Mayo Foundation for Medical Education and Research

Matthew R Callstrom

C Daniel Johnson

Bradley J Erickson

Judd E Reed

Zeynettin Akkus

In accordance with some embodiments, systems, methods, and media for automatically localizing and diagnosing thyroid nodules are provided. In some embodiments, a system for automatically diagnosing thyroid nodules comprises: an ultrasound machine; and a processor programmed to: receive a B-mode ultrasound of a thyroid from the ultrasound machine; provide the B-mode ultrasound to a classification model trained to automatically segment B-mode ultrasound; receive an output indicating which portions of the B-mode ultrasound correspond to a nodule; provide at least a portion of the B-mode ultrasound corresponding to the nodule to a second classification model trained to automatically classify thyroid nodules based B-mode, color Doppler, and shear wave elastography ultrasound; and receive, from the second trained classification model, an output indicative of the likelihood that the nodule is malignant.

Systems and media for automatically diagnosing thyroid nodules

A method for generating colonography images for colorectal screening of a patient. The patient is administered an amount of radiopaque stool marker which will enable stool present in the patient's colon to be distinguished from soft tissue. The patient's colon is imaged after the stool marker has been administered to generate colonography images. Preparation of the colon is simulated by processing the colonography images to remove marked stool before the images are observed during a diagnosis session.

Growing community of inventors

Rochester, MN, United States of America

Matthew R Callstrom