NDOL: An Enhanced Heart Disease Prediction System Using

Artificial Intelligence (AI) Assisted Neural Decision Optimization

Logic

K. Venkatasalam, M. Desikan, S. Kalai Selvan, B. Saranraj and A. Sowkath Basha

Department of Computer Science and Engineering, Mahendra Engineering College, Namakkal, Tamil Nadu, India

Keywords: Deep Learning, Heart Disease, Artificial Intelligence, Disease Prediction, AI, Neural Decision Optimization,

NDOL, Support Vector Machine, SVM.

Abstract: Since cardiovascular diseases (CVD) remain one of the global leading causes of mortality, the urgent need

for new diagnostics and preventative therapies is evident. These are classified as a significant physical

condition leading to a high probability of death. Heart disease must then be diagnosed effectively and quickly

to avoid further harm to individuals. This study thus investigates whether a learning technique derived from

Artificial Intelligence (AI)- Neural Decision Optimization Logic (NDOL) - can be used to predict cardiac

illnesses to enable timely treatments and personalized healthcare plans. The suggested model is cross-

validated with a common learning model, Support Vector Machine (SVM) to investigate how well the

proposed model performs. ML tech, applied to vast volumes of data to identify complex patterns and

possibilities of danger from certain combinations that the naked eye cannot see, is what enables these ML

models. Some of the important technologies used to predict the probability of heart disease are neural

networks, decision trees, and ensemble learning. Also, this study further demonstrates the implementation of

artificial intelligence technologies in the clinical workflow to give timely risk assessment and improve patient

care and resource distribution. The goal of proactive and precision medicine may be within our reach through

AI-enabled heart disease prediction which will combine technical ingenuity with medical expertise to

revolutionize cardiovascular health care.

1 INTRODUCTION

The advent of artificial intelligence has

revolutionized healthcare sector especially in

diagnosis, treatment planning and prevention

Archana Singh, et al., 2020). Predicting

cardiovascular disease is one of the most promising

applications for artificial intelligence. We cannot

stress enough the need for early identification and

appropriate management, as heart disease remains the

leading cause of death worldwide. Clinical exams,

imaging and blood testing are the hallmarks of

traditional heart disease diagnosis; however, those

tools don’t always catch threats early enough to

prevent a fatal event. This is where AI-powered

systems come into their own. AI-Powered Heart

Disease Forewarning: By applying machine learning

(ML) algorithms, this technology analyses extensive

and diverse datasets, such as clinical data, past

medical history, test results, genetic markers, and

lifestyle patterns, to identify trends that may indicate

a greater risk of cardiovascular conditions. AI data-

driven heart condition prediction enables early risk

detection of heart diseases so that care providers

initiate preventative therapeutic treatments or

recommend lifestyle modifications for these patients

Artificial intelligence methods used to predict

cardiac complications are brain networks, ensemble

learning, decision trees, neural networks.] These

models can manage and analyse complex data which

often uncover latent associations, unlike more

traditional techniques. Correlations between

predictors such as age, prior smoking, blood pressure,

cholesterol, and diabetes create a fuller picture of the

individual’s risk profile. After training, these artificial

intelligence (AI) models demonstrate remarkable

precision in predicting the risk of the cardiovascular

disease. AI systems continuously learn and improve

over time as they are exposed to new data, ensuring

that. In addition, these models analyse huge amounts

672

Venkatasalam, K., Desikan, M., Selvan, S. K., Saranraj, B. and Basha, A. S.

NDOL: An Enhanced Heart Disease Prediction System Using Artiﬁcial Intelligence (AI) Assisted Neural Decision Optimization Logic.

DOI: 10.5220/0013871100004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 1, pages

672-679

ISBN: 978-989-758-777-1

of data at once, allowing doctors to act promptly for

critical patients (

Hosam El-Sofany, et al., 2024). AI-

powered heart disease prediction systems have the

potential to greatly enhance outcomes by enabling

early intervention, reducing healthcare costs, and

providing a more personalized approach to

cardiovascular health. The incorporation of the AI in

clinical practices would allow pre-emptive screening

and diagnosis of cardiac ailments, making it possible

for practitioners to make informed clinical decisions

in the future 7.

The identification of these issues early on

provides a means of preventing long-term health

problems, including heart disease, the leading cause

of death in the world. Traditional methods of spotting

heart disease are not always speedy or dependable

because they can overlook early symptoms or

understated signs. For improvements, AI-based

cardiac disease prediction tools are essential for

faster and accurate assessments. These algorithms

might breezily scour mountains of data a patient’s

medical records, their lifestyle choices, their test

results to reveal hidden threats. This enables cardiac

disease to be diagnosed in its earliest stages before

encountering the symptoms when the doctors can take

steps to intervene. AI also helps tailor therapy, as it

can factor in individual risk factors and provide

personalized advice. And by automating routine but

necessary processes like risk assessments, AI

systems might help free clinicians’ time for patients

with more acute conditions. Not just that, some of

these AI-powered devices might work in real-time,

monitoring a patient’s vitals and altering their risk

profile as appropriate. Saving lives, making

healthcare much more efficient and better managing

cardiac disease across the Globe is also what this

kind of technology, 9.

Artificial intelligence-based heart disease

prediction primarily aims to create a reliable

technique to predict a person' risk of developing

cardiovascular disease. One of the goals is to create

an AI model capable of predicting the onset of

cardiovascular disease on the basis of data acquired

from multiple sources such as medical history,

lifestyle habits, and test findings. The point of this

approach is to detect any cardiac issues before

symptoms appear, so that treatment can start sooner

rather than later. The system will generate risk

assessments that, based on individual patient medical

history, can be used to determine customized

treatment plans. Once incorporated into processes,

healthcare providers will have much ease in

employing it during routine health checks. The

system automates the risk evaluation process,

enabling medical personnel to spend more time and

effort on the highest risk profile patients. The end

goal is improved forecasts, lighter workloads for

healthcare workers and better, more efficient

treatment.

2 RELATED WORKS

At this time, cardiovascular disease ranks as the

leading death globally. It is difficult to predict cardiac

illness since it requires expertise and specific

information (

Padmakumari Pitchal, et al., 2024). Medical

facilities have just recently started collecting sensor

data using Internet of Things (IoT) technologies to

improve cardiac disease diagnosis and prognosis.

The results may not be trustworthy, despite the

extensive research on heart disease diagnosis. The

three primary steps of the automated model for

predicting cardiac problems—preprocessing, feature

extraction and prediction—are laid forth in this article.

The input data is preprocessed by using an upgraded

Z-score normalization. In order to train the prediction

model, feature extraction is used to get the important

features from the preprocessed data. Statistical

features, information gain characteristics, and

enhanced entropy are among the features that were

retrieved. The Improved Quantum Convolutional

Neural Network (IQCNN) uses the retrieved

characteristics to make predictions. We compare the

IQCNN's performance against that of previous

systems using a number of parameters. The proposed

IQCNN model achieves a learning rate of 70% and an

accuracy of 0.91 when compared to more

conventional methods for predicting cardiac

problems.

Each year, millions of people die from heart

disease, making it one of the most recognized and fatal

diseases in the world (

Ahmad Ayid Ahmad, et al., 2023).

The only avenue to save people's lives is through early

detection of this disease. Machine Learning (ML) is

an artificial intelligence technique that can diagnose

diseases quickly, easily, and inexpensively. Our

objective is to create a machine learning model that

can accurately forecast the onset of heart disease by

analyzing the Cleveland heart disease dataset. Model

performance is very sensitive to the properties of the

training dataset and the ML technique chosen. The

Jellyfish optimization approach was used to reduce the

Cleveland dataset to a lower dimensional subspace in

order to avoid overfitting. Overfitting is caused by the

curse of dimensionality and happens when a dataset

has too many characteristics. Finding the optimal

NDOL: An Enhanced Heart Disease Prediction System Using Artiﬁcial Intelligence (AI) Assisted Neural Decision Optimization Logic

673

features is made easy with the Jellyfish algorithm's

fast convergence speed and flexibility.

According to the World Health Organization,

cardiac-related diseases have increased. Therefore,

each year 17.9 million people die (

Vijeta Sharma, et al,

2020). Detecting and treating these patients earlier is

getting more difficult with the growing population. On

the other hand, many studies have shown that the

recent growth in technology has caused machine

learning techniques to accelerate the health-care field.

Hence, the purpose of this work is to build a machine

learning model for the prediction of heart disease

utilizing these significant characteristics. The heart

disease prediction dataset at UCI served as a standard

for this research; it contains fourteen separate

characteristics related to cardiovascular disease.

While building the model, many machine learning

approaches were employed, including Decision Tree,

Naive Bayes, Support Vector Machine (SVM), and

Random Forest. As part of our study, we utilized

traditional Machine Learning methods to identify

correlations between the dataset's many properties,

with the goal of applying these findings to the

prediction of heart disease risk. As compared to other

ML approaches, Random Forest provides more

accurate predictions in less time, according to the

results. As a decision-support system, this model can

be useful for doctors in the clinic.

In the last several decades, cardiovascular illness

(heart disease) has become the leading cause of

mortality worldwide (

Devansh Shah, et al., 2020). It

includes a broad variety of cardiac conditions. There

are a lot of things that may go wrong with a heart

attack, and it's critical that we find ways to diagnose

the condition quickly so that we can start treating it

effectively. Healthcare organizations often use data

mining as a method for coping with large data sets. In

order to aid medical professionals in the prediction of

heart illness, researchers examine large medical data

sets using various data mining and machine learning

methods. This research study's model exhibits several

characteristics linked to heart illness; it is constructed

using supervised learning techniques such Naïve

Bayes, decision trees, K-nearest neighbors, and

random forest. It draws on the Cleveland database at

UCI, which already has information on people with

cardiac disease. With 303 cases and 76 characteristics,

the data is rather extensive. We can actually evaluate

fourteen of those seventy-six attributes the ones that

matter most for comparing algorithm performance.

This study aims to assess the potential occurrence of

cardiovascular disease in individuals. The results

demonstrate that K-nearest neighbor offers the highest

level of accuracy.

Important medical duties include cardiovascular

disease diagnosis and prognosis to guarantee accurate

categorization, which aids cardiologists in treating

patients appropriately (

Chintan M. Bhatt, et al., 2023).

The ability of machine learning to identify patterns in

data has led to an upsurge in its use in the medical

field. To help diagnosticians decrease misdiagnosis,

machine learning may be used to categorize the

occurrence of cardiovascular illness. In an effort to

lower the death toll from cardiovascular disorders, this

study builds a model that can accurately forecast these

conditions. In order to enhance classification

accuracy, this research suggests a k-modes clustering

algorithm that starts with Huang. We employ models

like XGBoost, multilayer perceptron, decision tree

classifier and random forest. In order to get the best

possible outcome, the parameters of the applied model

were hyper-tuned using GridSearchCV. We test the

suggested model on a Kaggle dataset with 70,000 real-

world examples. Here is how the models were trained

using an 80:20 split of data and how they attained

accuracy: In the decision tree model, 86.37% of the

trials used cross-validation, while 86.53% did not. In

the XGBoost model, 87.12% of the trials used cross-

validation, while 87.05% used random forest. In the

multilayer perceptron model, 87.28% used cross-

validation, while 86.44% used non-validation. The

models that have been suggested have AUC (area

under the curve) values: XGBoost: 0.95, decision tree:

0.94, random forest: 0.95, multilayer perceptron: 0.95.

Based on these foundational studies, we know that

multilayer perceptron with cross-validation is the most

accurate method currently available. With an accuracy

of 87.28%, it was the most accurate.

3 METHODOLOGY

The significance of AI-based heart disease prediction

is Early prediction of heart disease even before people

develop any serious symptoms. When it is diagnosed

early, patients may have a better treatment experience

and better outcomes than if diagnosed late. AI can

assist doctors in more accurately assessing a person’s

risk for heart disease by analysing a patient’s history

and other lifestyle factors. This technology will also

improve the effect of health care, as it can accelerate

the initial assessment of risk, and enable providers to

focus on patients determined to have the greatest

need. This technology is integrated seamlessly within

healthcare systems, would provide efficiencies for

hospitals and doctors. With AI system powered

heart disease prediction services, more loved ones can

be reached at varying environmental conditions

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

674

including under-resourced settings. Artificial

intelligence and machine learning have been applied

to the health sector to improve diagnosis, treatment,

and patient care in various forms. These algorithms

can whittle down huge databases in search of patterns

and trends, offering the potential for better healthcare

decision making.

Deep learning and AI have many possible

applications in healthcare from disease prediction to

therapy personalization and automating tedious tasks.

Deep learning algorithms, for example, can often

identify abnormalities such as tumours in X-rays and

MRIs significantly faster than human doctors. AI can

analyse factors such as blood pressure, cholesterol

and lifestyle to predict diabetes, heart disease and

other chronic diseases. Deep learning and artificial

intelligence (AI), when used in the context of the

health-care industry, could significantly reduce

processing times for high-volume patient datasets.

This will be extremely useful in risk assessment and

suggesting preventive measures. These technologies

also aid in personalized medicine by enabling

therapies targeted to the needs of the patient through

the analysis of their hereditary make-up and medical

journey. Despite the promise of AI-assisted

diagnostics, there are significant barriers to adoption,

including the need for reliable datasets and the need

to address patient privacy, as well as the issue of

black-box AI models. However, AI and ML could

transform healthcare to a more organized, accurate,

and individualized experience, leading to better

outcomes at lower costs.

Over the years, heart disease risk prediction has

improved with the development of new risk

assessments tools. Traditional risk factors, such as

age, blood pressure, cholesterol, and smoking habits,

are handled by models like the Framingham Risk

Score. Statistical methods like logistic regression

work, but can miss complex patterns in the data.

Deep learning and deep learning methods combine

big data to detect relationships that no one has

observed before, yielding better predictions. Real-

time health data (now available through smartwatches

and other wearables) makes it much easier to monitor

your health and catch any potential problems early.

These advanced methods of modelling significantly

enhance cardiac disease prevention and treatment

when compared to previous models. Predicting

cardiovascular disease is important, but not without

hurdles. Data quality is a major concern simply due

to the prevalence of erroneous or missing records in

the medical record. These include heredity, lifestyle,

and environment, all significant contributors to heart

disease but are difficult to accurately represent in

models. One complex model that performs well, but

is rarely employed in the clinic, is deep learning,

because it is simply too difficult for doctors to

understand. The second part of the problem of data

bias is that models trained on specific, rather than

diverse, populations can have results that will be

incorrect. And smaller clinics may not be able to

afford the pricey, cutting-edge models, and there are

concerns over the privacy and security of patient

data. Finally, a majority of models are trained on

outdated data, even though a user’s health can

change as time goes on. There are many problems that

still need to be addressed for predictive models for

heart diseases to become available and reliable.

AI-based heart disease prediction systems still

have numerous unresolved matters. Firstly, the lack

of diversity in training datasets results in many

models’ performing poorly when they are deployed in

populations in the real world. The integration of these

tools into pre-existing hospital workflows is also not

easy. To make it worse, doctors may be reluctant to

trust AI predictions if they’re based on models that

are hard to understand.

It is common for existing systems to depend on

static data, which fails to consider how a patient's

health evolves over time. Concerns around patient

data privacy and ethics have not been addressed

either. Having said that, a lot of room for growth

exists and improving the accuracy of models for

various groups of individuals may be achieved

through the use of distinct datasets. Through the

utilization of AI, wearable gadgets may offer the

convenience of real-time health monitoring.

Increasing confidence among healthcare practitioners

can be achieved by making AI forecasts more

understandable. Personalized treatment regimens for

patients can also be assisted by AI. One last thing that

can be done to enhance healthcare outcomes is to

create systems that collaborate with doctors. The

following figure 1 shows the system flow design and

the following figure 2 shows the architectural

diagram.

Filling such gaps and pursuing such opportunities

can improve and make artificial intelligence systems

more widely accepted. There are several steps

involved in using AI to predict the incidence of

cardiovascular disease. Data cleaning means combing

through databases and medical records for errors and

missing information. First the research team needs to

decide which of the variables age, cholesterol, blood

pressure, etc. deserve most of the researchers’

attention. This data can be useful in training deep

learning models such as neural networks and decision

trees. Numerous performance metrics such as

NDOL: An Enhanced Heart Disease Prediction System Using Artiﬁcial Intelligence (AI) Assisted Neural Decision Optimization Logic

675

precision and accuracy are used to valid and tune the

model to ensure that it is as accurate as possible.

Figure 1: System Flow Design.

Figure 2: Architectural Diagram.

Finally, the model is built into a system that

doctors can use to evaluate possible risks to their

patients. This method provides assurance of the

system's reliability and usefulness in predicting heart

disease. The proposed methods NDOL achieves

better prognostication with respect to CVDs over

present SVM and enhance issues due to imbalanced

data. The proposed system, NDOL, is an AI-driven

early prediction system for cardiovascular disease

that would help doctors to more accurately assess the

risk for their patients. To assess vital signs such as

age, cholesterol and blood pressure, the system first

collects data from the patients’ medical records and

wearables, cleans it up to eliminate errors, before

deploying it. The system has a very user-friendly user

interface which helps Doctors as well as patients to

easily view results and insights. In addition, it can

streamline processes by integrating with other

healthcare systems. This method can help improve

patient outcomes in the form of early diagnosis,

tailored treatment plans, and the management of

cardiac disease.

• Data Collection and Preprocessing: To

construct a system for the prediction of heart

disease, data collection is the initial stage.

Information is collected from a wide range of

resources, including health records,

questionnaires, publicly accessible databases,

and wearable technology. Some examples of

the types of information included in this data

set are gender, age, BP, cholesterol, heart rate,

and lifestyle choices. For reliable forecasts,

you need data that is both varied and of high

quality.

• Data Preprocessing: Before data can be

analyzed, it must first be cleaned and

structured. Cleaning and structuring comprise

the following steps: Statistical methods are

applied to impute missing data or otherwise

remove incomplete records from the dataset,

when missing data has been identified;

Searching for and removing inconsistent data

records; Normalizing values, when possible,

to make numerical values more comparable

with deep learning algorithms; and Encoding

categorical (non-numerical) data, such as

gender or lifestyle choice, into numerical

representations that the model can understand.

Pre-processing improves the prediction

model’s accuracy and reliability because it

ensures the dataset is clean, consistent, and

ready to train on.

• Feature selection and Data

Transformation: An essential part of

developing a system to anticipate cardiac

problems is feature selection. The first step is

to sort the dataset by the variables that affect

the likelihood of heart disease. Type 2

diabetes, smoking, exercise, cholesterol, blood

pressure, heart rate, and gender are all

common characteristics. The accuracy of the

model and the reduction of complexity caused

by the elimination of extraneous data are both

enhanced by selecting useful characteristics.

Now that we are looking at the selected,

previously discovered features, it is very

important that the data is transformed into

shapes that are suitable for deep learning

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

676

algorithms. This also includes Building

scalable and Normalization, which is basically

the process of outscoring the numerical

variables (cholesterol values and blood

pressure etc) so as to enhance the efficacy of

the model. Encoding of categorical variables

refers to converting categorical non-

numerical variables, such as gender or lifestyle

features, to numerical values, by techniques

such as one-hot-encoding. Extract or combine

features to improve the representation of a set

of data. An example of this is using age and

body mass index (BMI) to predict potential

health problems and determining a more

effective and accurate risk prediction in heart

disease by selection and transformation of

relevant features.

• Developing the Model: The model building

process consists of several simple steps. The

first of which is aggregating data from

multiple touchpoints such as health devices

and hospital data. The data is than cleansed for

any errors present and it is also formatted in

order to make it prepared for analyzing. Older

age and cholesterol values are identified as

features to be included in the model. Deep

learning algorithms take this data as an input

to train the model. After training the model,

fine tuning is done to make predictions more

accurate. Finally, the model is validated on

new data to be sure it is accurate and how good

the performance and reliability are. Such a

method can be employed to build a model

which accurately predicts whether or not a

person is likely to develop heart disease.

• Performance Measures: Evaluation

measures are used to evaluate the usefulness

of the model for predicting cardiac events.

Accuracy shows the total number of correct

predictions. By observing the precision we

can know how accurate the positive

predictions (heart disease patients) are. Recall

measures how well the model is capturing

true positives.

4 RESULTS AND DISCUSSION

The use of NDOL and support vector machines

(SVMs) to analyze complicated medical data for the

purpose of heart disease prediction has greatly

advanced the field of AI, which in turn has improved

patient outcomes and early detection. One example is

a research that suggested a new deep learning

architecture for determining if a person has heart

disease or not by utilizing optimization logics.

Compared to more conventional classification

algorithms, such as SVM, our model outperformed

them with a total prediction accuracy of more than

98.39%. Research has demonstrated that deep

learning models' prediction capabilities may be

significantly improved by utilizing feature

augmentation approaches. When combined with deep

learning methods, these approaches allowed

researchers to achieve a level of accuracy that was

noticeably higher than that of earlier approaches. This

paper's literature review sections emphasize the

significance of model selection and the integration of

diverse data sources to improve predictive accuracy,

and they also highlight the effectiveness of different

deep learning-based models in heart disease

prediction. Developing models with high accuracy

and reliability in heart disease prediction has been

made possible by integrating AI and deep learning. In

the long run, these innovations should lead to better

cardiovascular health outcomes for patients by

facilitating earlier diagnosis and treatment. Figures 3

and 4 show the proposed scheme's home page and

about us page designs, respectively.

Figure 3: Home Page.

The proposed scheme's health data gathering

portal and output prediction page architecture are

shown in Figures 5 and 6, respectively.

NDOL: An Enhanced Heart Disease Prediction System Using Artiﬁcial Intelligence (AI) Assisted Neural Decision Optimization Logic

677

Figure 4: About Us Page.

Figure 5: Health Data Collection Portal.

Figure 6: Output Prediction.

As shown in Figure 7, the proposed approach,

NDOL, is cross-validated with the traditional learning

model, Support Vector Machine (SVM), to see how

accurate it is in making predictions. Table 1 provides

a descriptive representation of the same.

Table 1: Comparison of Prediction Accuracy Between Svm and

Ndol.

No. of Days Tested SVM (%) NDOL (%)

5 77.12 98.09

7 78.54 97.71

10 77.64 96.63

14 79.52 97.42

15 79.89 98.39

18 76.17 97.48

27 74.45 96.57

29 75.25 96.26

33 76.44 97.19

36 76.12 98.14

Figure 7: Prediction Accuracy.

5 CONCLUSION AND FUTURE

SCOPE

Artificial intelligence (AI) offers enormous promise

to revolutionize healthcare with its application to the

prediction of cardiac disease. Artificial intelligence

models can aid in the early identification and

individualized therapy of cardiac disease by

providing predictions that are accurate, fast, and

consistent using sophisticated algorithms. Optimizing

resource allocation and overall care delivery are two

additional benefits of these systems that aid

healthcare practitioners in identifying high-risk

patients. AI's capacity to quickly and accurately

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

678

analyse huge amounts of data makes better clinical

decisions, with less room for human error and overall

better quality of care, possible. Additionally, its

scalability allows it to be employed in various

healthcare settings, from top hospitals to underserved

areas, thus broadening the reach of healthcare

globally. However, persistent oversight,

collaboration between medical professionals and

data scientists, and a focus on data security and

privacy are all essential for the successful integration

of AI into the healthcare system. AI heart disease

prediction: A promising need in heart disease

prediction leads to early diagnosis, improved patient

outcomes, and a revolutionized healthcare system.

Last but not least, AI-based prediction models of

cardiovascular disease have the potential to disrupt

the health-care system by providing more accessible,

faster, and accurate tools for prevention and

management of heart disease. Proper practice and

further development of AI have the ability to change

the approach to the treatment of cardiovascular

diseases. The model has paid off, in terms of correctly

and quickly predicting heart disease risk. It helps

healthcare practitioners to identify high risk

individuals at an early stage, through early detection.

The methodology enables better risk stratification to

categorize patients risk level clearly. It is faster and

yields more accurate predictions than more traditional

methods. Plus, it's scalable, so it can accommodate

large datasets, which is a benefit for all types of

healthcare organizations. The model is an important

tool in fighting and treating cardiovascular disease

because its findings are reliable and it can process

data quickly.

Moving forward, AI will enable continuous

monitoring of patients through wearables, helping to

assess heart disease risk in real time, and accelerating

the process to intervene if needed. These models will

also deliver more precise risk assessments, which will

assist doctors, aiming at high-risk patients and

allowing them to make educated decisions in regards

to their treatment options. Healthcare providers must

keep abreast of these latest advancements to harness

the full promise of AI to drive better diagnosis and

treatment of heart disease.

REFERENCES

Aashish Gnanavelu, et al., "Cardiovascular Disease

Prediction Using Machine Learning Metrics", Journal

of Young Pharmacists, 2025.

Ahmad Ayid Ahmad, et al., "Prediction of Heart Disease

Based on Machine Learning Using Jellyfish

Optimization Algorithm", Diagnostics, 2023.

Archana Singh, et al., "Heart Disease Prediction Using

Machine Learning Algorithms", International

Conference on Electrical and Electronics Engineering,

2020.

Chaimaa Boukhatem, et al., "Heart Disease Prediction

Using Machine Learning", Advances in Science and

Engineering Technology International Conferences,

2022.

Chintan M. Bhatt, et al., "Effective Heart Disease

Prediction Using Machine Learning Techniques",

Algorithms, 2023.

Dengqing Zhang, et al., "Heart Disease Prediction Based on

the Embedded Feature Selection Method and Deep

Neural Network", Journal of Healthcare Engineering,

2021.

Devansh Shah, et al., "Heart Disease Prediction using

Machine Learning Techniques", SN Computer Science,

2020.

Hosam El-Sofany, et al., "A proposed technique for

predicting heart disease using machine learning

algorithms and an explainable AI method", Scientific

Reports, 2024.

Kaushalya Dissanayake, et al., "Comparative Study on

Heart Disease Prediction Using Feature Selection

Techniques on Classification Algorithms", Applied

Computational Intelligence and Soft Computing, 2021.

M. Kavitha, et al., "Heart Disease Prediction using Hybrid

machine Learning Model", 6th International

Conference on Inventive Computation Technologies,

2021.

Md Mamun Ali, et al., "Heart disease prediction using

supervised machine learning algorithms: Performance

analysis and comparison", Computers in Biology and

Medicine, 2021.

Mohammed Amine Bouqentar, et al., "Early heart disease

prediction using feature engineering and machine

learning algorithms", Heliyon, 2024.

Padmakumari Pitchal, et al., "Heart disease prediction:

Improved quantum convolutional neural network and

enhanced features", Expert Systems with Applications,

2024.

Rohit Bharti, et al., "Prediction of Heart Disease Using a

Combination of Machine Learning and Deep

Learning", Computational Intelligence and

Neuroscience, 2021.

Vijeta Sharma, et al., "Heart Disease Prediction using

Machine Learning Techniques", 2nd International

Conference on Advances in Computing,

Communication Control and Networking, 2020.

NDOL: An Enhanced Heart Disease Prediction System Using Artiﬁcial Intelligence (AI) Assisted Neural Decision Optimization Logic

679