Session P92.4

ECG and Echocardiography Processing for Decision Support in Heart Failure

F Chiarugi, S Colantonio, D Emmanouilidou, D Moroni*,
F Perticone, O Salvetti, A Sciacqua

ISTI - CNR
Pisa, Italy

Signal and imaging investigations are a basic step of the diagnostic, prognostic and follow-up processes of heart diseases. Besides, the increasing availability of heterogeneous clinical data and the need of an integrated approach to their analysis have lead nowadays to a renaissance of Clinical Decision Support Systems (CDSS). The purpose of this work is to present an effective way to achieve a high-level integration of signal and image processing methods in the general process of care, by means of a CDSS. Among several heart diseases, we treat the specific yet paradigmatic example of heart failure, that for its complexity highlights best the benefits of this integration. After a careful investigation about the needs of practitioners and the formalization of the main decisional problems, the importance of considering and interpreting ECG signals and echocardiography images had come forth, since objective and reliable evaluation of important features can facilitate decisional problems in collaboration with the CDSS. Thus a signal and image processing suite was implemented. It consists in i) modules for the automatic and semi-automatic segmentation of echocardiographic sequences and the extraction of geometrical parameters (chamber sizes, volumes and ejection fraction) and ii) modules for QRS detection, identification of dominant beats and construction of the averaged dominant beat from the raw data acquired by ECG devices with different lead numbers and acquisition periods. Special care was given to the seamless integration of the signal and image processing suite in the general IT infrastructure: a composite repository was prepared and enabled with standard-compliant network services (DICOM, HL7, SCP-ECG) for storing in a structured way both the original and processed data. Then integration of signal and image processing models in the CDSS was assured by a dedicated formalization of the relevant acquisition modalities, diagnostic examinations and computable parameters within an ontology of the Domain Knowledge Base. Inferential rules able to process parameters extracted from both signals and images were encoded into the same Knowledge Base. Finally, a Meta Knowledge Base containing suitable procedural rules for integrating the application of the data processing methods into the inferential reasoning process was included.

(Abstract Control Number: 310)