A Methodology To Categorize Exam Series

DICOM is the global standard for managing and exchanging medical images and data but lacks a consistent series-naming scheme. I will use machine learning for the identification and classification task, and the detailed approach is mentioned in the proposal document. I will add a new module to run on DICOM images of the MRI modality. This module will execute on DICOM images acquired in real-time or on-demand by Niffler. It will identify and categorize the MRI series (orientation, weighting) into different brain sequences, which will help identify repeated or missing series in an MRI study. Additionally, it will check whether the sequence parameters meet standard thresholds. Lastly, a confidence score will be given.