Abstract: Systems and methods for identifying and presenting prior findings data in a radiology review workflow, based on natural language processing (NLP) of unstructured radiology report text, are disclosed. In an example, prior findings are generated with operations including: receiving text from a prior radiology report prepared for a patient in a prior radiology study; processing the text with a trained NLP engine to classify discrete prior findings in the prior radiology study that identify observed condition(s) and characteristics of the observed condition(s); and outputting or presenting the prior findings, such as in a radiology review user interface. In an example, the radiology review user interface displays the observed condition(s) and the characteristic(s) of the observed condition(s) to a user (e.g., a radiologist), and can apply sorting, filtering, or grouping based on pathology phrases or attribute words associated with the observed conditions or the characteristics of such observed conditions.

Abstract: A co-axial filtering device for removing particles from arteries and veins comprises an outer catheter that can be inserted into a blood vessel and an inner catheter with a filter at its distal end. The filter is a radially expandable receptacle made of an elastic mesh structure of spring wires or plastic monofilaments. The filter expands into engagement with the walls of the blood vessel when it is pushed from the distal end of the outer catheter. A syringe is attached to the proximal end of the inner catheter and particles entrapped within the filter are aspirated through the inner catheter into the syringe. When filtering is no longer necessary, the filter can easily be removed by pulling the elastic material through the outer catheter.