Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Healthy Individuals
Abstract:
Existing system:
Aortic valve closes, a slight rebound of heart acceleration gives the AC point. In
ECG, the ventricular repolarization stops and produces offset of T-wave.
Furthermore, atrial pressure increases when blood enters into atrium. It causes a
rise in the acceleration followed by a sudden fall in the acceleration.
Proposed system:
Advantages:
The systolic profiles are enhanced using RSDMK weights. The envelope of a
systolic profile is extracted using the SE operator followed by the autocorrelation
feature. Finally, AO peaks are approximated using a FOGD filtering based
approach. The method is tested and validated on publically available CEBS
database.
Finally, AO peaks are detected using Shannon energy (SE) and autocorrelation
features based envelope construction and Gaussian derivative filtering based peak
detection logic. The rest of the paper is organized as follows. In Section II, the
proposed AO peak detection method is presented. In Section III, we evaluated the
performance of the proposed method. Finally, conclusions are drawn in Section IV.
Disadvantages:
This action generates an is electric line. As soon as mitral valve opens, the atrium
volume decreases, causes generation of MO valley point. After a certain interval of
time, ventricle starts filling with blood and heart walls move outward, which
augments the acceleration and produces RF peak point.
At this moment, the electric impulse travels from SA node to AV node for atrial
depolarization. It produces P-wave. In this way, a SCG cycle is generated.
Modules:
Seism cardiogram:
Signal the heart has two different pumps separated by a septum. Each of these
pumps comprises of two chambers- atrium and ventricle. These chambers are again
separated by atrioventricular and semi-lunar valves. Different phases of a cardiac
cycle can be identified with the help of a SCG signal, and these instances/phases
are mitral valve opening (MO), mitral valve closure (MC), aortic valve opening
(AO), aortic valve Closure (AC), isovolumic contraction time (IVCT), isovolumic
relaxation time (IVRT), rapid filling (RF) and rapid ejection (RE) of blood through
ventricles. In Fig. 1, a typical SCG signal is shown for two cardiac cycles along
with simultaneously recorded ECG. The mitral valve is closed when the blood
completely enters from left atrium to left ventricle. As a result, MC notch point is
seen as a sudden positive deflection in SCGsignal. At this moment, electric
impulse depolarizes purkinje fibers, which results major ventricular depolarization
(Rwave in ECG) in the electro bio-potential measurement. Then ventricles start
contracting, which causes sharp inward wall motion and decreases acceleration
waves.
However, the proposed method performs well in the presence of baseline drifts.
Records b001, b003, b006, b018, and b020 contain abrupt changes and
unrecognizable distorted beats. Fig. 9 illustrates the detection performance of SCG
signal, which consists of distorted beats and unrecognizable heart rhythms. These
signals are even very difficult to annotate manually. For such instants, the
proposed algorithm gives a large number of misdetections. In addition, the records
b001, b007, and b020 have smaller AO peak morphologies than RE, and the record
b011 has similar diastolic structures as systolic profile. For these cases, the
proposed method produces more FPs and FNs. The performance of the proposed
method for the SCG signal having numerous smaller systolic profiles is presented
in Fig. 10. Records b001 and b006 have larger variations in AO amplitudes. Our
method gives more FPs for two records 003 and 020 across all the records in the
database due to their distorted beats and spurious spikes. Thus, by eliminating
these two records (003 and 020), the overall performance are achieved as Se =
94.5%, +P = 93.4% and ACC = 88.65%.