Development of an Analytic Technique for Unstructured Data: Applications to Knowledge Extraction From User Reports on Medical Device Failures
These researchers are developing and demonstrating an analytic technique for large, unstructured data. The technique will be used to analyze data generated through user-level reporting at various stages of the medical device supply chain. The data are available through FDA’s “Manufacturer and User Facility Device Experience (MAUDE)” database. The technique’s application will extract knowledge for identifying the sources of device failures, unraveling the conditions under which the failures occur, and informing how the failures can be prevented. This knowledge will of interest to the key stakeholders of the medical device supply chain: device users, manufacturers, distributors, and regulators.