due to limited attributes.
• Manual assessment of treatment efficacy for common and plantar warts is difficult.

Purpose of the Study:

• To develop an evolutionary-based computer-aided diagnosis (CAD) system for classifying WD treatment response.
• To enhance the classification accuracy of WD using machine learning algorithms.
• To provide a reliable tool for clinical WD management.

Main Methods:

• Proposed a CAD system integrating the improved adaptive particle swarm optimization (IAPSO) algorithm with the artificial immune recognition system (AIRS).
• Utilized a database of 180 records from immunotherapy and cryotherapy treatments.
• Applied cross-validation with five partition protocols (K2, K3, K4, K5, K10) to evaluate the system's performance.

Main Results:

• The IAPSO-AIRS system achieved high performance metrics, including precision (0.8908), recall (0.8943), and F-measure (0.8916) using the K10 protocol.
• The system demonstrated an accuracy of 90% and a reliability of 98.68%.
• The best results were obtained with the K10 cross-validation protocol.

Conclusions:

• The developed IAPSO-AIRS system shows significant potential for accurate WD treatment response classification.
• This AI-driven approach can assist clinicians in managing wart disease effectively.
• The system's high reliability and accuracy suggest its applicability in clinical environments.