A group of scientists in the Spain public research council has developed and registered a new method for detection of events in biomedical signals of interest in areas such as phonocardiography and electrocardiography, which can be useful in other areas as well such as seismic analysis and inspection by sound. The method is based on a Rényi entropy measure extracted from a time-frequency representation. Inventors are looking for interested companies to license that technology.

Detalles

The present innovation relates to techniques for digitally capturing and processing biomedical signals of cardiac interest in phonocardiography and electrocardiography. The procedure is based on a time-frequency analysis of the signal of interest followed by extraction of a particular normalisation of the Rényi entropy. This method proposes one specific entropy normalisation, under an instantaneous-wise basis, as the most suitable among others for detection of normal and abnormal patterns and their singularities. This procedure originates a new temporal sequence of data where normal and anomalous events can be better detected and visualised by providing a more discriminative distinction between them. It has direct applications in many biomedical signals of interest such as electrocardiograms (ECG), electroencephalograms (EEG) as a help for detection i.e. precursors of epilepsy, electromyograms, as well as in other areas such as seismic signal analysis and detection of failures in inspection by sound.

This technology provides better-localised information about normal and pathological events hence early diagnosis of diseases and phenomena are facilitated this way. Innovative Aspects: A phonocardiogram (PCG) is a graphical representation of the sounds produced by the heart. A normal cardiac cycle has two main sounds, called S1 and S2, which define the start of systole and diastole, respectively. A common problem in the processing of phonocardiographic signals is to detect the main events of the cardiac cycle (S1, S2, murmurs, etc) as a tool for helping the diagnosis of the valvular state of the heart. The delineation of the individual events is a challenging task, because the events are usually very close in time, and their separation may be not clear using threshold techniques. Classical methods based on magnitude thresholding are frequently not able to delineate and separate the events of interest. The new developed method is able to accurately delineate the presence of events, even if two or more events appear joined or even if murmurs have amplitude larger than S1 and S2. This method of biomedical signal analysis offers a new way to analyse cardiac signals. Other advantages include the low computational cost due to the fact it implements a 1D Wigner-Ville distribution what facilitates its implementation in portable devices. Another advantage if that it will allow an objective measure of abnormal events without the presence of an expert.

Estado de desarrollo

Available for demonstration - field tested

Propiedad Intelectual

Patent(s) applied for but not yet granted

Nichos de Mercado

Categorías Information Processing, Information System, Workflow Management , Telephone Related , Computer Software , Sound Engineering/Technology , Analytical and Scientific Instrumentation , Electromedical and Medical Equipment   Language en

Información relacionada

• automatic separation of neural signals in ecg measurement
• nondestructive acoustic method and device, for the determination of detachments of mural paintings
• analysis of three-phase partial discharge signals
• transcriptome analysis in non-model species: a new method for the analysis of heterologous hybridization on microarrays
• stationary and not stationary hybrid plants with molten carbonate fuel cells (mcfc) or solide oxide fuel cells (sofc) for the simultaneous production of electric, thermal and cooling energy
• innovative acoustic materials for room & building acoustics, acoustic doors
• sequence method for analysis of blood
• fast and reliable calibration method for sample analysis
• fast and quantitative method for chloramphenicol analysis in honey.
• innovative analysis method for distributed embedded systems