Bioscience Methods 2024, Vol.15, No.3, 124-138 http://bioscipublisher.com/index.php/bm 133 conditions like pulmonary hypertension (PH) and heart failure (HF). One study reported an area under the receiver operating characteristic curve (AUC) of 0.859 and 0.902 for internal and external validation, respectively, in predicting PH using a 12-lead ECG (Kwon et al., 2020). Another study found that AI models applied to cardiac MRI achieved an AUC of 0.90 with 89% sensitivity and 81% specificity for diagnosing acquired pulmonary arterial hypertension (PAH) (Hardacre et al., 2021). In contrast, traditional methods such as transthoracic echocardiography (TTE) have lower sensitivity and specificity. A meta-analysis of TTE for diagnosing PH reported a pooled sensitivity of 85% and specificity of 74%, with an AUC of 0.88 (Hardacre et al., 2019). Similarly, right heart catheterization (RHC), although considered the gold standard, has its limitations in terms of procedural risks and lower diagnostic odds ratios compared to AI-based methods (Ullah et al., 2020). 7.2 Speed and efficiency AI-based diagnostic systems significantly enhance the speed and efficiency of diagnosing hypertensive heart disease. Traditional diagnostic methods like TTE and RHC are time-consuming and require specialized equipment and trained personnel. In contrast, AI algorithms can process large datasets rapidly and provide real-time diagnostic support. For example, an AI algorithm developed for predicting PH using ECG data was able to identify high-risk patients efficiently, significantly reducing the time required for diagnosis (Kwon and Kim, 2019). Moreover, AI systems can continuously monitor patients using wearable technologies, providing ongoing assessment and early detection of hypertensive conditions. This continuous monitoring is particularly beneficial for managing chronic conditions like hypertension, where timely intervention can prevent disease progression (Visco et al., 2023). 7.3 Cost-effectiveness and accessibility AI-based diagnostic systems offer significant advantages in terms of cost-effectiveness and accessibility. Traditional diagnostic methods often involve high costs due to the need for specialized equipment and trained personnel. For instance, RHC and cardiac MRI are expensive and not readily available in all healthcare settings, particularly in low-resource environments (Ullah et al., 2020). In contrast, AI algorithms can be integrated into widely available and inexpensive diagnostic tools like ECG machines. A study demonstrated that an AI-based system using ECG data could effectively detect hypertension with an accuracy of 84.2%, making it a cost-effective alternative to more expensive diagnostic methods (Angelaki et al., 2022). Additionally, AI systems can be deployed in remote and underserved areas, improving access to diagnostic services for populations that might otherwise be excluded from advanced healthcare. Furthermore, AI-based systems can reduce the overall healthcare costs by enabling early detection and intervention, thereby preventing the progression of hypertensive heart disease and reducing the need for more expensive treatments and hospitalizations (Visco et al., 2023). In summary, AI-based diagnostic systems for hypertensive heart disease offer significant improvements over traditional methods in terms of accuracy, speed, and cost-effectiveness. These systems provide highly accurate and sensitive diagnostic capabilities, enhance the efficiency of the diagnostic process, and offer a more accessible and cost-effective solution for managing hypertensive conditions. As AI technology continues to advance, its integration into clinical practice is likely to revolutionize the diagnosis and management of hypertensive heart disease, ultimately improving patient outcomes and reducing healthcare costs. 8 Challenges and Limitations of AI-Based Diagnostics for HHD 8.1 Data privacy and security concerns One of the foremost challenges in the deployment of AI-based diagnostic systems is ensuring the privacy and security of patient data. AI models require vast amounts of data to train and validate their algorithms, often necessitating the collection and storage of sensitive health information. This raises significant concerns about data
RkJQdWJsaXNoZXIy MjQ4ODY0NQ==