In recent years, machine learning models have been introduced to solve many problems in medical imaging, such as segmentation, classification, and disease prediction. Unfortunately, most of them are useless for the physician due that they are not available tools that are part of their workflow. At present, Picture Archiving and Communication Systems (PACS) are the standard platforms used in clinical environments to store and transmit electronic images and the reading reports. Therefore, the integration of the automatic analysis tools with these systems is required to allow validation by physicians and the use in clinical and medical research. This paper presents a simple way of adding the use of machine learning models for the automatic analysis of medical images in the radiological workflow using DICOM services provided by open source tools. An implementation case study is also presented, in which a deep learning architecture was trained for classifying chest X-ray images as normal, bacterial pneumonia or viral pneumonia, including in the last case images of COVID-19 patients.
CITATION STYLE
Osorno-Castillo, K., Fonnegra, R. D., & Díaz, G. M. (2020). Integration of Machine Learning Models in PACS Systems to Support Diagnostic in Radiology Services. In Communications in Computer and Information Science (Vol. 1274 CCIS, pp. 233–244). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-61834-6_20
Mendeley helps you to discover research relevant for your work.