Big Data from CT Scanning

Over 100-million of x-ray CT scans are performed worldwide each year. In most cases, a scan projection or sonogram data are discarded after images are read. This represents a huge waste of big data, and an opportunity to develop new methods for better image reconstruction and high dose efficiency. Here we present an initial attempt to archive, utilize and share big data from CT scanning. In this project, CT scans of several cadavers are used as examples. The data were collected at Massachusetts General Hospital at multiple different radiation dose levels for different x-ray spectra, and with representative reconstruction techniques. Hence, this database is more informative than others of this kind as prior knowledge to improve image reconstruction and image analysis, and reduce radiation dose.