Optimizing Data Quality and Decision-Making in IoT with AI-Driven Data Reduction in DICOM

The extensive adoption of Internet of Things (IoT) applications increases the need to strategically deploy sensor devices, generating vast volumes of data. This extensive data flow can overwhelm network capacities, highlighting the need for efficient data reduction techniques. This paper introduces a multi-model AI-based data reduction solution to optimize data quality and decision-making processes in IoT environments. By predicting critical analytical metrics such as reduction and distortion ratios, our approach allows for the dynamic selection of a suitable DR algorithm, thereby enhancing both storage efficiency and data utility. Our experimental validation, conducted using Digital Imaging and Communications in Medicine (DICOM) images, demonstrates the need of our solution in processing high-density data, thereby avoiding exhaustive processing and ensuring optimal data management.