AI in the Healthcare Industry by Susanna Lamadieu

Personalized Overview

susannalamadieu — 2025-03-16 19:45:09 UTC

Healthcare is an industry focused on maintaining or improving the overall physical and mental health of the public. Its methods of doing so has evolved over centuries. The industry has achieved numerous achievements from their endeavors such as vaccination, prevention, and improved sanitation. All these changes were accompanied by the improvement of technology. Pharmaceuticals, hospitals, and clinics have adapted to pathogens that have been introduced to humans and will continue to do so. As time goes on, the healthcare industry will, no doubt, make advances to better serve humanity including finding methods to make various areas of healthcare such as research, testing, and diagnoses easier and more efficient. Within the past decade, there has been one uprising, novel aspect of technology that can be observed today and is now being used for good in healthcare today. Artificial Technology (AI) experienced a surge of recognition worldwide. With many companies taking up AI and creating new models, it has revolutionized industries including retail and manufacturing via algorithms to make the user experience more seamless and efficient. In healthcare, AI has been used to aid patients and further research. It can glean findings and existing research to make a larger sample size using similar criterion, it can monitor patients from afar using algorithms tuned to each patient’s individual needs, and it can be used in imaging to recognize anomalies not easily visible to the human eye. Healthcare will evolve just as AI will, and both are potentially beneficial assets of humankind if used as such.

I’m interested in working in the healthcare industry, specifically in metal health. I’ve always loved studying behavior and the brain, and I think that AI will impact this aspect of healthcare just as much as any other. I plan on covering how AI is used in remote patient monitoring, medical diagnoses, and imaging as well as studies that lie within those areas. Patients don’t always live on the premises of hospitals or clinics. It wouldn’t be right to bar them of their daily lives to monitor their health every hour of the day if it isn’t needed. In the future, health professionals may have to use AI while they monitor, diagnose, or scan a patient.

AI and Remote Patient Monitoring

susannalamadieu — 2025-03-16 19:50:17 UTC

In remote patient monitoring (RPM), AI is used to monitor a patient’s vitals such as heart rate, blood pressure, and blood oxygen levels at home. It is said to identify “early detection of complications” without having to worry about being on the premises of a healthcare provider. Someone with diabetes may use an “AI-powered remote patient monitoring system” that tracks their dietary habits and exercise. (Tenovi, 2024). The AI algorithm will analyze patterns, presumably through a device, and suggest meals and physical activities based on the patient’s results. Some models may use chatbots and virtual assistants to aid patients as well. However, the use of AI RPM may be limited compared to imaging. There have been several studies taken, but according to a study published in the NIH, among 64 AI/ML (Artificial Intelligence/Machine Learning) algorithms used for RPM, “only 29% mentioned any algorithm.” (Dubey, 2023). It is said that the “absence of an AI algorithm was the most common reason for rejection.” (Dubey, 2023). Healthcare companies must be as transparent as possible especially where AI is used. If an AI model is being used, the company’s policy must explicitly state so.

I feel positive about AI patient monitoring if the patient themselves are aware of data privacy and their uses. It is a relatively easy way for both healthcare providers and patients to monitor their health without needing to commute to the clinic to check in. However, I believe that remote patient monitoring shouldn’t be AI independent; there should always be a human operator behind AI monitoring systems to mitigate bias or address a possible error in the AI’s algorithm or output.

Picture

AI and Medical Diagnosis

susannalamadieu — 2025-03-16 20:08:52 UTC

AI can be used to diagnose and protect patients through various methods such as medical imaging, virtual assistants, and disease prevention. AI models in these areas are trained to recognize “patterns in radiology images” such as signs of certain cancers or tumors that can be missed by human radiologists. They can also be trained to glean information and identify patterns from electronic health records (EHRs) and compare them to patients to flag health risks such as “diabetes, heart disease, or stroke.” (Park, 2024). AI can aid healthcare professionals in identifying rare diseases using data sets to find similar cases. (Ellis, 2024). There is no one set way to make diagnosis easier, but rather, it is the culmination of data from different methods that can be used to diagnose patients. The larger the sample size, the more accurate data will be. However, data privacy may pose a large issue in AI patient diagnosis. For data to be taken, there must be a database where thousands of patients’ information resides. Healthcare providers must adhere to regulations such as the Health Insurance Portability and Accountability (HIPAA) to protect patient data. (Park, 2024). Data must be diverse to be effective and beneficial to all; algorithms may have a bias if there isn’t enough diversity within data sets (Park, 2024). If any misuse occurs, one must take accountability for it. An AI is a machine and cannot do so, but the human operators who provide input should help accountable for any unintentional or malicious use of AI.

I think that AI can be used for good in diagnosis, especially in the case of rare diseases. Many diseases, ailments, or conditions may appear to be similar to each other on the surface and may require further investigation. AI can be used to spot the minute differences between each of them using datasets of similar cases in places where healthcare professionals may not be able to easily distinguish between them. It can help the healthcare industry in that way.

Picture

AI and Imaging

susannalamadieu — 2025-03-16 20:10:44 UTC

Medical imaging is used in a variety of different areas of healthcare such as radiology and cardiology. It uses “deep learning algorithms, convolutional neural networks, and generative adversarial networks” to increase the efficiency of imaging analyses. (Pinto-Coelho, 2023). This will allow healthcare professionals to accelerate their process and allow for more accurate results. This can allow for early detection which can be followed by treatment options for patients. Generative adversarial networks (GANs) and variational autoencoders (VAEs) are especially used in image-to-image translation and generation. (Pinto-Coelho, 2023). According to a study of AI imaging, there may be “limited work examining the interventional impact that different types of AI educational programs… can have on improving acceptability.” (Hua, 2024). Low digital literacy may be associated with negative attitudes towards novel innovations in healthcare.

I think it’s an interesting approach to diagnosis in medical imaging. I’m typically suspicious of photos and their relevance to AI since many times, AI may either misinterpret or generate photos that don’t “make sense.” While my suspicions come from places outside of healthcare, I can’t help but express worry about how AI-generated photos may lead to misinformation even said photos are taken from a large database. I think human monitors is vital here; with human intervention, there should be less chances of inviable output from AI algorithms.

Picture

Future Trends and Ethical Considerations

susannalamadieu — 2025-03-16 20:11:17 UTC

AI will improve significantly as time goes on in industries of all kinds. As more data sets are logged, research findings will become more accurate, diagnoses will become more accurate, and patients will have better access to the healthcare they need. I think that, as people acclimate to AI in everyday life, it will become more widely accepted. However, because of how novel the recent developments of AI are, there are some things that need to be considered when using AI. While there have been laws and policies about AI that are meant to protect the user such as the European AI Act that is meant to ensure AI’s proper use amongst AI model developers, there are still concerns of data privacy and intentional misuse. AI must have access to data across different data sets to be as accurate as possible. However, where and how the AI model extracts data must be explicitly stated as to maintain its credibility and trustworthiness. It may draw from sensitive or inappropriate places whether the AI algorithm intends to or not. This may also lead to questioning the human operator or developer behind the AI model. The person may willingly choose to take information for their own malicious intentions from other sites across the internet. It can lead to skewing the algorithm to the person’s favor instead of having an unbiased viewpoint or even being used against other people in cybersecurity. They may use AI to intrude or steal other people’s sensitive information using the same methods hackers may use. In healthcare, this may look like stealing someone’s social security number or health information off electronic health records (EHRs) for fraudulent activities under the victim’s name. There must be extra measures taken to prevent malicious AI use such as having ethical hackers test security or cybersecurity that flags potential “AI-operated hackers.”

At this point in time, most people can have access to AI tools to use to our own benefit such as ChatGPT or Microsoft Copilot. However, there are certain areas where AI should not be taken lightly. Data that is logged into EHRs is sensitive by nature. While data may be taken from EHRs to aid patients for good by looking at similar data to them and recommending healthier choices, it can be stolen by others. AI must have a human operator to ensure that AI is credible and trustworthy for all.

Societal Impact

susannalamadieu — 2025-03-16 20:11:49 UTC

AI has been incorporated in almost every facet of everyday life. It can be seen from the television news headlines to simple searches on Google. AI can make customer service and feedback much quicker than waiting for a person to answer or allow a person to quickly and conveniently find answers to their questions. If AI is in the hands of trusted healthcare professionals and operators, AI in healthcare will be a net positive for everyone. If implemented in other industries like robotic engineering, it could aid patients in clinical settings as well. There could be automated machines acting as a nurse or companion. While it may certainly seem helpful to clients and lower costs for health companies, it may lead to an employment shortage. There should certainly be people who manage AI models and the algorithms it produces and there will certainly be more jobs pertaining to AI moderation, but, if AI can be managed by few people and complete the jobs of many, there’d be less of a reason to employ those who are qualified for those positions. There have been stories in other industries that many can find online detailing how they had been laid off presumably because AI could do their job with greater accuracy than a human.

Those who are familiar with AI seem to have a better understanding of what AI is capable of. I’ve heard those of the older generations complain or express their disdain for AI. Many believe that it has made people even lazier than before AI’s seemingly sudden appearance within the last decade; anyone could search for anything with the touch of a button, but nobody is obligated to search for their answers anymore. Many believe that AI has made things earning money through traditional means like having a 9-5 job much harder since a machine can do their job for them. One must understand both the benefits and risks while using AI as a tool. This is no different in the healthcare industry.

Video Overview

susannalamadieu — 2025-03-16 20:57:02 UTC

Reflection

susannalamadieu — 2025-03-16 20:57:52 UTC

As I thought about ways to research, I quickly realized that the use of AI in Healthcare was extremely vast. I had to search for “AI in Healthcare” to find broad examples of different areas where AI may be used. I searched for specific areas in healthcare where AI could be used. I only had vague ideas of what AI could be used for like imaging or virtual nursing, but this opened my eyes. I think that AI can be incorporated in every aspect of healthcare including meticulous surgeries or detailed conducted research. Since I want to be in the healthcare industry, this research has left me intrigued by how AI will affect the specific domain I’d like to be a part in. I think that mental health is becoming more mainstream, and with AI becoming more prominent, I believe AI and mental health will coincide at some point in the future.

References

susannalamadieu — 2025-03-16 20:58:52 UTC

Tenovi. (2024, August 23). How is AI Used in Remote Patient Monitoring? Tenovi. https://www.tenovi.com/ai-in-remote-patient-monitoring/ [Rationale: This article provides examples of how AI can be used in Remote Patient Monitoring and how it can become the standard of healthcare worldwide.]

Dubey, A., Tiwari, A. (2023, May 3). Artificial intelligence and remote patient monitoring in US healthcare market: a literature review. National Library of Medicine. https://pmc.ncbi.nlm.nih.gov/articles/PMC10158563/ [Rationale: This scholarly source provides explanations and statistics on AI and its use in Remote Patient Monitoring. This includes how much more often people will accept AI use if it is stated explicitly within a policy.]

Ellis, L. D. (2024, August 30) The Benefits of the Latest AI Technologies for Patients and Clinicians. Harvard Medical School. https://postgraduateeducation.hms.harvard.edu/trends-medicine/benefits-latest-ai-technologies-patients-clinicians [Rationale: This source comes from a highly renowned institution. It provides numerous cases on how AI can be used in multiple practices such as imaging or research.]

Park University. (2024, December 2). AI in Healthcare: Enhancing Patient Care and Diagnosis. Park University. https://www.park.edu/blog/ai-in-healthcare-enhancing-patient-care-and-diagnosis/ [Rationale: This source outlines the potential benefits and disadvantages of AI in Healthcare. It reviews ethical considerations one must consider as the healthcare industry begins to adopt AI into its practices.]

Pinto-Coelho, L. (2023, December 18). How Artificial Intelligence Is Shaping Medical Imaging Technology: A Survey of Innovations and Applications. National Library of Medicine. https://pmc.ncbi.nlm.nih.gov/articles/PMC10740686/ [Rationale: This scholarly source provides and describes imaging techniques that are used by AI to compare, identify, and flag anomalies found in medical scanning procedures.]

Hua, D., Petrina, N., Young, N., Cho, J., & Poon, S. K. (2024, January) Understanding the factors influencing acceptability of AI in medical imaging domains among healthcare professionals: A scoping review, vol. 147. Science Direct. https://www.sciencedirect.com/science/article/pii/S0933365723002129 [Rationale: This scholarly source provides an explanation as to why there may be low acceptance rates of AI among healthcare professionals, and why it may set AI’s progress, as it pertains to healthcare, back.]