Medical images are inherently high-resolution and contain locally varying structures crucial for diagnosis. Efficient compression must preserve diagnostic fidelity while minimizing redundancy. Low-rank matrix approximation (LoRMA) techniques have shown strong potential for image compression by capturing global correlations; however, they often fail to adapt to local structural variations across regions of interest. To address this, we introduce an adaptive LoRMA, which partitions a medical image into overlapping patches, groups structurally similar patches into clusters using k-means, and performs SVD within each cluster. We derive the overall compression factor accounting for patch overlap and analyze how patch size influences compression efficiency and computational cost. While applicable to any data with high local variation, we focus on medical imaging due to its pronounced local variability. We evaluate and compare our adaptive LoRMA against global SVD across four imaging modalities: MRI, ultrasound, CT scan, and chest X-ray. Results demonstrate that adaptive LoRMA effectively preserves structural integrity, edge details, and diagnostic relevance, measured by PSNR, SSIM, MSE, IoU, and EPI. Adaptive LoRMA minimizes block artifacts and residual errors, particularly in pathological regions, consistently outperforming global SVD in PSNR, SSIM, IoU, EPI, and achieving lower MSE. It prioritizes clinically salient regions while allowing aggressive compression in non-critical regions, optimizing storage efficiency. Although adaptive LoRMA requires higher processing time, its diagnostic fidelity justifies the overhead for high-compression applications.

"For only four hours of simulated data from the telescope, my code could run for five days before outputting the results," Atemkeng says. "I became passionate about the amount of data the SKA would produce in the future – a data stream likely to be in the order of PB/s. At this data scale, even using the most powerful supercomputers, the computation will always remain a major challenge. So I became interested in developing tools for big data, which can be implemented with AI and machine learning algorithms." For Atemkeng, most of Africa's problems – from poverty to food access for its rapidly growing population, education and healthcare access in rural areas – could be solved using machine learning and big data analytics.