ChromoEnhancer: An Artificial-Intelligence-Based Tool to Enhance Neoplastic Karyograms as an Aid for Effective...

Area of Science:

• Cytogenetics
• Medical Imaging
• Artificial Intelligence

Background:

• Cytogenetic laboratory tests are crucial for diagnosing genetic diseases, particularly hematological malignancies.
• Manual karyotyping is labor-intensive, time-consuming, and expensive, with current image enhancement methods having limitations.
• Poor-quality karyograms, common in cancer samples, hinder accurate diagnosis and analysis.

Purpose of the Study:

• To develop a novel artificial intelligence-based method, ChromoEnhancer, for enhancing neoplastic karyogram images.
• To improve the accuracy and efficiency of cytogenetic analysis for detecting chromosomal abnormalities.

Main Methods:

• Developed ChromoEnhancer using Generative Adversarial Networks (GANs) with a data-centric approach.
• Applied GANs to convert poor-quality karyogram images into high-quality images.
• Evaluated ChromoEnhancer by comparing enhanced and original images using cytogeneticist ratings and quantitative metrics (PSNR, SSIM).

Main Results:

• ChromoEnhancer successfully enhanced neoplastic karyogram images, improving quality.
• Enhanced images facilitated robust routine cytogenetic analysis and accurate detection of cryptic chromosomal abnormalities.
• Quantitative metrics and expert evaluations confirmed the effectiveness of the enhancement method.

Conclusions:

• ChromoEnhancer offers a novel, AI-driven solution for improving karyogram image quality.
• The method supports more accurate and efficient cytogenetic analysis, crucial for diagnosing genetic diseases.
• This artificial intelligence approach has the potential to advance diagnostic capabilities in cytogenetics.