Corresponding Author: Hafiz
Khawar Hussain
E-mail: [email protected]
INTRODUCTION
In the entire world, cardiovascular diseases (CVDs)
are a significant source of morbidity and mortality. The World Health
Organisation (WHO) estimates that 17.9 million deaths worldwide each year are
attributable to CVDs, or 31% of all fatalities (Bengio & LeCun,
2007). To improve patient outcomes and lower healthcare
costs related to CVDs, early identification and effective treatment are
essential (Silver et al., 2017). However, due to the complexity of the condition and
the broad range in how patients exhibit and respond to treatment, diagnosing
and managing CVDs can be difficult. Artificial intelligence (AI) and machine
learning (ML) have become potent tools in healthcare, notably in the diagnosis
and management of CVDs. Large volumes of clinical and imaging data may be processed
by ML and AI algorithms to produce precise and individualised diagnosis, risk
stratification, and therapy planning. An overview of the current state and
potential applications of ML and AI in cardiovascular healthcare will be
provided in this review article.
The importance of timely and accurate diagnosis of
cardiovascular diseases coronary artery disease, heart failure, and stroke are
among the conditions known as cardiovascular diseases (CVDs), which affect the
heart and blood arteries. For lowering the risk of complications and improving
patient outcomes, early CVD detection and diagnosis are essential (Berner et al., 1994). However, because of the disease's complexity and the
broad range in how patients exhibit and respond to treatment, diagnosing CVDs
can be difficult.
Historically, patient history, physical examination,
lab testing, and imaging techniques have all been used to diagnose CVDs.
Although these methods have been useful in many situations, they are frequently
constrained by their subjectivity and the possibility of human mistake.
Additionally, the interpretation of imaging tests such cardiac magnetic
resonance imaging (MRI), echocardiography, and electrocardiography (ECG) can be
challenging and time-consuming, necessitating specialized knowledge and skills.
By offering precise and unbiased analysis of clinical and imaging data, ML and
AI systems can get around these constraints. For instance, ML systems can be trained
to recognize ECG signal patterns that signify heart diseases like atrial
fibrillation or myocardial infarction. Similar to this, artificial intelligence
(AI) algorithms can be used to analyse imaging images to find minute variations
in heart structure and function that might be signs of early disease (Krittanawong et al.,
2017).
ML and AI algorithms can aid in the identification of
individuals who are at a high risk of developing CVDs in addition to enhancing
the precision and speed of diagnosis. ML algorithms are able to categories
individuals into different risk groups and forecast the likelihood that they
will acquire CVDs by combining clinical and genetic data analysis (Deo, 2015). This can assist medical professionals in creating
individualised treatment programs and closely monitoring patients for
indications of illness development. Overall, the use of ML and AI to
cardiovascular healthcare has the potential to increase diagnosis precision and
timeliness, lower medical expenses, and enhance patient outcomes (Johnson et al., 2018). The incorporation of ML and AI algorithms into
current clinical workflows, the validation and generalization of machine learning
models across a range of patient populations, as well as the ethical and legal
ramifications of AI-based diagnosis and treatment planning are some of the
challenges and limitations that still need to be addressed.
Based on the background above, the purpose of this
study was to identify and analyze the role of AI in cardiovascular health care.
Research on the role of AI in cardiovascular health care has significant merit.
First, research can help improve the effectiveness of cardiovascular care by
identifying new, innovative, and efficient ways to diagnose, treat and prevent
cardiovascular disease. Second, research can help reduce the cost of
cardiovascular care by introducing more efficient and more cost-effective
technologies. Third, research can raise awareness about the potential benefits
and risks of using AI in cardiovascular care and help address the ethical and
privacy issues associated with using this technology. Fourth, research can open
new opportunities for the development of more sophisticated AI technologies
that can adapt to the specific needs of patients. Thus, research on the role of
AI in cardiovascular health care can make a significant contribution in
improving the overall health and well-being of society.
METHOD
The methodology of this review paper involved an
extensive literature review of the existing research on the topic of AI in
cardiovascular health care. We identified and analyzed relevant peer-reviewed
articles, research studies, and other sources of information related to the role
of AI in cardiovascular health care. Once the relevant literature was
identified, we conducted a thorough analysis to identify key trends,
challenges, and opportunities in the field. Our goal was to provide a
comprehensive overview of the current state of research on AI in cardiovascular
health care and identify key areas for further investigation. To ensure the
accuracy and reliability of our analysis, we used a rigorous methodology that
included a systematic search of relevant databases, a detailed analysis of the
research studies, and a critical evaluation of the findings. We also consulted
with experts in the field to gain insights into the current state of research
and best practices. In summary we would like to say the methodology of this
review paper aimed to provide a comprehensive and objective overview of the
role of AI in cardiovascular health care. Our analysis helped identify key
opportunities and challenges in this field and provided valuable insights for
researchers, clinicians, and policymakers.
RESULTS AND
DISCUSSION
Medical
Imaging and AI
The diagnosis and treatment of many medical disorders
depend heavily on medical imaging. Medical imaging has improved in
sophistication and accuracy as a result of technological advancements (Dilsizian & Siegel,
2018). However, because it depends on the knowledge and
experience of the interpreter, medical image interpretation is a time-consuming
and frequently subjective procedure. Artificial intelligence (AI) algorithms
have been created recently to help in the analysis and interpretation of data
from medical imaging (Kalinin et al., 2018). An overview of the various AI algorithms utilized in
the analysis of medical imaging will be given in this section. Both supervised
learning and unsupervised learning are common types of AI algorithms used in
the analysis of medical imaging (Cavallari & Weitzel,
2015). Using labelled training data, the algorithm is taught
to recognize particular patterns or characteristics in the images using
supervised learning. The system can learn from a series of examples in the
labelled training data, which also enables it to accurately categories fresh
photos. Contrarily, using unsupervised learning, the system must independently
find patterns or features using unlabeled training data (Sibbing et al., 2019).
Convolutional neural networks (CNNs) are one typical AI
method utilized in the examination of medical imaging. CNNs are a subset of
deep learning algorithms that are particularly beneficial for image analysis
because they can find patterns in massive amounts of data. CNNs recognize
features in the photos, such as edges or textures, using a number of layers of
filters, and then utilize these features to categories the images. For
instance, CNNs have been used to spot lung cancer on CT scans by locating tiny
nodules that human radiologists could miss. Support vector machines (SVMs) are
an additional AI algorithm that is frequently utilized in medical image
analysis. Using their features, SVMs, a form of supervised learning algorithm,
may categories images into several groups. SVMs operate by determining the best
border between various categories in the data space. SVMs have been utilized
for several imaging-related medical applications, including the detection of
breast cancer in mammograms and the prognosis of Alzheimer's disease
progression in MRI scans. Other AI algorithms utilized in medical imaging
analysis, in addition to CNNs and SVMs, include random forests, decision trees,
and k-nearest neighbors (Kitzmiller et al., 2016), (Pirmohamed et al., 2013). While k-nearest neighbors is an unsupervised learning
technique that may detect comparable images based on their characteristics,
random forests and decision trees are supervised learning algorithms that can
categories images based on their properties (Syn et al., 2018).
Although AI algorithms have the potential to completely
change the way that medical imaging analysis is done, they also have major
drawbacks. The requirement for a significant volume of high-quality training
data is one restriction (Li et al., 2020). It takes a sizable dataset of labelled photos that
faithfully depict the disease or condition of interest to train an AI
algorithm. It can be difficult to find such datasets, especially for rare
diseases or ailments. The potential for data bias is still another drawback (Shah et al., 2015). The AI system may produce inaccurate predictions or
fail to recognize specific scenarios if the training data is not representative
of the population. The application of AI algorithms to the study of medical
imaging also brings up moral and legal issues, such as patient confidentiality,
data ownership, and accountability for wrong diagnoses. Medical image analysis
could benefit from the accuracy and efficiency of AI algorithms (Shah, 2017). Support vector machines, convolutional neural networks,
and other AI techniques can be used to find patterns in medical images and
assist in the diagnosis and treatment of a variety of medical conditions (Shameer et al., 2018). However, there are also drawbacks to using AI
algorithms for medical imaging analysis, such as the requirement for a
substantial amount of high-quality training data and the possibility of bias in
the data. As a result, ongoing research and development are required to
guarantee the secure and efficient application of AI algorithms in the analysis
of medical imaging.
Application of AI in cardiovascular Health Care
In order to effectively treat and manage cardiovascular
diseases (CVDs), early detection and precise diagnosis are essential. CVDs are
a primary cause of death and disability worldwide. Artificial intelligence (AI)
and machine learning (ML) approaches have become effective tools for analyzing
sizable patient data sets and discovering patterns in illness development and
therapeutic response (Lee et al., 2017). We will examine the various uses of ML and AI in
cardiovascular healthcare in this section, including the detection and
diagnosis of CVDs, risk stratification and prediction of cardiovascular events,
personalised treatment planning, and monitoring of cardiovascular health and
disease progression (Krittanawong et al.,
2018).
Cardiovascular Diseases: Detection
and Diagnosis Using Machine Learning and AI
For the purpose of identifying and diagnosing CVDs, ML
and AI algorithms have been applied to a variety of medical imaging
investigations, such as cardiac MRI, CT scans, and echocardiography
(Przewlocka-Kosmala et al., 2019). These algorithms can examine images and spot minor
alterations in the structure and operation of the heart that might be signs of
early illness. To detect individuals with diastolic dysfunction, a disorder in
which the heart muscle stiffens and is unable to relax adequately between
contractions, for instance, and ML algorithms can analyse echocardiogram images
(Zellweger et al., 2018). Healthcare professionals can launch early therapies to
stop the progression of disease via early recognition of diastolic dysfunction.
In addition to medical imaging, ML and AI algorithms have
been used to detect and diagnose CVDs in other types of patient data, such as
electrocardiograms (ECGs). For instance, using ECG data analysis, ML algorithms
may spot patients who have arrhythmias, a disease in which the heart beats
erratically. Healthcare professionals can start early therapies to stop the
onset of more serious CVDs, like heart failure, by early detecting arrhythmias (Zellweger et al., 2014).
Prediction of Cardiovascular Events and Risk Stratification
Predictive models that can identify individuals who are
at a high risk of developing CVDs or experiencing cardiovascular events, such
as heart attacks or strokes, can be created using ML and AI algorithms. These
algorithms may examine sizable patient data sets, including demographic,
clinical, and genetic information, to find patterns in illness progression and
potential risk factors for CVDs. In order to identify patients who are at a
high risk of developing heart failure, a condition in which the heart is unable
to properly pump blood, for instance, and ML algorithms can analyse patient
data. Healthcare professionals can start early interventions to stop disease
development and lower the risk of hospitalization and mortality via early
identification of high-risk individuals.
Personalized Treatment Planning and Decision-making
Based on each patient's specific traits and stage of the
disease, ML and AI algorithms can be utilized to create individualised
treatment programs. To find patterns in illness development and therapeutic response,
these algorithms may analyse enormous databases of patient data, including
demographic, clinical, and genetic information. For individuals with
hypertension, a condition in which the blood pressure is constantly raised, ML
algorithms can examine patient data to determine the most efficient therapy.
These algorithms can determine the best treatments and dosages for specific
individuals by examining trends in disease progression and therapeutic
response, lowering the risk of side effects and enhancing treatment outcomes (Khamis et al., 2017).
Monitoring of cardiovascular health and disease progression
Finally, AI and machine learning can be used to follow
the development of diseases over time and monitor patients' cardiovascular
health. These technologies are able to provide real-time insights into patient
health and identify early warning signals of impending cardiovascular events by
analyzing data from wearable, medical imaging, and electronic health records (van Rosendael et al.,
2018). Machine learning algorithms, for instance, can be
trained to analyse electrocardiogram (ECG) data and find anomalies that could
point to a higher risk of heart attack or stroke. Similar to this, AI can be
employed to examine data from medical imaging and spot changes in the
composition or operation of the heart that might portend the beginning of
cardiovascular illness. Machine learning and AI can assist in improving patient
outcomes and lowering healthcare costs by avoiding expensive hospital stays and
emergency interventions by monitoring patients in real-time and offering early
warnings of potential health problems (van Rosendael et al.,
2018).
By enhancing the precision and effectiveness of disease
detection and diagnosis, forecasting cardiovascular events and risk,
personalizing treatment planning and decision-making, and tracking patient
health in real-time, machine learning and AI have the potential to completely
transform the field of cardiovascular healthcare. Although there are still
difficulties and restrictions to be resolved, the quick development of these
technologies suggests that they will become more crucial to cardiovascular
healthcare in the years to come (Al�Aref et al., 2019). We can work towards a future where cardiovascular
illness is recognized and treated earlier, more efficiently, and with fewer
consequences by utilizing the potential of machine learning and AI (Peng et al., 2016).
Challenges and Limitations
For machine learning and AI to be successfully
implemented in cardiovascular healthcare, a number of issues and restrictions
must be resolved (Zhou et al., 2011). We will go through some of the biggest obstacles to
machine learning and artificial intelligence in cardiovascular healthcare in
this part.
Integration with existing clinical workflows and electronic health records:
The integration of machine learning and AI with current
clinical procedures and electronic health records (EHRs) is one of the biggest
obstacles to its use in cardiovascular healthcare. Electronic health records,
or EHRs, are crucial for delivering prompt and efficient care because they
contain information about a patient's medical history, diagnosis, medications,
and other pertinent clinical data. However, the lack of standardization in EHRs
makes it challenging to incorporate machine learning techniques into the
current healthcare system. The adoption of new technology may also be hampered
by healthcare practitioners' possible resistance to change (Balanescu et al., 2018).
Validation and generalization of machine learning models in diverse patient
populations:
Validating and generalizing machine learning models
across a range of patient groups is a significant obstacle to the application
of AI and machine learning in cardiovascular healthcare. Large datasets are
necessary for machine learning algorithms to train and improve their models.
These statistics, however, frequently exhibit bias towards particular
demographics, such as age, gender, or ethnicity. This may restrict how broadly
the algorithms may be applied and produce erroneous predictions in some patient
populations (Powles & Hodson,
2017). Additionally, over fitting in machine learning models
might produce outcomes that are excessively hopeful.
Privacy and security concerns in handling patient data:
Large volumes of patient data are used by machine
learning and AI algorithms to train and improve their models. However, handling
patient data in this manner presents privacy and security issues. To avoid
unauthorized access or data breaches, healthcare organizations must make sure
that patient data is adequately de-identified and protected (Constantinides &
Fitzmaurice, 2018). In order to preserve patient privacy, healthcare
professionals must also make sure that they are adhering to laws like HIPAA
(Health Insurance Portability and Accountability Act) and GDPR (General Data
Protection Regulation).
Ethical and legal implications of AI-based diagnosis and treatment
planning:
Algorithms that use machine learning and artificial
intelligence (AI) can diagnose patients more quickly and with greater accuracy.
However, there are a number of moral and legal issues with using these
technology in healthcare. For instance, if these algorithms are being used to
make life-or-death choices, there may be issues about their openness.
Additionally, there can be issues with access to these technologies being
equally distributed, especially for vulnerable patient populations. The
liability of healthcare practitioners in the event of a negative incident or a
malpractice claim may also be a source of concern. For machine learning and AI
to be successfully implemented in cardiovascular healthcare, a number of issues
and restrictions must be resolved. These difficulties include incorporating these
tools into clinical workflows and electronic medical records as they currently
stand, validating and generalizing machine learning models across a range of
patient populations, ensuring the privacy and security of patient data, and
dealing with the moral and legal ramifications of AI-based diagnosis and
treatment. To overcome these obstacles and take use of the promise of machine
learning and AI to enhance patient outcomes and quality of care, healthcare
organizations and providers must collaborate (Laser et al., 2014).
CONCLUSION
In summary, artificial intelligence (AI)
and machine learning have the potential to revolutionize cardiovascular
healthcare in the years to come. The integration with current clinical
workflows and electronic health records, validation, and generalization of
machine learning models across a range of patient populations, and privacy and
security issues when handling patient data are just a few of the obstacles that
need to be overcome. However, the potential benefits are substantial. Further
study is required in several areas before machine learning and AI can
completely realize their potential in cardiovascular treatment. The creation of
more precise and trustworthy predictive models, generalization of these models
over a range of patient groups, and incorporation of machine learning and AI
into current clinical procedures are a few of these. To guarantee that new
technologies are created and applied in a responsible and ethical manner, there
is also a need for collaboration between physicians, researchers, and
industrial partners. This includes resolving the privacy and security issues
related to handling patient data as well as making sure that all patients,
regardless of socioeconomic situation, have access to this technology. Overall,
machine learning and AI have a bright future in cardiovascular healthcare and
hold enormous promise for raising patient outcomes and lowering costs. These
technologies have the potential to completely alter how we detect, diagnose,
and treat cardiovascular illnesses with continuous investment in research and
development.
REFERENCES
Al�Aref, S. J., Anchouche, K., Singh, G., Slomka, P. J.,
Kolli, K. K., Kumar, A., Pandey, M., Maliakal, G., Van Rosendael, A. R., &
Beecy, A. N. (2019). Clinical applications of machine learning in cardiovascular
disease and its relevance to cardiac imaging. European Heart Journal, 40(24),
1975�1986. https://doi.org/10.1093/eurheartj/ehy404
Balanescu, D. V., Monlezun, D. J., Teodora Donisan, M. D.,
David Boone, M. D., Frances Cervoni-Curet, M. D., Nicolas Palaskas, M. D., Juan
Lopez-Mattei, M. D., Peter Kim, M. D., Cezar Iliescu, M. D., & Balanescu,
S. M. (2018). A cancer paradox: machine-learning backed propensity-score
analysis of coronary angiography findings in cardio-oncology. Journal of
Invasive Cardiology, 31(1).
Cavallari, L. H., & Weitzel, K. (2015). Pharmacogenomics
in cardiology�genetics and drug response: 10 years of progress. Future
Cardiology, 11(3), 281�286.
Constantinides, P., & Fitzmaurice, D. A. (2018).
Artificial intelligence in cardiology: applications, benefits and challenges. Br
J Cardiol, 25(3), 86�87. doi:10.5837/bjc.2018.024
Deo, R. C. (2015). Machine learning in medicine. Circulation,
132(20), 1920�1930.
Dilsizian, M. E., & Siegel, E. L. (2018). Machine meets
biology: a primer on artificial intelligence in cardiology and cardiac imaging.
Current Cardiology Reports, 20, 1�7.
Johnson, K. W., Torres Soto, J., Glicksberg, B. S., Shameer,
K., Miotto, R., Ali, M., Ashley, E., & Dudley, J. T. (2018). Artificial
intelligence in cardiology. Journal of the American College of Cardiology,
71(23), 2668�2679.
Kalinin, A. A., Higgins, G. A., Reamaroon, N., Soroushmehr,
S., Allyn-Feuer, A., Dinov, I. D., Najarian, K., & Athey, B. D. (2018).
Deep learning in pharmacogenomics: from gene regulation to patient
stratification. Pharmacogenomics, 19(7), 629�650. https://doi.org/10.2217/pgs-2018-0008
Khamis, H., Zurakhov, G., Azar, V., Raz, A., Friedman, Z.,
& Adam, D. (2017). Automatic apical view classification of echocardiograms
using a discriminative learning dictionary. Medical Image Analysis, 36,
15�21. https://doi.org/10.1016/j.media.2016.10.007
Kitzmiller, J. P., Mikulik, E. B., Dauki, A. M., Murkherjee,
C., & Luzum, J. A. (2016). Pharmacogenomics of statins: understanding
susceptibility to adverse effects. Pharmacogenomics and Personalized
Medicine, 97�106.
Krittanawong, C., Bomback, A. S., Baber, U., Bangalore, S.,
Messerli, F. H., & Wilson Tang, W. H. (2018). Future direction for using
artificial intelligence to predict and manage hypertension. Current
Hypertension Reports, 20, 1�16.
Krittanawong, C., Zhang, H., Wang, Z., Aydar, M., &
Kitai, T. (2017). Artificial intelligence in precision cardiovascular medicine.
Journal of the American College of Cardiology, 69(21), 2657�2664.
Laser, K. T., Horst, J.-P., Barth, P., Kelter-Kl�pping, A.,
Haas, N. A., Burchert, W., Kececioglu, D., & K�rperich, H. (2014).
Knowledge-based reconstruction of right ventricular volumes using real-time
three-dimensional echocardiographic as well as cardiac magnetic resonance
images: comparison with a cardiac magnetic resonance standard. Journal of
the American Society of Echocardiography, 27(10), 1087�1097. https://doi.org/10.1016/j.echo.2014.05.008
Lee, K.-T., Hour, A.-L., Shia, B.-C., & Chu, P.-H.
(2017). The application and future of big database studies in cardiology: a
single-center experience. Acta Cardiologica Sinica, 33(6), 581.
Li, Q., Wang, J., Tao, H., Zhou, Q., Chen, J., Fu, B., Qin,
W., Li, D., Hou, J., & Chen, J. (2020). The prediction model of warfarin
individual maintenance dose for patients undergoing heart valve replacement,
based on the back propagation neural network. Clinical Drug Investigation,
40, 41�53. https://doi.org/10.1007/s40261-019-00850-0
Peng, P., Lekadir, K., Gooya, A., Shao, L., Petersen, S. E.,
& Frangi, A. F. (2016). A review of heart chamber segmentation for
structural and functional analysis using cardiac magnetic resonance imaging. Magnetic
Resonance Materials in Physics, Biology and Medicine, 29, 155�195. https://doi.org/10.1007/s10334-015-0521-4
Pirmohamed, M., Burnside, G., Eriksson, N., Jorgensen, A. L.,
Toh, C. H., Nicholson, T., Kesteven, P., Christersson, C., Wahlstr�m, B., &
Stafberg, C. (2013). A randomized trial of genotype-guided dosing of warfarin. N
Engl J Med, 369, 2294�2303.
Powles, J., & Hodson, H. (2017). Google DeepMind and
healthcare in an age of algorithms. Health and Technology, 7(4),
351�367. https://doi.org/10.1007/s12553-017-0179-1
Przewlocka-Kosmala, M., Marwick, T. H., Dabrowski, A., &
Kosmala, W. (2019). Contribution of cardiovascular reserve to prognostic
categories of heart failure with preserved ejection fraction: a classification
based on machine learning. Journal of the American Society of
Echocardiography, 32(5), 604�615. https://doi.org/10.1016/j.echo.2018.12.002
Shah, S. J. (2017). Precision medicine for heart failure with
preserved ejection fraction: an overview. Journal of Cardiovascular
Translational Research, 10(3), 233�244. https://doi.org/10.1007/s12265-017-9756-y
Shah, S. J., Katz, D. H., Selvaraj, S., Burke, M. A., Yancy,
C. W., Gheorghiade, M., Bonow, R. O., Huang, C.-C., & Deo, R. C. (2015).
Phenomapping for novel classification of heart failure with preserved ejection
fraction. Circulation, 131(3), 269�279.
Shameer, K., Johnson, K. W., Glicksberg, B. S., Dudley, J.
T., & Sengupta, P. P. (2018). Machine learning in cardiovascular medicine:
are we there yet? Heart, 104(14), 1156�1164.
Sibbing, D., Aradi, D., Alexopoulos, D., Ten Berg, J., Bhatt,
D. L., Bonello, L., Collet, J.-P., Cuisset, T., Franchi, F., & Gross, L.
(2019). Updated expert consensus statement on platelet function and genetic
testing for guiding P2Y12 receptor inhibitor treatment in percutaneous coronary
intervention. JACC: Cardiovascular Interventions, 12(16),
1521�1537.
Silver, D., Schrittwieser, J., Simonyan, K., Antonoglou, I.,
Huang, A., Guez, A., Hubert, T., Baker, L., Lai, M., & Bolton, A. (2017).
Mastering the game of go without human knowledge. Nature, 550(7676),
354�359.
Syn, N. L., Wong, A. L.-A., Lee, S.-C., Teoh, H.-L., Yip, J.
W. L., Seet, R., Yeo, W. T., Kristanto, W., Bee, P.-C., & Poon, L. M.
(2018). Genotype-guided versus traditional clinical dosing of warfarin in
patients of Asian ancestry: a randomized controlled trial. BMC Medicine,
16(1), 1�10.
van Rosendael, A. R., Maliakal, G., Kolli, K. K., Beecy, A.,
Al�Aref, S. J., Dwivedi, A., Singh, G., Panday, M., Kumar, A., & Ma, X.
(2018). Maximization of the usage of coronary CTA derived plaque information
using a machine learning based algorithm to improve risk stratification;
insights from the CONFIRM registry. Journal of Cardiovascular Computed
Tomography, 12(3), 204�209. https://doi.org/10.1016/j.jcct.2018.04.011
Zellweger, M. J., Brinkert, M., Bucher, U., Tsirkin, A.,
Ruff, P., & Pfisterer, M. E. (2014). A new memetic pattern based algorithm
to diagnose/exclude coronary artery disease. International Journal of
Cardiology, 174(1), 184�186. https://doi.org/10.1016/j.ijcard.2014.03.184
Zellweger, M. J., Tsirkin, A., Vasilchenko, V., Failer, M.,
Dressel, A., Kleber, M. E., Ruff, P., & M�rz, W. (2018). A new non-invasive
diagnostic tool in coronary artery disease: artificial intelligence as an
essential element of predictive, preventive, and personalized medicine. EPMA
Journal, 9, 235�247. https://doi.org/10.1007/s13167-018-0142-x
Zhou, J., Qu, F., Sang, X., Wang, X., & Nan, R. (2011).
Acupuncture and auricular acupressure in relieving menopausal hot flashes of
bilaterally ovariectomized Chinese women: a randomized controlled trial. Evidence-Based
Complementary and Alternative Medicine, 2011.
