We built a scalable image processing,
post-processing, and extraction system tailored
for healthcare, handling diverse modalities like
CT scans, MRIs, and X-rays. Pre-processing
involved normalizing and standardizing datasets,
enhancing visibility with contrast adjustments,
reducing noise using filters, and segmenting key
areas like tumors with advanced algorithms such
as U-Net and FCNs. Our pipeline ingested medical
images from PACS and other devices, preprocessed
them, and leveraged AI models to extract
features. Post-processing enhanced contrast and
highlighted critical areas for clinicians, while
automated reports were generated. All datasets
were securely stored in compliance with privacy
regulations.
