Researchers at the University of Jyväskylä, in collaboration with the University of Turku’s Institute of Biomedicine, University of Helsinki, and Nova Hospital of Central Finland, have developed an advanced artificial intelligence (AI) tool to automatically analyze colorectal cancer tissue slides.
Superior Accuracy in Cancer Tissue Analysis
The AI tool, powered by a neural network model, outperformed all previous models in classifying tissue microscopy samples. The study, published in the journal Heliyon, demonstrated remarkable accuracy.
“Our model identifies all tissue categories relevant for cancer detection with 96.74% accuracy,” said Fabi Prezja, the lead researcher behind the method’s design.
Transforming Pathology with AI
Traditionally, pathologists manually examine scanned digital microscopy slides from intestinal tissue samples, marking cancerous and related tissues point by point. This process is time-consuming and labor-intensive.
As reported by medicalxpress, the newly developed AI tool automates this process by analyzing tissue samples and highlighting areas containing different tissue categories. Its high accuracy can significantly reduce the workload of histopathologists, leading to faster diagnoses, improved prognoses, and better clinical insights.
Freely Available for Global Research Collaboration
To encourage further advancements, the research team has made the AI tool freely available.
“By sharing our tool, we hope to accelerate progress and inspire scientists, developers, and researchers worldwide to enhance it and discover new applications,” Prezja explained.
Ensuring Safe Clinical Integration
Despite the promising results, researchers emphasize that introducing AI tools into clinical settings must be gradual and cautious. Before becoming routine in medical practice, AI-based solutions must undergo rigorous validation to ensure compliance with clinical and regulatory standards.
With continued development and validation, this AI breakthrough could revolutionize colorectal cancer diagnostics, offering more efficient and accurate pathology workflows for the medical community.