Introduction

The healthcare industry is among the most profoundly impacted by cutting-edge technologies, particularly artificial intelligence (AI), which remains at the vanguard of technological innovation [1]. As healthcare challenges intensify globally exacerbated by aging populations, rising costs, and pandemics, AI technologies are promising to enhance disease detection, optimize service delivery, and accelerate drug discovery [2]. AI encompasses a range of technologies, including machine learning (ML), natural language processing (NLP), computer vision, and advanced computing, all of which are revolutionizing various facets of medicine. The growing need for greater efficiency, accuracy, and personalized care in healthcare has driven the exploration of AI's potential applications [3]. AI is transforming every facet of medicine, from disease diagnosis to personalized treatment plans, providing unparalleled opportunities to enhance patient outcomes and optimize healthcare delivery. AI-powered systems, leveraging ML algorithms, can swiftly and accurately analyse vast datasets, including genetic, physiological, imaging, and electronic healthcare records (EHRs) far exceeding human capabilities [1]. These technologies utilize algorithms to analyse medical images, genetic data, EHRs, and patient histories, helping clinicians diagnose conditions more quickly, accurately, and efficiently [4]. In specialties such as radiology, pathology, and genomics, AI and ML significantly enhance the interpretation of medical data, enabling more accurate, efficient, and cost-effective healthcare solutions.

Significance of AI and ML in Modern Diagnostics

AI and ML technologies are particularly transformative in medical diagnostics. Traditional diagnostic methods, which largely rely on human expertise and standard diagnostic tools, have faced significant challenges, including limitations in accuracy, time efficiency, and the potential for human error [5]. These issues not only expose the vulnerabilities of existing healthcare systems but also highlight the urgent need for innovative solutions to enhance care delivery and improve administrative processes. AI and ML technologies (Table 1) overcome these limitations by facilitating the fast processing of large, diverse datasets, which aids in early detection, disease prediction, and clinical decision-making. By utilizing advanced algorithms capable of analyzing medical images, genomic data, and patient histories, AI-driven systems assist clinicians in diagnosing conditions with greater speed and precision [4]. In fields such as radiology, pathology, and genomics, AI and ML enable the detection of subtle patterns and abnormalities that might otherwise be overlooked, thereby improving diagnostic accuracy and reducing the risk of misdiagnosis [6]. AI is playing a crucial role in optimizing drug discovery and treatment selection, offering a new era of personalized medicine where interventions are tailored to individual patient profiles [7]. The integration of AI in diagnostics is enhancing the accuracy of medical assessments while also streamlining healthcare workflows, leading to improved patient outcomes and accelerated innovation in healthcare [8]. This comprehensive review aims to explore the transformative role of AI and ML technologies in the field of medical diagnosis.

AI and ML in Cancer Diagnosis

Breast cancer

AI and ML tools have significantly advanced breast cancer detection and diagnosis by enhancing the performance of traditional methods [27]. ImageChecker M1000, a traditional CAD system, uses computer vision for mammography pattern recognition [28] but has limitations in accuracy. In contrast, deep learning solutions from initiatives like the dream challenge, along with companies such as Therapixel and Kheiron Medical Technologies, leverage convolutional neural networks (CNNs) to improve breast cancer screening. There is also growing development in AI tools for other imaging modalities such as ultrasound (US) and magnetic resonance imaging (MRI), further improving diagnostic accuracy. These AI/ML tools are applied across four key imaging modalities—mammography (MG), US, MRI, and histopathology (HP) to support tasks like screening, detection, segmentation, and classification [29, 30].

Lung cancer

AI and ML tools have shown substantial promise in lung cancer diagnosis, treatment, and prognosis [31]. Technologies such as Artificial Neural Networks (ANN), Support Vector Machines (SVM), and Random Forest Neural Networks (RFNN) have been applied to accurately differentiate between

malignant and benign lung lesions, identify subtypes such as small-cell and non-small-cell lung cancer, and enhance imaging analysis, often surpassing traditional methods [32]. DL models, especially those based on CT imaging, have demonstrated high accuracy in early-stage lung cancer detection [33].

Table 1: Characteristics of AI/ML models

Abbreviations: CNNs: Convolutional Neural Networks, DL: Deep Learning, SVMs: Support Vector Machines, ANNs: Artificial Neural Networks, RFNNs: Random Forest Neural Networks, ML: Machine Learning, EHRs: Electronic Health Records, RF: Random Forest, RNNs: Recurrent Neural Networks, MRI: Magnetic Resonance Imaging, CT: Computed Tomography, ECGs: Electrocardiograms, COVID-19: Coronavirus Disease 2019.

AI is also playing a pivotal role in predicting treatment responses, with models that integrate CT-based radiomic features and machine learning algorithms predicting patient responses to therapies such as nivolumab and gefitinib, as well as responses to EGFR-tyrosine kinase inhibitors in NSCLC patients [34]. Radiomics-based AI models have further shown potential in predicting PD-L1 expression levels, progression-free survival, and chemotherapy outcomes, helping guide treatment decisions [35]. Additionally, AI tools such as recurrent neural networks (RNN) are being used to track tumor progression over time by analyzing longitudinal imaging data, providing valuable insights into treatment response and tumor changes. While these AI models show great promise, further optimization and real-world validation are necessary for their broader clinical implementation, ultimately improving the precision and personalization of lung cancer care.

Skin cancer

AL tools has been emrged for skin cancer diagnosis and treatment. ML models assist in predicting patient outcomes, personalizing treatment plans, and even identifying drug candidates [36]. AI and machine learning tools are revolutionizing skin cancer detection and treatment by improving

diagnostic accuracy and aiding early detection. Examples of these tools include SkinVision, a mobile app that assesses the risk of skin lesions being malignant, and DeepDerm, a DL model for classifying skin lesions [37, 38]. MelanomaAI by IBM Watson Health and Google Health’s AI model also use image recognition to identify melanomas with high accuracy [39]. Additionally, VisualDx assists clinicians by providing differential diagnoses for skin conditions, including malignancies [40]. These tools enhance dermatology practices by supporting healthcare professionals in making faster, more accurate diagnoses, ultimately improving patient outcomes.

Cardiovascular Diseases

AI in Heart Disease Detection and Monitoring

AI and ML are transforming heart disease detection and monitoring by enhancing diagnostic accuracy, enabling early detection, and improving personalized treatment [41]. AI algorithms, particularly DL models applied to electrocardiograms (ECGs), allow for rapid and precise analysis of heart function, detecting previously undetectable conditions like atrial fibrillation and left ventricular dysfunction [42]. In heart failure (HF), AI integrates with digital health tools to monitor patients in real time, predicting exacerbations and enabling timely interventions [43]. AI also plays a crucial role in risk stratification, identifying high-risk patients and guiding preventive care [44]. Despite challenges like data quality, algorithm transparency, and regulatory issues, AI has the potential to revolutionize cardiovascular care, improving early diagnosis, reducing healthcare costs, and ensuring more effective, targeted treatments. As AI technologies evolve, they promise to significantly enhance heart disease management and patient outcomes [45].

Machine Learning Models for Predicting Cardiac Events

ML models have shown significant promise in predicting cardiovascular events, offering potential improvements over traditional statistical methods, while statistical models like logistic regression have widely been used for event prediction [46]. ML methods are more effective in identifying complex patterns and making individual-level predictions, without requiring assumptions about the sample or population [47]. ML algorithms, such as decision trees, random forests, support vector machines, and neural networks, have been used to create prognostic models for HF and other cardiovascular diseases (CVDs) [48]. These models help clinicians predict risks like heart failure events, mortality, and readmissions, facilitating early intervention and personalized treatment [48]. ML can analyze vast amounts of clinical data, including patient demographics, medical histories, and lab results, to predict the likelihood of cardiovascular events, improving patient outcomes [49, 50]. The application of ML in cardiovascular disease prediction is rapidly growing, particularly in areas like heart failure, where the complexity of the disease requires advanced models for accurate prognosis.

Echocardiography and AI Integration

Echocardiography plays a crucial role in diagnosing and managing cardiovascular diseases, but its interpretation is often subjective and prone to inter-operator variability, leading to potential diagnostic errors [51]. AI, particularly ML and DL, offers significant potential to enhance echocardiography by providing consistent, accurate, and automated interpretation of echocardiograms, thereby reducing human error [52]. AI algorithms, such as convolutional neural networks (CNNs), have shown the ability to accurately identify and diagnose a range of cardiac pathologies, including hypertrophic cardiomyopathy, pulmonary hypertension, and cardiac amyloidosis, with performance on par with or exceeding that of expert clinicians [50] These models also automate tasks like view identification, image segmentation, and cardiac chamber quantification, making echocardiography more efficient and accessible [53]. However, the widespread adoption of AI in echocardiography is still in its early stages, and significant challenges remain. These include the need for large-scale, outcome-focused clinical trials to confirm AI's real-world impact, as well as addressing legal, ethical, and regulatory concerns [54]. Additionally, while AI has the potential to reduce clinician workload and address the growing demand for cardiovascular care, its integration into clinical practice requires careful refinement and validation.

Neurological Disorders

AI Applications in Neuroimaging

AI, particularly DL, is revolutionizing the field of neuroimaging, enhancing diagnostic capabilities, image quality, and clinical workflows [45].

DL, known for its superior feature extraction and classification abilities, has demonstrated significant improvements in detecting and diagnosing various neurological conditions, such as brain metastases, glioblastoma mutations, and stroke [55]. AI algorithms, including CNNs, are increasingly applied to automate the detection and segmentation of anatomical structures and lesions, reducing the burden on neuro-radiologists and increasing diagnostic efficiency [56]. Moreover, AI has shown promise in improving image quality by removing artifacts, harmonizing images, and reducing radiation exposure, leading to enhanced patient care [57]. A key area of development is radiomics, where AI models extract quantitative features from imaging data, providing insights beyond human perception. These features, when combined with clinical and molecular data, are helping in non-invasive disease prediction and prognosis. Additionally, AI’s potential to predict genetic alterations and tumor behavior, particularly in brain tumors like meningiomas, marks a significant advancement in personalized medicine. While challenges such as model interpretability and regulatory hurdles remain, AI’s transformative impact on neuroimaging continues to expand, offering new opportunities for early diagnosis, improved clinical outcomes, and optimized healthcare delivery.

ML Algorithms for Early Detection of Neurodegenerative Diseases

ML algorithms are transforming the early detection, diagnosis, and prognosis of neurodegenerative diseases (NDs) like Alzheimer's Disease (AD) and Parkinson's Disease (PD) [58]. These diseases often progress without noticeable symptoms, leading to irreversible neuronal damage before clinical signs appear. ML, particularly through techniques like CNNs, has shown significant promise in analyzing digital biomarkers such as eye tracking and facial expression changes [58], achieving high diagnostic accuracies (up to 0.88 ROC-AUC for PD detection). ML is also being applied to neuroimaging, motor function, and language analysis, reducing the time required for clinical assessments and improving patient stratification. Studies have demonstrated ML's potential for multiclass disease diagnosis, achieving up to 87.89% classification accuracy [59]. However, challenges such as data accuracy, privacy issues, and the need for robust validation remain, hindering full integration into clinical practice. Despite these obstacles, ML's growing capabilities and integration with digital health tools promise to revolutionize the early detection of NDs, making diagnosis more accessible, cost-effective, and globally applicable.

Alzheimer’s and Parkinson’s Disease

AI and ML tools are increasingly transforming AD diagnosis, progression monitoring, and treatment. Techniques like Convolutional Neural Networks (CNNs) are used to analyze brain imaging (MRI, PET) for early detection [60], while Support Vector Machines (SVMs) predict cognitive decline [61] and the transition fro mild cognitive impairment (MCI) to Alzheimer's. Random Forests (RF) analyze genomic and biomarker data to assess disease risk [62], and Recurrent Neural Networks (RNNs) track disease progression over time [63]. Additionally, deep learning models are applied to fluid biomarkers and speech analysis for non-invasive diagnostics, while Natural Language Processing (NLP) helps detect early cognitive changes. AI also aids drug discovery by analyzing clinical and molecular data for potential treatment targets. Together, these AI-driven approaches are enabling earlier diagnosis, more precise monitoring, and personalized interventions, ultimately improving Alzheimer's care

Parkinson’s disease

AI and ML tools are revolutionizing the diagnosis, monitoring, and treatment of PD. Techniques such as CNNs are applied to brain imaging (MRI, PET) to detect early signs of PD, often before symptoms become apparent [64]. Support Vector Machines (SVMs) are used to analyze motor function data, such as tremor and gait patterns, to distinguish between Parkinson's and other neurodegenerative disorders [65]. Machine learning models also help predict the progression of the disease by analyzing longitudinal data from clinical assessments and wearable sensors. Deep learning is utilized for speech and handwriting analysis, which are early indicators of motor dysfunction in Parkinson’s patients. Natural Language Processing (NLP) is being applied to assess changes in speech patterns, while AI-powered tools like wearable sensors and smartphones are enabling continuous monitoring of symptoms [66]. In drug discovery, machine learning models are used to identify potential therapeutic targets and predict treatment responses. Collectively, these AI-driven technologies are advancing early diagnosis, improving symptom tracking, and

facilitating personalized treatments for Parkinson’s disease, offering more effective care and better patient outcomes.

Diabetes Management

Predictive Models for Diabetes Onset and Progression

ML models are increasingly being applied to predict the onset and progression of type 2 diabetes (T2D), offering more accurate and data-driven alternatives to traditional prediction methods [67]. A recent systematic review of studies from 2018 to 2022 revealed that Random Forest (RF) models were most used and provided superior performance in predicting T2D progression [68]. By analyzing patient data from electronic medical records, these models can identify high-risk individuals and enable targeted interventions, improving disease management and resource allocation. Despite the promise of ML, challenges persist, including the need for more effective feature reduction techniques and the development of more interpretable models. Future research should explore integrating novel biomarkers, such as cardiac biomarkers, to enhance risk stratification and improve outcomes for patients with T2D.

AI in Blood Glucose Monitoring and Control

AI is transforming blood glucose monitoring (BGM) and diabetes management by enabling non-invasive, real-time monitoring and personalized, closed-loop insulin delivery systems [69]. Machine learning algorithms, like random forest and neural networks, have shown promise in predicting glucose levels with varying accuracy [70], while AI-driven continuous glucose monitoring (CGM) systems can dynamically adjust insulin delivery, improving patient outcomes and minimizing complications such as hypoglycaemia [71]. Advanced models like ANFIS predict glucose dynamics, offering early warnings of hypoglycemic events [72]. However, challenges like sensor calibration, cost, and lag between blood glucose and interstitial fluid readings remain. Despite these issues, AI's integration into CGM technology paves the way for personalized, more effective diabetes care, promising better management and outcomes for patients.

Personalized Treatment Plans and AI

AI is revolutionizing diabetes management by enabling personalized treatment plans tailored to individual patient needs [73]. By leveraging advanced diagnostic tools, predictive modeling, and continuous data analysis, AI helps customize therapies that optimize blood glucose control, lifestyle adjustments, and dietary management [74]. This patient-centric approach enhances clinical decision-making and patient engagement, offering a more holistic solution to diabetes care. The integration of AI fosters a shift towards data-driven, adaptive therapies, improving outcomes and quality of life for diabetics. However, its successful implementation relies on robust research, secure data practices, and interdisciplinary collaboration to ensure ethical, responsible use and maximize its potential in transforming diabetes treatment.

Infectious Diseases

AI in the Diagnosis and Epidemiology of Infectious Diseases

AI is transforming the diagnosis and epidemiology of infectious diseases by leveraging vast datasets to enhance early detection, improve diagnostic accuracy, and predict disease trends [75]. AI, particularly through ML and Bayesian networks (BN), can analyze complex health data, identify weak signals, and forecast epidemics, enabling more efficient resource allocation and timely intervention [76]. This is especially valuable in resource-limited settings, where AI tools can compensate for gaps in traditional healthcare infrastructure. AI can also support personalized medicine and optimize treatment strategies by assessing factors like pathogen transmission, host susceptibility, and environmental conditions [77]. However, for AI’s full integration into healthcare, global collaboration is needed to develop standardized guidelines and regulatory frameworks that ensure equitable, ethical, and effective use. Harmonizing AI approaches across institutions and ensuring the quality of input data, especially from IoT devices, will be critical for its success. In the context of pandemics, AI-powered bio-surveillance systems could offer significant advancements in monitoring and controlling infectious diseases, requiring a concerted effort in data management and policy development to fully realize its potential.

Machine Learning in Vaccine Development and Response Prediction

By analyzing large datasets, ML algorithms can identify patterns and predict immune system responses to various pathogens, significantly speeding up the vaccine discovery process [78]. In vaccine development, ML is used to predict optimal antigens, design effective vaccine candidates, and model the interaction between vaccines and the immune system [22]. Additionally, ML models can forecast individual or population-level vaccine responses, enabling personalized vaccine strategies and more effective public health responses. As the need for rapid vaccine development grows due to emerging infectious diseases and pandemics, ML provides a powerful tool to accelerate the process and improve vaccine efficacy.

COVID-19 and AI

During the COVID-19 pandemic, AI and ML tools played a crucial role in various aspects of public health and safety [79]. These technologies were employed for predictive modeling to forecast infection rates, optimize resource allocation in healthcare facilities, and identify potential outbreaks through data analysis. AI-driven algorithms facilitated the rapid development of vaccines by analyzing genetic sequences of the virus, while ML tools helped in diagnosing COVID-19 through medical imaging and symptom assessment [80]. Additionally, chatbots and virtual assistants powered by AI provided timely information and support to the public, enhancing communication and reducing misinformation. Overall, AI and ML tools significantly contributed to managing and mitigating the impact of the pandemic.

Dermatological Conditions

Image Recognition for Skin Lesion Analysis

Image recognition technology, powered by AI, has become a key tool in skin lesion analysis, particularly in the early detection of skin cancers like melanoma [81]. By training machine learning algorithms on vast datasets of labeled images, AI systems can accurately identify and classify skin lesions based on visual patterns, often detecting irregularities that may be missed by the human eye [81]. These AI-driven tools provide dermatologists with valuable diagnostic support, enabling faster and more accurate assessments. Image recognition in skin lesion analysis is not only improving the precision of diagnoses but also enhancing early intervention, which is crucial for better treatment outcomes in conditions like skin cancer [82].

AI in the Treatment of Chronic Skin Conditions

AI is increasingly being used in the treatment of chronic skin conditions by enhancing diagnosis, treatment planning, and monitoring. Machine learning algorithms can analyze images of the skin to identify conditions like eczema, psoriasis, acne, and melanoma with high accuracy, often surpassing human dermatologists in certain cases [83]. AI-driven tools help personalize treatment plans by analyzing patient data, including genetics, lifestyle, and previous responses to treatments, ensuring more effective and tailored interventions [84]. Additionally, AI is being used to monitor disease progression through wearable devices that track skin health in real-time, providing valuable data for ongoing treatment adjustments. Overall, AI is improving both the speed and precision of managing chronic skin conditions, leading to better patient outcomes

Melanoma Detection

AI and ML tools have become invaluable in the detection and management of melanoma, offering advanced methods to improve diagnostic accuracy, predict progression, and personalize treatment [85]. For example, deep learning algorithms, particularly CNNs, are commonly used to analyze skin lesions from dermoscopic images, allowing for early detection of melanoma with a level of accuracy comparable to dermatologists. Tools like Skin Cancer AI or MoleScope use AI to classify skin lesions and assess the risk of malignancy based on image features [86]. Additionally, ML models have been developed to integrate clinical data, such as Breslow thickness and serum biomarkers, with imaging data to predict melanoma progression and the potential for metastasis [87]. These tools can also analyze genetic markers or protein expression profiles, identifying high-risk patients who may benefit from closer monitoring or early intervention. By combining multiple data sources, AI and ML offer a more comprehensive, accurate, and efficient approach to melanoma diagnosis and treatment planning.

Ophthalmology

Automated Analysis of Retinal Images

Automated analysis of retinal images using AI and ML has revolutionized the field of ophthalmology, offering efficient and precise methods for diagnosing and monitoring retinal diseases [88]. AI algorithms, particularly deep learning models, can analyze retinal scans to detect a variety of conditions, including diabetic retinopathy, age-related macular degeneration, and glaucoma, with remarkable accuracy [89]. These tools are designed to automatically segment and classify retinal structures, identify abnormalities such as hemorrhages or exudates, and assess the severity of disease progression [90]. By reducing the need for manual interpretation, automated retinal image analysis not only speeds up the diagnostic process but also improves access to healthcare, especially in regions with limited access to ophthalmologists. Furthermore, it allows for early detection and personalized treatment plans, ultimately enhancing patient outcomes.

AI in the Screening for Diabetic Retinopathy

The integration of artificial intelligence (AI) into diabetic retinopathy (DR) screening is progressing rapidly, with machine learning (ML) systems already validated for detecting diabetic retinopathy-related lesions [91]. Traditional ML models, which classify DR based on features like lesion shape, color, and location, show high sensitivity (87–95%) but lower specificity (50–69%), leading to false positives and limiting cost-effectiveness [92]. Deep learning (DL), particularly through convolutional neural networks (CNNs), represents the next generation of AI for DR screening. DL models require less human guidance and can learn directly from ground truth-labeled data, improving classification accuracy. However, challenges remain, such as the need for standardized, high-quality datasets and the development of a unified regulatory and evaluation system for AI products in clinical practice. Despite these hurdles, AI systems like IDx-DR have already gained FDA approval for autonomous DR diagnosis [93], and numerous AI tools are emerging globally, especially in China, where the healthcare system and large population provide a strong foundation for AI development. The future of DR screening will likely involve AI-driven telemedicine, real-time assessments, and improved algorithms, expanding the potential for AI to not only prevent sight-threatening diseases but also contribute to broader systemic diagnoses in healthcare

Predictive Models for Glaucoma

Predictive models for glaucoma development and progression are designed to assess the risk of glaucoma in individuals by incorporating multiple risk factors into a cohesive, data-driven evaluation [94]. These models, often using statistical methods or risk calculators, help clinicians make more objective decisions about patient care. For example, a risk calculator developed in 2005, based on the Ocular Hypertension Treatment Study (OHTS), was used to predict the likelihood of ocular hypertensive patients developing glaucoma [95]. Similarly, predictive models for glaucoma progression estimate the risk of existing glaucoma patients experiencing further damage over time. By analyzing longitudinal data and identifying key risk factors, these models improve risk assessment and guide treatment strategies, ultimately enhancing the management of glaucoma and helping prevent vision loss.

Ethical, Legal, and Social Implications

Data Privacy and Security Concerns

The growing integration of AI presents both opportunities and challenges in the realm of privacy and data security. As businesses leverage AI to analyze vast data, safeguarding sensitive information becomes increasingly critical [96]. Innovations like differential privacy and federated learning offer promising solutions to protect data while still enabling AI advancements. At the same time, evolving regulations such as GDPR and CCPA emphasize the need for transparency, accountability, and compliance in data practices. While AI systems do not directly infringe on privacy by forming perceptions about individuals, the vast amounts of personal data they process create risks to privacy and security if accessed or misused [97]. The true danger lies in the potential for security breaches, where the misuse of personal data can harm individuals, even if the AI itself does not "understand" the data. Achieving a balance between innovation and privacy requires organizations to adopt ethical data practices and privacy-preserving technologies, which not only comply with regulations but also build consumer trust. The future of privacy in AI will depend on collaboration between businesses, regulators, and consumers, ensuring that AI technologies contribute positively to society while respecting individual privacy right

Bias and Fairness in Machine Learning Models

Bias and fairness in ML models are critical concerns, as algorithms can unintentionally perpetuate or amplify existing societal biases [98]. These biases often arise from skewed or unrepresentative training data, leading to unfair predictions or decisions that disproportionately affect certain groups based on attributes like race, gender, or socioeconomic status [99]. Addressing these issues involves identifying and mitigating bias at various stages of the ML lifecycle, including data collection, model training, and deployment. Techniques like reweighting training data, adjusting algorithms, and implementing fairness constraints are used to promote more equitable outcomes. Ensuring fairness in ML models not only improves their ethical standing but also enhances their trustworthiness and effectiveness in diverse real-world applications [100]. As ML systems become more pervasive, prioritizing fairness is essential to avoid reinforcing harmful stereotypes and ensure that the benefits of AI are distributed justly across all demographic groups.

Conclusion

AI and ML hold immense potential to revolutionize various sectors, especially healthcare, by improving accuracy, efficiency, and personalization of services. These technologies can enhance diagnostics, streamline operations, and drive innovations in treatments and patient care. However, there are also notable limitations, AI and ML systems depend heavily on the quality and quantity of data they are trained on, and biases in this data can lead to inaccurate or unfair outcomes. Additionally, the lack of transparency in some AI models, often referred to as the "black-box" issue, raises concerns about trust and accountability. Regulatory frameworks and ethical considerations are still evolving, and a balance must be struck between innovation and privacy protection. Ultimately, while AI and ML offer powerful tools for advancing technology, their success depends on addressing these limitations through rigorous validation, ethical guidelines, and continued human oversight.

Abbreviations

AI: Artificial intelligence

ANN: Artificial Neural Networks

CNNs: Convolutional neural networks

DL: Deep learning

RFNN: Random Forest Neural Networks

Supporting information: None

Ethical Considerations: Not applicable

Acknowledgments: None

Funding: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author contribution statement: All authors (AG, SR, RS, MJ, RSV, AD, SM, DSL, PK, SY, SD, AAT, SM, SBM, CG, SB, KCA) contributed equally and attest they meet the ICMJE criteria for authorship and gave final approval for submission.

Data availability statement: Data included in article/supp. material/referenced in article.

Additional information: No additional information is available for this paper.

Declaration of competing interest: The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Clinical Trial: Not applicable

Consent for publication: Note applicable

References