Sign in Subscribe
Artificial intelligence

AI Models Improve Pathological Diagnoses Efficiency with Machine Learning

Strategic Insights

— 1 min read

Researchers have developed an innovative artificial intelligence (AI) model for the classification of colon polyps using AutoML algorithms from cloud-based machine learning platforms. The AI model demonstrated high accuracy, identifying tubular adenoma and hyperplastic polyps with 100% success and normal colon with 97% success. This study highlights the potential of AI models in improving diagnostic efficiency in digital pathology.

Key Takeaways:

  • Researchers developed an AI model for the classification of colon polyps using AutoML algorithms from Google's VertexAI platform.
  • The AI model was trained on whole-slide images from public and institutional databases, and it achieved a high accuracy rate in identifying various types of colon polyps.
  • The model displayed 100% accuracy in identifying tubular adenoma and hyperplastic polyps, and 97% accuracy in identifying normal colon.
  • The study demonstrates the potential of AI models in improving diagnostic efficiency in digital pathology.
  • The AI model was developed using an AutoML algorithm, which is readily available from cloud-based machine learning platforms.
  • The study suggests that such AI models could be used by pathologists to improve diagnostic efficiency.

Statistics:

  • 100% accuracy in identifying tubular adenoma.
  • 100% accuracy in identifying hyperplastic polyps.
  • 97% accuracy in identifying normal colon.
  • The AI model was trained on a dataset of whole-slide images from public and institutional databases.

Sources:

  • Automating Colon Polyp Classification in Digital Pathology by Evaluation of a "Machine Learning as a Service" AI Model: Algorithm Development and Validation Study. JMIR Formative Research, 2025;9.
  • NewsRx. Findings on Machine Learning from Evan Delancey and Colleagues Provide New Insights (Automating Colon Polyp Classification in Digital Pathology by Evaluation of a "Machine Learning as a Service" AI Model: Algorithm Development and Validation ...). Health & Medicine Week. August 22, 2025; p 1823.

Read more