Page 407 - Kaleidoscope Academic Conference Proceedings 2024

Innovation and Digital Transformation for a Sustainable World

a. Identify the principal fiducial points on a generated
ECG report.
b. Compute the statically calculated variance in the
generated R—R gap by applying a certain threshold ECG reflects AI-enabled A.I. adoption of
value. electrical ECG methods AI-enabled
activity and analysis ECG
c. Identify abnormal activity – with the help of P waves
checks the
 No regular atrial activity was identified. condition of
the cardiac
 Any irregular abnormality activity reported. ?
cycle [15]
 High-frequency abnormality waves are present?
Using the above algorithmic approach, the results are ECG approach Detecting DL, NN ECG and
generated to detect the abnormality of the heart and used for ECG arrhythmia
reported accordingly [8, 9]. identifying anomalies classification
CVD using DL .
RELATED WORK problems.[16] models.

Different researchers apply distinct ML approaches to solve ECG analysis ECG CNN, classification
heart disorder-related predictions using ECG signals using and analysis for DL, ML, accuracy is
classification models. Table - 2 below describes related classification classification LSTM 99.13%
work, different ML methods, classifiers used, and for CVD [17]
inferences taken from the paper in tabular form.
Automated ECG data CNN, A system
TABLE 2: DESCRIBES RELATED WORK, diagnosis of analysis SVM that helps for
DIFFERENT ML METHODS, CLASSIFIER USED, CVD diseases the patient's
AND INFERENCE [18] autonomy

Related work Method Classifie Inference Interpretable Identify and ML, DL Progress for
r used ML characterize ECG signal
techniques for ECG signals. identificatio
ECG, ML and Time series ALL ML Performance CVD [19] n
time series and meta- algorith of ML for
analysis [10] analysis of ms ECG Machine Meta- ML, ML is
algorithms classification Learning for analysis of effective for
Detecting diagnostic DL, detecting
ML for ECG Patients with ML with R.F. model Atrial accuracy CNN A.F. from
risk and occlusion classifier shows good Fibrillation ECGs
treatment [11] myocardial results. from ECGs
infarction [20]
(OMI)
ML-based Optimized ML, comparing
configuration Developing R.F., Better disease framework CNN, the
of electrodes algorithms N.N. and education classification named DL outcomes of
affects the to detect D.T. regarding techniques. WbGAS for different
result [12] electrode ECG [21] prediction. ML-based
misplacemen acquisition. approaches
t
Arrhythmia MMPA to SVM, For ECG
AI-enabled Adoption of AI Better can modify the identify A.F. A.F. recordings,
ECG [13] the AI- methods Diagnosis heart's in brief ECG HRV is
enabled rhythms and data. effective and
ECG its potential reliable for
impact [22] A.F.
AI-Enabled A.I. using A.I. with Health identificatio
Electrocardiog ECG data ML for monitoring n
ram Analysis accompanied classifica and early
for Disease by modern tion diagnosis. Classifying Building SVM, The
Diagnosis [14] wearable patients using DSS for Grid proposed
biosensor ML to build a prediction search DSS is the

– 363 –