Tumor Spheres Quantification with Smoothed Euclidean Distance Transform.

Area of Science:

• Oncology
• Biomedical Imaging
• Computational Biology

Background:

• Accurate tumor sphere quantification is crucial for cancer research and drug screening.
• Manual methods are time-consuming, error-prone, and not scalable for large datasets.

Purpose of the Study:

• To develop an automated method for precise tumor sphere quantification.
• To address the over-segmentation problem in existing automated methods.

Main Methods:

• A novel segmentation technique utilizing a smoothed distance transform to refine watershed algorithm seeding.
• The method identifies and eliminates superfluous minima before watershed segmentation.

Main Results:

• The proposed method effectively overcomes over-segmentation issues inherent in traditional distance transform watershed algorithms.
• Validation on pancreatic tumor sphere images demonstrated high efficiency and accuracy in quantification.

Conclusions:

• The developed automated method offers a robust and efficient solution for tumor sphere quantification.
• This technique has significant implications for advancing cancer research and drug discovery pipelines.