Improved gated recurrent unit-based osteosarcoma prediction on histology images: a meta-heuristic-oriented...

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Osteosarcoma is a prevalent primary bone cancer requiring effective treatment strategies.
• Current diagnosis and treatment rely on chemotherapy response, but chemotherapy alone can lead to persistent disease.
• Early diagnosis and individualized care are crucial for improving osteosarcoma patient survival rates.

Purpose of the Study:

• To develop a novel deep learning approach for predicting osteosarcoma directly from histology images.
• To enhance the accuracy and efficiency of osteosarcoma diagnosis through advanced computational methods.

Main Methods:

• Utilized the UT Southwestern/UT Dallas osteosarcoma dataset for image collection.
• Applied Weiner filter for image pre-processing and 2D Otsu's method for segmentation.
• Employed Linear Discriminant Analysis (LDA) for feature extraction.
• Developed an Improved Gated Recurrent Unit (IGRU) model optimized by the Osprey Optimization Algorithm (OOA) for prediction.

Main Results:

• The developed deep learning model demonstrated significant effectiveness in osteosarcoma prediction.
• Comparative analysis showed superior performance against conventional methods.
• The IGRU model, optimized by OOA, achieved high accuracy in identifying osteosarcoma from histology images.

Conclusions:

• The proposed deep learning framework offers a promising tool for the early and accurate diagnosis of osteosarcoma.
• This approach has the potential to guide individualized treatment strategies and improve patient outcomes.
• Further research and validation are warranted to integrate this technology into clinical practice for osteosarcoma management.