Page 114 - AIH-2-3

Artificial Intelligence in Health Opportunities for AI-based arrhythmia screening

with costs projected to reach US $70 billion by 2030. The typically consists of five distinct peaks labeled P, Q, R, S,
cost of heart failure varies widely, from less than US $1,000 and T, with a frequently observed sixth peak, the U wave.
per patient in low-income countries to €5,000 – 15,000 The P wave is the result of the depolarization of the atria,
in Europe and US $17,000 – 30,000 in the US. Accurate while the remaining waves are caused by the depolarization
and timely diagnostic screening, coupled with advanced of the ventricles. The R wave represents the synchronized
warning systems, could save approximately 1,250 lives per depolarization of the right and left ventricles over time.
day in the US, 40 in the United Kingdom (UK), and 1,100 Cardiac muscles contract as a direct result of cellular
across the EU. Notably, data from Asia and Russia remain electrical excitation, mediated by active and passive
incomplete. SCD, often caused by ventricular fibrillation ion transport across cell membranes. Cardiac cells are
(VF), affects individuals of all ages, depending on factors not directly stimulated by neurons but instead initiate
such as congenital disorders and lifestyles. The incidence depolarization through an intrinsic excitation process.
in pediatric and juvenile populations can be as high as The depolarization of one cell triggers a cascade of
1 in 10,000. Women are generally disproportionately depolarization and contraction in neighboring cells,
affected and often face underdiagnoses or misdiagnoses. resulting in coordinated atria and ventricular contractions
In addition, people of African descent and Asian Indian that pump blood through the heart. The electrical
men are more frequently affected than Caucasians. The use activation of each cardiac cell serves as an indication of
of AI in medical diagnostics can save time, enhance the its health and functional state. The summation of these
level of detail, and improve early detection of pathological electrical activities can be represented as a vector, which
conditions. It can uncover warning signs of deviations in changes direction as depolarization propagates through
the image and signal patterns that may require follow-up the heart (Figure 2). The details of this vector rotation can
investigations and potentially additional tests. 2-7 be captured by placing multiple electrodes on the skin.
The limitations of manual analysis of diagnostic results Modern ECG systems employ advanced signal
include personal preferences, bias, high dependency on processing techniques to enhance the signal-to-noise ratio,
professional experience, skills, and abilities, and personal enabling precise analysis of arrhythmogenic tendencies.
health conditions, such as fatigue, eyesight problems, The ECG is, therefore, the cumulative result of the
and personal medical history (e.g., color blindness). For depolarization of individual cardiac muscle cells over
example, traditional X-ray imaging alone may fail to fully time, occurring in a controlled and repetitive manner.
assess coronary stenoses or predict the progression of life- The 12-lead ECG, the most detailed configuration, offers
threatening cardiovascular conditions. One important spatial information about depolarization patterns over
characteristic of coronary health, particularly cardiac time (Figure 3). In contrast, most other detection and
depolarization health, is the recognition of fatal factors monitoring techniques only provide temporal information.
underlying disease progression. AI-driven computational By recording the depolarization of cardiac muscle cells over
analysis can enhance pattern recognition and identify extended periods, clinicians can diagnose abnormalities,
clusters of subtle abnormalities that may be overlooked identify malfunctioning regions, and determine the need
by visual inspection, such as those in peak-to-peak for further medical intervention. Deviations from a typical
sequencings and groupings like image-clustering. 8-11 ECG pattern, whether detected visually or through signal
One area of interest in diagnostics that may benefit processing, can be classified as specific cardiac disorders.
from automated screening through software algorithms, During VF, the depolarization vector will no longer be
often mathematically based, is electrocardiogram (ECG, recognizable due to the random and chaotic nature of
also known as electrocardiogram in German) acquisition. electrical activity in the ventricles. 9,10,14,15
The ECG was introduced in 1903 by Willem Einthoven This paper introduces the pathological conditions that
from the Netherlands, who pioneered the use of a specially can potentially be identified using various mathematical
designed galvanometer to record the action potentials. techniques under AI-driven signal processing. Each
Early recordings were obtained by immersing hands condition may require unique technical approaches. The
and feet in saltwater to enhance conductivity. The ECG ultimate goal is to provide predictive mechanisms for early
captures the electrical depolarization and repolarization of detection and prevention of cardiac disorders. The early
cardiac muscle cells. 12,13 These electrical signals, known as detection and intervention of cardiac disorders require a
action potentials, result from the flow of ions across the cell highly regulated control system, which is currently under
membrane, leading to muscle contraction. A single ECG development and may take several years for market release.
cycle is illustrated in Figure 1A, while a representative It is important to recognize that AI-based diagnosis is
healthy ECG is shown in Figure 1B. The ECG waveform not a standalone solution. Medical screening requires a

Volume 2 Issue 3 (2025) 108 doi: 10.36922/aih.8468