Clustering algorithms benchmarked for medical image analysis and compression

A new research paper evaluates five clustering algorithms—k-means, mini-batch k-means, agglomerative hierarchical clustering, BIRCH, and bisecting k-means—for their effectiveness in medical image analysis. The study specifically examines their performance in preserving local details crucial for diagnosis while exploring potential for adaptive image compression. Results indicate that standard k-means and bisecting k-means generally perform well but can have high intra-cluster variability. Agglomerative clustering showed superiority for MRI and ultrasound in maintaining homogeneity, while mini-batch k-means balanced quality and compactness for chest X-rays. BIRCH was found to underperform across all tested modalities. AI

IMPACT This research could lead to improved medical image compression techniques, potentially reducing storage and transmission costs while preserving diagnostic detail.

RANK_REASON The cluster contains a research paper detailing the performance benchmarking and optimization of clustering algorithms for medical image analysis. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.CV →

• agglomerative hierarchical clustering
• BIRCH
• bisecting k-means
• chest X-ray
• k-means
• mini-batch k-means
• ultrasound

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →