This post is the second part of a two-part guide to artificial intelligence (AI) for medical science liaisons (MSLs), which explains the basics of AI and provides real-world applications. Part one defined general AI information and clarified machine learning’s (ML) role in healthcare and medical affairs.

As a quick review, AI is the umbrella term for any technology with the ability to reason and adapt based on data and algorithms. Essentially, AI strives to mimic human intelligence, reduce the manual burden of tasks, and enhance the quality of life. AI has transformed medicine, enabling faster, more accurate diagnoses and reducing the burden of manual tasks for healthcare providers (HCPs). AI transforms data into action and drives decision-making in both healthcare and medical affairs. In the life sciences, HCPs use AI for patient diagnostics, health services management, predictive medicine, clinical decision-making, and surgery.3,8 In medical affairs, AI is used to gain insights into the market, enhance the capabilities of a medical affairs team, and ultimately produce more profitable and meaningful outcomes.1By 2030, the AI healthcare market will reach a value of $10.4 billion. A lack of AI experts, however, hinders the market's growth. As scientific field experts disseminating knowledge, MSLs can accelerate the adoption of AI technology. Regardless of an MSL’s specialty, understanding AI and subfields is essential. AI will shape the future of healthcare and medical affairs. Understanding AI, the subfields of AI, and their functions within medical affairs and healthcare will ensure you remain relevant in a constantly evolving industry. After reading part two of the guide, you will understand two more subfields of AI, artificial neural networks and deep learning, and their impact on healthcare and medical affairs.

What Is an Artificial Neural Network (ANN)?

An artificial neural network (ANN), also known as a neural network, is a subset of machine learning (ML). In part one of the guide, the role of ML in healthcare and medical affairs was explained. AI and its subfields build on each other to produce ever-more powerful and efficient technologies. ANNs are biologically-inspired computational networks consisting of thousands of artificial neurons, also known as 'processing units.' An ANN imitates the brain's ability to interpret knowledge by using algorithms to understand and analyze immense volumes of data at an extraordinary rate. An example of a neural network is Google's search algorithm. However, ANNs are used across many industries, including healthcare and medical affairs. Healthcare ANNs have a wide range of applications in the life sciences industry. Today, ANN is used for nearly all medical therapeutic areas but has revolutionized the fields of cancer and cardiology in particular.9 ANN applications in health care include:9

• Developing diagnoses
• Predicting cancer
• Detecting arrhythmias
• Recognizing speech
• Predicting hospital stay lengths
• Analyzing medical imaging

In cardiology, ANNs are integrated into everyday diagnostic and surgical practices. Diagnosing coronary artery disease, detecting arrhythmias, analyzing imaging, and determining appropriate drug dosages are four ways ANN is applied in treating cardiovascular disease. Figure 1: Applications of ANNs for the Treatment of Cardiovascular Disease.9

Aside from clinical applications, ANNs have administrative applications. The ability of ANN to recognize speech can reduce physicians’ manual tasks and improve workflows. HCPs spend nearly one-third of their time completing electronic healthcare record (EHR) documentation. Speech recognition technology with ANNs can now auto-fill forms for HCPs, improving the EHR documentation process. Medical Affairs ANNs' role in medical affairs is similar to that of ML. The ANNs can analyze past and current data to estimate future values, uncovering hidden correlations in the data. Many predictive analytic tools use ANNs to create data-driven insights. The insights gained from predictive analytic tools are translated into action and determine how MSLs communicate with HCPs. Interacting with HCPs is more effective when medical affairs and MSL teams can access actionable data. Medical affairs use ANN and ML in the following ways:

• Identifying and engaging health care providers
• Integrating data streams for analysis and optimization of strategic objectives
• Analyzing field activity to gain insights
• Aligning interactions to product lifecycle and preferred channels

ACMA Engage™ is an extensive, searchable database powered by AI with insights informing the medical affairs team about HCPs’ activities. In addition to improving how MSLs provide information to HCPs, the platform increases the rate at which field medical teams collect feedback and produce market insights.Experts from the life sciences and AI industries contributed to the design and development of the Engage platform. By understanding each stakeholder's needs, preferences, and interactions, ACMA Engage helps MSLs foster more profound, insightful communication.

What Is Deep Learning?

Deep learning (DL) is a subfield of ML and ANNs. DL uses three or more layers of neural networks to mimic the human brain’s ability to process information.4 DL and ANN are often used interchangeably, which is confusing. Due to this, it's important to note that the “deep” in DL refers to the depth of layers in a neural network.AI applications use deep learning technology to automate and perform physical or analytical tasks without human interaction. Deep learning technology drives the following products and services:

• digital assistants
• voice-enabled TV remotes
• credit card fraud detection
• self-driving cars

Healthcare While DL has already transformed our everyday lives, its impact on healthcare is just beginning. For instance, the first drug created entirely through AI entered Phase 1 clinical trials in December 2021. Insilico Medicine uses AI during every step of the drug discovery process and created therapy for idiopathic pulmonary fibrosis (IPF) with AI. IPF is a progressive, irreversible, and fatal decline in lung function with unknown origins. DL was first used in medical imaging, and its use has quickly spread to every facet of the life sciences industry.5 For instance, DL is used to analyze potential skin cancer lesions and measure the growth of a tumor. DL is utilized to analyze images from every therapeutic area, but it is highly used in oncology. The following are 4 of the main applications of DL: 5

• Drug discovery
• Diagnostics
• Medical Imaging
• Genomics

Figure 2: Deep Learning’s Impact on Healthcare

For genomic medicine, DL has revolutionized scientists’ ability to interpret DNA data, facilitating the discovery of new drug targets and faster diagnoses.5 Recently, DL helped to uncover the complexity of genetic expression. A study published in 2020 concluded coding and non-coding regions of DNA, as well as the entire gene regulatory structure, is responsible for a gene’s expression level.10 DL is also used to process EHRs, including both structured (e.g., diagnosis, medications, laboratory tests) and unstructured (e.g., free-text clinical notes) data. Using DL, this data can help predict patient outcomes. Studies have found that DL produces more accurate clinical predictions than ML when forecasting a patient’s disease progression.5

Medical Affairs Medical affairs and MSLs use DL, similar to how HCPs use DL to discover trends in unstructured data in EHRs. Social media data is considered unstructured but holds a wealth of information for life science organizations. By using DL, life science organizations can learn what customers say about their products and how people feel about the company. ACMA Sentiment™ is a digital AI-powered digital solution that provides MSLs with real-time customer insights. ACMA Sentiment™ offers the following insights:

• Customer feedback on products, disease state, and the company
• Share of Voice (SoV) of digital KOLs by therapeutic area
• Strategic planning and anticipating future market needs
• Providing impactful scientific engagement

ACMA Sentiment™ is a part of ACMA Insights™, a suite of digital solutions available to MSL teams. ACMA Insights™ offers five platforms, each specializing in insightful data for medical affairs and MSLs to optimize field activity. After reading the complete Guide to Artificial Intelligence for Medical Science Liaisons, you now have a basic understanding of the subfields of AI, how AI impacts healthcare, and how AI affects medical affairs. Check back in the coming weeks for more about AI in healthcare and medical affairs on the ACMA blog.

Artificial Intelligence Resources

• Learn more about ACMA Insights™.
• Read Predictive Analytics in Medical Affairs: Leveraging data in the age of Artificial Intelligence.
• Read 5 Ways Artificial Intelligence Will Change Medical Affairs

References1.Bedenkov A, Moreno C, Agustin L, et al.. Customer Centricity in Medical Affairs Needs Human-centric Artificial Intelligence. Pharmaceutical Medicine. 2021;35(1):21-29. doi:10.1007/s40290-020-00378-1.2. Bohr A, Memarzadeh K. The rise of artificial intelligence in healthcare applications. Artificial Intelligence in Healthcare. 2020;25-60. doi:10.1016/B978-0-12-818438-7.00002-2.3.Fazlollahi AM, Bakhaidar M, Alsayegh A, et al.. Effect of Artificial Intelligence Tutoring vs Expert Instruction on Learning Simulated Surgical Skills Among Medical Students. JAMA Network Open. 2022;5(2):e2149008. doi:10.1001/jamanetworkopen.2021.49008.4. Jiang F, Jiang Y, Zhi H, et al Artificial intelligence in healthcare: past, present, and future. Stroke and Vascular Neurology 2017;2:doi:10.1136/svn-2017-000101.5. Miotto R, Wang F, Wang S, Jiang X, Dudley JT. Deep learning for healthcare: review, opportunities, and challenges. Brief Bioinform. 2018;19(6):1236-1246. doi:10.1093/bib/bbx044.6. Nelson SD, Walsh CG, Olsen CA, et al. Demystifying artificial intelligence in pharmacy. Am J Health Syst Pharm. 2020;77(19):1556-1570. doi:10.1093/ajhp/zxaa218.7. Rong G, Mendez A, Bou Assi E, Zhao B, Sawan M. Artificial intelligence in healthcare: Review and prediction case studies. Engineering (Beijing). 2020;6(3):291-301. doi:10.1016/j.eng.2019.08.015.8.Secinaro S, Calandra D, Secinaro A, Muthurangu V, Biancone P. The role of artificial intelligence in healthcare: a structured literature review. BMC Medical Informatics and Decision Making. 2021;21(1). doi:10.1186/s12911-021-01488-9.9. Shahid N, Rappon T, Berta W. Applications of artificial neural networks in health care organizational decision-making: A scoping review. PLoS One. 2019;14(2):e0212356. Published 2019 Feb 19. doi:10.1371/journal.pone.0212356.10. Zrimec, J., Börlin, C.S., Buric, F. et al. Deep learning suggests that gene expression is encoded in all parts of a co-evolving interacting gene regulatory structure. Nat Commun 11, 6141 (2020). doi:10.1038/s41467-020-19921-4.