Curvelet initialized level set cell segmentation for touching cells in low contrast images

Area of Science:

• Biomedical imaging
• Computational biology
• Image analysis

Background:

• Accurate cell segmentation is crucial for automated cell analysis.
• Low contrast images present significant challenges for distinguishing cell structures.
• Existing methods often struggle with segmenting touching cells and nuclei in such images.

Purpose of the Study:

• To develop and validate a robust method for cell segmentation, specifically targeting cell nuclei and boundaries in low contrast images.
• To improve the accuracy and reliability of automatic cell analysis pipelines.

Main Methods:

• Image enhancement using a combination of multiscale top hat filter and h-maxima to improve contrast.
• Cell nuclei and boundary detection via a curvelet-initialized level set method.
• Quantitative evaluation of image enhancement using Peak Signal to Noise Ratio (PSNR) and segmentation performance using accuracy, sensitivity, and precision.

Main Results:

• The proposed image enhancement technique effectively improved the contrast of low contrast cell images.
• The curvelet-initialized level set method demonstrated superior performance in segmenting cell nuclei and boundaries, especially for touching cells.
• Validated performance metrics showed significant improvements in accuracy, sensitivity, and precision compared to existing methods.

Conclusions:

• The proposed method offers a significant advancement in cell segmentation for low contrast microscopy images.
• This technique enhances the capabilities of automatic cell analysis, particularly in challenging imaging conditions.
• The validated improvements in performance metrics suggest broad applicability in biological research requiring precise cell delineation.