Page 53 - AIH-2-3

Artificial Intelligence in Health AI in medical diagnostics: A multi-disease approach

generate responses to medical queries that are perceived 3.1. Cardiovascular medicine
as more empathetic and of higher quality than responses AI has demonstrated significant potential in diagnosing
from healthcare professionals, though these are not in coronary artery disease and predicting outcomes such
the context of established patient-physician relationships. as patient mortality and adverse effects following acute
EHR systems, widely adopted in healthcare, have become coronary syndrome treatment. Wearable devices and
5,6
essential for storing and sharing patient data. AI enhances smartphones are expanding the ability to monitor cardiac
EHR functionality by utilizing natural language processing
(NLP) to standardize medical terminology, improve the health, potentially enabling earlier detection of events like
7,8
readability of medical notes, and predict patient risks heart attacks outside hospital settings. AI has also been
based on historical data. By identifying trends in patient applied to analyze heart sounds and diagnose valvular
9
data, AI can offer predictive insights, alerting physicians disease; however, challenges remain due to limited
to potential health risks and allowing for preemptive training data, especially regarding social determinants of
interventions. These predictive models have achieved cardiovascular health. In some areas, AI is non-inferior to
significant accuracy in assessing treatment responses, humans, such as echocardiogram interpretation, and has
further demonstrating the value of AI in patient care even outperformed physicians in diagnosing heart attacks
management. 10-12 With the volume of EHRs doubling every in emergency settings. 10
5 years, AI offers the necessary bandwidth to analyze this 3.2. Dermatology
data effectively and assist healthcare providers in making
informed clinical decisions. AI has made strides in processing medical images for
dermatological diagnoses, such as skin cancer detection.
AI has also made significant strides in addressing Studies show that ML models can achieve dermatologist-
drug-drug interactions (DDIs), a critical issue in patients level accuracy in some cases. However, many studies
32
taking multiple medications. Advanced algorithms can have not adequately engaged with external validation
scan medical literature and user-generated content, such or considered skin tone disparities, which are crucial
as EHRs and adverse event reports, to identify potential for equitable diagnosis and treatment. AI also shows
interactions between drugs. 13-15 These innovations have the potential in evaluating the outcomes of maxillofacial
potential to prevent harmful drug interactions, improving 33
patient safety. Competitions such as the DDIExtraction surgeries.
challenge have helped standardize and evaluate the 3.3. Gastroenterology
effectiveness of these AI-driven algorithms, driving further
research and development in this field. AI has improved the detection of abnormal tissues during
endoscopic procedures like colonoscopies, with the early
Telemedicine, which has surged in popularity, offers stomach cancer detection showing sensitivity close to
another area where AI is transforming healthcare. Through expert endoscopists. AI tools are being developed to
34
the use of sensors and wearable devices, AI can monitor predict ulcerative colitis flare-ups with similar accuracy
patients remotely, identifying subtle changes in health that to human pathologists, offering promising support for
may go unnoticed by human caregivers. 16-18 These devices disease management. 35
allow for constant patient monitoring, alerting physicians
to potential issues in real time. 3.4. Obstetrics and gynecology
AI-powered chatbots have also been introduced for AI is enhancing imaging techniques such as ultrasound and
mental health therapy, though some experts argue that MRI in obstetrics, assisting in diagnosing and monitoring
they cannot replace the human connection necessary for pregnancies. Its applications are expanding in areas like
effective care. 19-21 As life expectancy increases and the fetal monitoring, with AI improving diagnostic capabilities
aging population grows, AI can help caregivers monitor for various obstetrical issues. 38
elderly patients through personal and environmental
sensors, though these technologies raise privacy 3.5. Infectious diseases
concerns. 22-24 During the COVID-19 pandemic, AI contributed to
Despite these limitations, AI’s role in healthcare early detection and monitoring of virus spread. 39,40 Other
will likely continue to expand, offering solutions to applications include detecting antimicrobial resistance
complex medical challenges while improving patient and malaria and improving point-of-care diagnostics
outcomes. AI is showing increasing promise in various for diseases such as Lyme disease and sepsis. 41,42 AI has
clinical applications across a wide variety of medical also been used in analyzing blood smears and predicting
specialties. 25-75 complications in viral infections like hepatitis. 44,45

Volume 2 Issue 3 (2025) 47 doi: 10.36922/aih.5173