Relevance of eHealth standards for big data interoperability in radiology and beyond

Abstract

The aim of this paper is to report on the implementation of radiology and related information technology standards to feed big data repositories and so to be able to create a solid substrate on which to operate with analysis software. Digital Imaging and Communications in Medicine (DICOM) and Health Level 7 (HL7) are the major standards for radiology and medical information technology. They define formats and protocols to transmit medical images, signals, and patient data inside and outside hospital facilities. These standards can be implemented but big data expectations are stimulating a new approach, simplifying data collection and interoperability, seeking reduction of time to full implementation inside health organizations. Virtual Medical Record, DICOM Structured Reporting and HL7 Fast Healthcare Interoperability Resources (FHIR) are changing the way medical data are shared among organization and they will be the keys to big data interoperability. Until we do not find simple and comprehensive methods to store and disseminate detailed information on the patient's health we will not be able to get optimum results from the analysis of those data.

Introduction of Systematized Nomenclature of Medicine–Clinical Terms Coding Into an Electronic Health Record and Evaluation of its Impact: Qualitative and Quantitative Study

“…4. Clinicians are driven by patient safety, data exchange across the health system, and research for population health. 5. Sophisticated and CDS or artificial intelligence algorithms cannot be leveraged without this data collection.…”

Mobile Health Technologies for Diabetes Mellitus: Current State and Future Challenges

“…Big data analytics tools (e.g., Apache Spark, Apache Mahout, and Storm) and techniques (e.g., data mining, machine learning, and statistics) can very much benefit from this unified, consistent, and integrated big data. Marcheschi [125] evaluated the existing e-Health standards (i.e., DICOM, HL7 vMR, and HL7 FHIR) for big data interoperability, and asserted that the success of big data analytics is tightly connected with handling interoperability between heterogeneous data sources. Rathore et al [128] proposed a real-time CDSS for emergency response.…”

Structured radiology reporting on an institutional level—benefit or new administrative burden?

“…It therefore remains challenging to use these promising techniques in the current healthcare IT environments. Although various efforts have already been made to improve interoperability of healthcare IT systems, a large proportion of relevant information is still not easily accessible for further data analysis …”

“…Although various efforts have already been made to improve interoperability of healthcare IT systems, a large proportion of relevant information is still not easily accessible for further data analysis. 43 Recently, extending the idea of big data analysis, machine learning and deep learning have emerged as a highly interesting research topic in radiology. Especially with the advent of deep convolutional neural networks in computer vision, more and more researchers are focusing on teaching computers to see patterns in the imaging data, potentially allowing for fully automated diagnosis in different settings.…”