Grading of glioma tumors using digital holographic microscopy

Area of Science:

• Neuro-oncology
• Medical imaging
• Computational pathology

Background:

• Gliomas are common, aggressive brain tumors with high mortality.
• Accurate and rapid tumor grading is crucial for effective treatment.
• Current histopathological grading is time-consuming and subjective.

Purpose of the Study:

• To develop a supervised machine learning algorithm for glioma grading.
• To utilize quantitative phase images from digital holographic microscopy.
• To improve diagnostic accuracy and speed for brain tumor characterization.

Main Methods:

• Acquired quantitative phase images of unstained tissue samples.
• Computed statistical and texture parameters from the images.
• Developed a machine learning model for classification and grading.

Main Results:

• Classified six image classes: normal tissue and five glioma subtypes.
• Distinguished glioma grades II, III, and IV with high accuracy.
• Achieved highest sensitivity and specificity for grade II astrocytoma and grade III oligodendroglioma.

Conclusions:

• The proposed method offers a swift and reliable approach to glioma grading.
• This technique enhances clinical diagnostic accuracy for brain tumors.
• Improved tumor characterization supports better treatment decision-making.