Published In

IEEE Access

Document Type

Article

Publication Date

7-2018

Subjects

Cooperating objects (Computer systems), Nervous system -- Diseases -- Detection, Computer science -- Medical applications

Abstract

Paroxysmal diseases of inpatients are globally recognized as one of the top challenges in medicine. Poor clinical outcomes are primarily caused by delayed recognition, especially due to diverse clinical diagnostic criteria with complex manifestations, irregular episodes, and already overloaded clinical activities. With the proliferation of measuring devices and increased computational capabilities, cyber-physical characterization plays an increasingly important role in many domains to provide enabling technologies. This paper presents a cyber-physical system (CPS) framework to assist physicians in making earlier diagnoses of paroxysmal sympathetic hyperactivity based on existing medical knowledge. We propose a configurable diagnostic knowledge model to characterize clinical criteria to reduce domain knowledge deficiency between physicians and computer scientists. We present a component-based medical CPS framework to employ the knowledge models and integrate medical devices. Our approach aims to relieve medical staff from the heavy burden of clinical activities and to provide timely decision support. We evaluate our approach on 128 realworld clinical cases. Compared with the state-of-the-art approach, the results demonstrate that we enable early detection in 11.02% more patients and detect the condition 16.57 hours earlier on average.

Description

© 2018 IEEE.

Originally appeared in IEEE Access, volume 6, 2018. May be accessed at https://doi.org/10.1109/ACCESS.2018.2850039

DOI

10.1109/ACCESS.2018.2850039

Persistent Identifier

https://archives.pdx.edu/ds/psu/26159

Share

COinS