Smartpath^k: a platform for teaching glomerulopathies using machine learning

Area of Science:

• Medical Education
• Digital Pathology
• Artificial Intelligence in Medicine

Background:

• The COVID-19 pandemic accelerated the adoption of digital learning in medical education.
• Pathology teaching is adapting traditional microscopy with computational tools and digitized images.
• Machine learning algorithms are being developed to identify histological patterns in kidney biopsies for pathology diagnosis.

Purpose of the Study:

• To describe and evaluate SmartPath^k, an AI-powered tool for teaching glomerulopathies.
• To leverage machine learning for creating computational models for renal pathology identification.
• To support medical education by providing accessible training in nephropathology.

Main Methods:

• Developed SmartPath^k, an intelligent system utilizing machine learning.
• Employed the J48 algorithm for automatic knowledge acquisition and decision tree model creation.
• Trained the system on digitized kidney biopsy slides to recognize histological patterns.

Main Results:

• SmartPath^k demonstrated 89.47% accuracy in identifying renal pathologies.
• The system functions as a complementary remote tool for pathology teaching and learning.
• The machine learning algorithms were based on decision trees.

Conclusions:

• The AI system, SmartPath^k, effectively assists in teaching renal pathology.
• It has the potential to increase the training capacity of medical professionals in nephropathology.
• This tool offers a solution for quality medical training, even with limited access to expert nephropathologists.