An Innovative Solution Based on TSCA-ViT for Osteosarcoma Diagnosis in Resource-Limited Settings.

Area of Science:

• Oncology
• Biomedical Imaging
• Artificial Intelligence

Background:

• Osteosarcoma diagnosis is challenging, particularly in resource-limited settings.
• Accurate cell nucleus segmentation is crucial for advanced diagnostics.
• Existing methods struggle with image noise and cost-effectiveness.

Purpose of the Study:

• To develop an efficient and accurate AI solution for osteosarcoma cell nucleus segmentation.
• To address challenges of image noise and high costs in pathological analysis.
• To improve diagnostic capabilities for osteosarcoma in developing nations.

Main Methods:

• Introduction of the Twin-Self and Cross-Attention Vision Transformer (TSCA-ViT) AI system.
• Utilized a directed filtering algorithm for noise reduction.
• Employed a transformer architecture with twin and cross-attention mechanisms for segmentation and spatial information augmentation.

Main Results:

• Achieved an average precision of 97.7% on 1000 osteosarcoma pathology slide images.
• Demonstrated superior performance compared to traditional segmentation methods.
• TSCA-ViT showed enhanced computational efficiency with fewer parameters, reducing time and equipment costs.

Conclusions:

• TSCA-ViT offers a highly effective and efficient solution for osteosarcoma cell nucleus segmentation.
• The AI system addresses critical challenges in image quality and cost.
• Presents a promising approach for advancing osteosarcoma diagnosis and management in resource-constrained environments.