News & Events
Coming soon with more information.
WideMed has developed an advanced signal processing technology for the analysis of highly non-stationary life-sign signals such as EEG, Respiratory, ECG, Blood Saturation, and more. The automatic signal analysis enables the processing of very large volumes of data, performing event identification and classification with high precision, robustness and repeatability that are essential for diagnostics and research. Unlike human scoring, WideMed's signal processing performs quantification of signals and events, which enables advanced trend analysis and data mining for diagnostics and treatment management.
The innovative patent of EEG analysis, performed by the Morpheus system is based on patient specific adaptation of a dynamic frequency states model. EEG signals are adaptively segmented into quasi-stationary segments, enabling micro-sleep and sleep fragmentation analysis. Patient intrinsic frequency states are recognized by utilizing fuzzy clustering of quasi-stationary EEG segment features.
Not restrained by 30 second epochs, WideMed’s analysis provides improved microstructure and state analysis.
The brain frequency spectral analysis is then converted to a 30 seconds epoch view based on R&K rules, allowing its user to be in a familiar environment.
Respiratory Events Analysis
Analysis of the respiratory events is being preformed by an adaptive segmentation algorithm, applied to the respiration envelope signal in order to detect amplitude reduction caused by respiration cessations. By using fuzzy logic approach, all the detected respiration events are classified into apnea, hypopnea and RERA classes, where events classified into the apnea class are being sub-classified into central, mixed and obstructive subclasses.
Limb Management Analysis
Leg movements are detected by applying burst detection algorithm on the summed leg channels. Only bursts with duration between 0.5-8 seconds are regarded as valid leg movements.
Periodic LEG Movements (PLMS) - Periodic leg movements (PLMS) are recognized by detection of frequent patterns of leg movements i.e. 4 or more successive limb movements, where the duration of each limb movement is between 0.5 and 8 seconds and the time interval between successive leg movements is greater than 4 seconds, but no more than 90 seconds.
The bursts are isolated by detecting a short differentiation in the energy behavior of the microphone recording. Only bursts with duration less than 6 seconds are regarded as valid snore events. Successive events, which occur in less than 0.5 second from one another, are regarded as one EMG burst episode. All the numbers are customizable.