Development and Evaluation of Techniques for Computer Aided Detection and Diagnosis From Existing Radiologic Images

Overview

Background: Radiologic images are getting more complex. They are also being used more often. Medical personnel are overwhelmed with data. Computer Aided Detection (CAD) and diagnosis may be able to improve medical care. Researchers want to create and test ways to use CAD. To do this, they want to use data from the Clinical Center s Department of Diagnostic Radiology. Objective: To create algorithms and software that accurately detect and characterize lesions, model anatomy, and monitor diseases on radiologic studies. Eligibility: People of all ages who have radiologic exams stored in the clinical PACS (picture archiving system) of the Clinical Center since July 6, 1953 with no end date. People with target lesions of any kind will be included. Design: This study will use existing data. Participants will include males and females of all ages. They will be chosen by keyword search on NIH databases. The data that is used may include CT, MRI, ultrasound, and other images. It may include a participant s name, date of birth, and date of exam. It may include the name of the doctor, radiologist, and hospital. Data such as age, gender, race, disease, and treatment may be used. Other data from charts or studies may be used. Imaging data of all organs of the body will be studied. Data will be kept in computers and servers. The equipment will be password protected. Printouts will be stored in locked rooms. This study will last 10 years.

Study Description: This study uses artificial intelligence techniques to improve radiology diagnosis. Objective: Development of Computer Aided Detection techniques to improve radiology imaging. Study Population: Up to 1,000,000 NIH Clinical Center patients. Description of Sites/Facilities conducting research: NIH Clinical Center Study Duration: 10 years

Conditions