Page 59 - AIH-2-3

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

of robust legal and ethical frameworks to govern AI use in the use of remote medical facilities to maintain healthcare
healthcare. delivery. 55,56 AI-powered mHealth (AIM) has emerged as
a promising subfield that integrates AI techniques with
6.6. AI in specialized medical fields: Dentistry mHealth applications to address key healthcare challenges.

AI’s role in medical diagnostics extends beyond general These AI techniques, including DL, federated learning (FL),
healthcare and into specialized fields like dentistry. and explainable AI (XAI), offer more accurate diagnostic
Orthodontics, for instance, relies on AI to diagnose insights while ensuring patient privacy and data security.
57
malocclusions and plan treatments. By analyzing Researchers have noted the increasing integration of
cephalometric radiographs, AI systems can assess AI in mHealth, especially during the pandemic, with
abnormalities in dental and craniofacial structures more advancements such as real-time disease progression
accurately than traditional methods. 22-32 AI-driven models monitoring and chronic disease management. 58-61
help that orthodontists make precise diagnoses and These technologies are instrumental in providing
improve treatment outcomes by identifying subtle patterns non-invasive care and enabling emergency responses,
in dental data. AI technologies are reshaping medical particularly for at-risk communities and individuals
diagnostics, offering innovative solutions for disease with limited access to healthcare. 62,63 AI’s integration
detection, treatment planning, and patient management. into mHealth offers significant advantages, including
As AI continues to evolve, its applications in healthcare automated chronic disease detection, suicide prediction
will likely expand, leading to more accurate, efficient, and and intervention, and reduction of medical errors. Medical
personalized medical care. However, continuous research errors remain one of the leading causes of preventable
is required to address the technological, ethical, and deaths in the U. S., and AI-enabled clinical decision-making
regulatory challenges associated with AI integration in systems, which use real-time data from wearable sensors,
medical diagnostics, ensuring that these systems benefit can substantially reduce these errors. 64,65 AIM solutions
both healthcare providers and patients.
also hold the potential to extend high-quality medical
6.7. The application of mHealth care to underserved populations, addressing healthcare
inequities and improving patient outcomes on a broader
The rapid evolution of information technology in scale. However, challenges remain in the implementation
66
healthcare has progressed beyond simple data collection to of AIM technologies. The healthcare industry has
the sophisticated use of AI for advanced diagnostics and historically been resistant to automation, largely due to
preventative medicine. Initially, healthcare information concerns surrounding data privacy, interpretability of AI
systems were designed solely for the purpose of gathering models, and regulatory constraints. 67
patient data. However, with the rise of ML techniques and
data analytics, healthcare providers are now able to leverage Nevertheless, recent advances in DL and FL have
this data to make smarter, faster, and more accurate opened new possibilities for AI in healthcare, allowing
decisions. One key area that has seen significant growth for secure data management and knowledge transfer
48
68
is the field of mobile health (mHealth), which utilizes AI across decentralized systems. FL, in particular, ensures
to enhance patient care through real-time monitoring, that patient data remains within healthcare organizations,
particularly for life-threatening conditions such as asthma, while still enabling the training of powerful AI models on
diabetes, and sleep apnea. 49,50 distributed datasets. 69,70 This allows for the development of
personalized care algorithms without compromising data
The growing presence of wearable technologies,
Internet of Things (IoT) devices, and mobile sensors has security.
fueled the expansion of mHealth. This trend is reflected The growing importance of XAI in the mHealth domain
in the increasing adoption of remote in-home care and cannot be overstated. The Defense Advanced Research
telemedicine, particularly in response to the COVID-19 Projects Agency has been instrumental in promoting
pandemic. 51,52 The healthcare information technology the development of AI models that are interpretable,
market has seen substantial growth in the use of these trustworthy, and usable by healthcare professionals. XAI
71
technologies, enabling real-time monitoring of patients plays a key role in fostering trust and acceptance of AI
outside traditional healthcare facilities. Not only does models in medical practice, as it allows clinicians to better
mHealth facilitate early detection and treatment of understand the rationale behind AI-driven decisions. This
chronic diseases but it also enhances patient safety and transparency can improve decision-making in clinical
well-being through remote monitoring solutions. 53,54 The settings, reduce medical errors, and enhance the overall
pandemic has further accelerated the adoption of mHealth efficacy of healthcare delivery. 72,73 The convergence of
technologies, as social distancing measures necessitated AI and mHealth also points toward the potential for AI

Volume 2 Issue 3 (2025) 53 doi: 10.36922/aih.5173