What is a "Good" figure: Scoring of biomedical data visualization.

Area of Science:

• Biomedical data visualization
• Scientific communication
• Bioinformatics

Background:

• Interpreting complex biomedical datasets relies heavily on data visualization.
• Current visualization tools and applications exhibit significant variability in clarity and quality.
• Lack of standardized methods hinders effective evaluation of biomedical figures.

Purpose of the Study:

• To develop a comprehensive framework for evaluating biomedical figures.
• To benchmark the performance of various data visualization platforms.
• To provide solutions for enhancing figure design in biomedical research.

Main Methods:

• Developed Metrics for Evaluation and Discretization of Biomedical Visuals using an Iterative Scoring algorithm (M.E.D.V.I.S.) to quantify figure quality.
• Assessed figures based on complexity, color usage, whitespace, and number of visualizations.
• Integrated dimensionality reduction, clustering, and thresholding for figure classification and feedback generation.
• Conducted a comparative analysis of 26 visualization tools based on usability, customizability, cost, and expertise required.

Main Results:

• The M.E.D.V.I.S. algorithm provides a systematic approach to assess and improve biomedical figure quality.
• Comparative analysis revealed varying strengths and weaknesses across 26 visualization tools.
• Demonstrated real-world applicability through case studies and introduced SpatioView for spatial omics data exploration.

Conclusions:

• Standardized evaluation methods are crucial for advancing biomedical data visualization.
• The developed framework and tools offer practical solutions for improving figure design.
• Findings support enhanced figure quality in biomedical research, education, and industry.