Artificial Intelligence in Medicine*

Table of Contents

• Introduction

• The Motivation behind Artificial Intelligence in Healthcare

• Different Fields of Artificial Intelligence in Healthcare: general machine learning, neural networks, natural language processing

• How is Artificial Intelligence Currently Used in Healthcare?

• Ethical Considerations with Artificial Intelligence in Healthcare

• Conclusion

Introduction

Differing from the relatively backward ideals of the past, modern healthcare has developed into an incredibly complex field; with advancements in such intricate occupations, the complexity of the tools used must also rise. Davenport and Kalakota (2019) reinforce this idea, mentioning, "The complexity and rise of data in healthcare means that artificial intelligence will increasingly be applied within the field." Several types of AI are already being employed by payers and providers of care, and life sciences companies." In essence, with the continuous rise and expansion of artificial intelligence, we can expect it to have a significant impact on the healthcare industry, because "according to Statista, the artificial intelligence healthcare market, valued at $11 billion in 2021, is projected to be worth $187 billion in 2030. That massive increase means we will likely see considerable changes in how medical providers, hospitals, pharmaceutical and biotechnology companies, and others in the healthcare industry operate." (IBM Education, 2023)

The Motivation behind Artificial Intelligence in Healthcare

Artificial intelligence presents numerous benefits for healthcare, as discussed extensively in Medical Literature. Through analyzing large volumes of healthcare data, AI can utilize sophisticated algorithms to uncover patterns and derive insights to inform clinical practice. Additionally, AI systems possess learning and self-correcting capabilities, allowing them to improve accuracy based on feedback. By accessing the latest medical information from diverse sources, including journals, textbooks, and clinical practices, AI can assist physicians in delivering optimal patient care. (Hong et al. (2022)) AI holds promise in reducing the diagnostic and therapeutic errors innate to human clinical practice. The ability to synthesize useful information from extensive patient populations empowers AI systems to generate real-time health risk alerts and outcome predictions. In summary, AI offers healthcare providers an invaluable tool for enhancing clinical decision-making and patient care through its skills in processing healthcare data, continuously learning, minimizing human errors, and identifying trends across large groups of patients. (Jiang F, Jiang Y, Zhi H, et al, 2017).

As the authors of the previous paragraph indicated, the uses of AI in healthcare will inarguably bring forth a new age in healthcare, as Barth (2023) stated: "The emergence of artificial intelligence in healthcare has been groundbreaking, reshaping the way we diagnose, treat and monitor patients. This technology is drastically improving healthcare research and outcomes by producing more accurate diagnoses and enabling more personalized treatments."

Different Fields of Artificial Intelligence in Healthcare

As Davenport and Kakalota (2019) stated: "Artificial intelligence is not one technology, but rather a collection of them. Most of these technologies have immediate relevance to the healthcare field, but the specific processes and tasks they support vary widely. Some particular AI technologies of high importance to healthcare are defined and described below." The following are some of the few most influential AI fields that will have applications in healthcare:

-General Machine Learning:

Machine Learning is a statistical technique for fitting models to data to 'learn' by training models with data. Machine learning is one of the most common forms of AI; it is a broad technique with many versions at the core of many approaches to AI. In healthcare, the most common application of traditional machine learning is precision medicine – predicting what treatment protocols will likely succeed on a patient based on various patient attributes and the treatment context. “Most machine learning and precision medicine applications require a training dataset for which the outcome variable is known; this is called supervised learning.” (Davenport and Kalakota, 2019).

-Neural Networks:

While being considered a subdivision of machine learning, I believe NNs (neural networks) deserve to be mentioned separately as to how large of an impact they might have on diagnosis and the reduction of mistreatments. According to Väänänen et al. (2021), neural networks (NN) consist of digitized inputs, such as an image or speech, which proceed through several hidden layers of connected artificial neurons, each layer responding to different features progressively detecting features and providing an output. Deep NNs (DNN) require special mention as they are a subcategory of NN with more complex variations used in various AI solutions in field medicine. “A typical use case for DNN is when there is a need to interpret data to certain patterns from different types of clinical images, such as pathology, skin lesions, retinal, and endoscopy images, and to find out patterns from datasets, such as medical scans, electrocardiograms, and vital signs.” (Väänänen, et al (2021). From this, we can establish comfortably that neural networks are a promising field for healthcare and medical development, especially in how proficient they might become in deducing illnesses based on images alone.

-Natural Language Processing (NLP):

NLP has primarily been used to create speech recognition programs, translations, text analysis, and many other goals related to having a machine make sense and interpret the human language. "In healthcare, the dominant applications of NLP involve creation, understanding, and classification of clinical documentation and published research. NLP systems can analyze unstructured clinical notes on patients, prepare reports (example: on radiology examinations), transcribe patient interactions, and conduct conversational AI." (Davenport and Kalakota, 2019)

How is Artificial Intelligence Currently Used in Healthcare?

Modern healthcare is being transformed and augmented by artificial intelligence-based technologies that have the capability to comprehend vast amounts of complex data, learn from examples and experience, and act autonomously, whether assisting physicians in detecting subtle patterns and anomalies across genetic profiles that humans may overlook, or powering sophisticated robotics to execute surgeries with unparalleled precision and care.

The three primary applications currently revolutionizing healthcare through artificial intelligence are AI-driven drug discovery, clinical trials support, and enhanced patient care. Pharmaceutical researchers have seen considerable success applying AI to tasks like automating the identification of novel drug targets and accelerating the drug design process by evaluating millions of potential candidates in silico. Machine learning algorithms also help streamline drug development by automating time-consuming monitoring and data analysis during pre-clinical and clinical testing. (Väänänen, et al (2021))

When applied to clinical trials, AI has demonstrated an unmatched capacity for synthesizing enormous datasets involving countless individual patient profiles, genetic factors, treatment responses, and outcomes. Powerful algorithms can efficiently generate trial results with extreme accuracy and reliably identify subtle effects that may otherwise go undetected. Additionally, healthcare AI companies are developing sophisticated platforms that utilize patient medical data and individual factors to not only improve diagnosis and treatment recommendations but also gain insights to help individuals optimize their well-being and quality of life on a personal level. (Shaheen, 2021)

Ethical Considerations with Artificial Intelligence in Healthcare

The use of AI in healthcare has fundamentally altered the medical industry. However, there are numerous legal and ethical problems that this scientific field must deal with. Despite the enormous progress of AI in society and its contribution to bettering the therapeutic process, it is not available to everyone. Many low-income and developing nations continue to lack access to cutting-edge technology. We notably face several issues when employing AI, including ethical dilemmas, privacy and data protection, informed consent, social gaps, medical consultation, empathy, and sympathy. Therefore, professionals and practitioners should consider the four medical ethics principles, including autonomy, beneficence, nonmaleficence, and justice in all facets of healthcare, before integrating artificial intelligence with the healthcare system. (Thompson, 2009).

Conclusion

In summary, it is evident that artificial intelligence holds both promise and peril for the future of medical care. By augmenting human expertise through more precise diagnoses, tailored therapies, and constant attention to individual needs, AI can help deliver revolutionary gains in patient outcomes. At the same time, such transformative technologies also demand thoughtful governance to safeguard medical ethics and ensure all may equally benefit. If stakeholders prioritize responsible development through an open-minded yet discerning review of AI systems, addressing concerns of bias, oversight, and access proactively, we can strive to realize health technology's benefits for all. With diligence to uphold principles of compassion and well-being above algorithms and humans guiding progress prudently, artificial intelligence may serve healthcare as a faithful tool to advance its highest purpose - preserving and improving human lives. In this spirit of prudent yet progressive change, AI shows the potential to elevate medicine to new heights of personalization, participation, and care.

References

Jiang, F. et al. (2017) Artificial Intelligence in Healthcare: Past, present and future, Stroke and Vascular Neurology. Available at: https://svn.bmj.com/content/2/4/230 (Accessed: 05 September 2023).

Väänänen, A. et al. (2021) Ai in healthcare: A narrative review, F1000Research. Available at: https://f1000research.com/articles/10-6 (Accessed: 06 September 2023).

Hong, G., Smith, M. and Lin, S. (2022) The AI will see you now: Feasibility and acceptability of a conversational AI medical interviewing system, JMIR formative research. Available at: https://pubmed.ncbi.nlm.nih.gov/35759326/ (Accessed: 06 September 2023).

Shaheen, M. (2021) Applications of artificial intelligence (AI) in Healthcare: A Review, ScienceOpen. Available at: https://www.scienceopen.com/hosted-document?doi=10.14293%2FS2199-1006.1.SOR-.PPVRY8K.v1 (Accessed: 09 September 2023).

Barth, S. (2023) Artificial Intelligence (AI) in Healthcare & Hospitals, ForeSee Medical. Available at: https://www.foreseemed.com/artificial-intelligence-in-healthcare (Accessed: 09 September 2023).

Thompson, D.F. (2009) The challenge of conflict of interest in medicine, Zeitschrift fur Evidenz, Fortbildung und Qualitat im Gesundheitswesen. Available at: https://pubmed.ncbi.nlm.nih.gov/19554887/ (Accessed: 09 September 2023).