@AI_4_Healthcare: Learning about #aiinhealthcare applications for Patients, Clinicians, Administrators, Researchers, Media, and Humanity.

Abstract

The integration of Artificial Intelligence (AI) into healthcare represents a paradigm shift towards more personalized, efficient, and effective patient care. The @AI_4_Healthcare initiative on X (formerly Twitter), founded and solely authored by Walter Robinson, emerges as a pioneering project aimed at democratizing knowledge about AI’s role in healthcare. Through a unique blend of personal experience, professional insights, and a dedicated social media strategy, this initiative aims to foster an informed and engaged community. This interview of ML in Health Science with @AI_4_Healthcare founder Walter Robinson outlines the project’s inception, approach, outcomes, and future direction, underpinning its contribution to health equity and the promotion of patient-centric AI applications.

Opinion Piece

Introduction

YR Question 1: Please explain the aim of your project and why it was created. I personally believe that the main aim is to summarize data about AI applications in healthcare and life science settings from public content for patients, researchers, and clinicians.

WR Answer 1: The primary objective is to summarize and present links to data (articles, videos, journal abstracts, financing trends, regulatory developments, and legislative initiatives) about AI applications in healthcare and life science settings from publicly available sources for an audience of patients, researchers, clinicians, media, and the general public. This data comes from daily scans of over 500 sources and this "library” continues to grow.

YR Question 2: You mentioned in our chat that you had (or still have) medical experience through charitable activities. Could you provide more information about it here? That can help readers better understand the motivations and goals of your project.

WR Answer 2: After spending 15+ years in pharmaceutical policy (regulation, reimbursement, and stakeholder relations) from an association, consultant, and corporate perspective alongside 25+ years in volunteer healthcare governance, charitable leadership, and fundraising, I started @AI_4_Healthcare on X (formerly Twitter) to share my excitement and learnings about the many and growing uses and applications of #aiinhealthcare. The goal is to share my perspective from a biopharma industry background and infuse it with my passion for #patients as the point-of-care and locus of all #healthpolicy solutions and deeply personal experience as a child in a #clinicaltrial in 1970 and the fact that a clinical trial provided me with the greatest gift of life, a child.

Material and Methods

YR Question 1: If I understand correctly, the main method and material of your project is @AI_4_Healthcare platform (Figure1). Please provide information about how you generate your content, what criteria you use to add content for publication on your platform, and what methods you use to increase the visibility of your content. I personally believe that you are a highly communicative person who can easily communicate with people. Do you have any educational backgrounds or connections that help you build your AI database.

Figure 1: Landing Page X.

WR Answer 1: Content is sourced through subscriptions to e-letters, biopharma-medical-ai-legacy/digital-tradepress-financial trackers sites, some webscraping searching for relevant "keywords" and #hashtags, and using Feeder to track several hundred publications/information points. My experience in legislative relations, public affairs, and 30-years of communications (writing for all types of audiences – CEOs, politicians, legislative testimony, newspaper columns and OPEDs, clients’ briefs, and advocacy organizations) along with being a National Spokesperson and “political pundit" in local and national media have provided me a degree of communications expertise. My Honors Bachelor of Commerce degree and exposure to various industries over my career also gives me a grounding in the world of business and finance. Finally, I have the privilege through my work to have a “front-row” seat to interact with and influence public policy via key decision makers (elected and officials) at the national, provincial, and city levels of government.

YR Question 2: If I understood correctly, the second main component of your platform is your official LinkedIn profile (Figure 2). Here, you provide information under your name. This approach is honest, as you don't hide behind the facade of a specific company. It would be beneficial here if you could describe your philosophy behind not having an official website.

Figure 2: Landing page LinkedIn.

WR Answer 2: I used X.com as it is still the most robust social media platform for digestible bytes of #aiinhealthcare information. Cross posting to other social media platforms would be a draw on my time. I build out some of my posts and observations on LinkedIn. On a weekly basis, I ensure a reference or link to my profile, so people know who is running/managing the @AI_4_Healthcare handle and always identify myself in any direct message exchanges on X.

YR Question 3: Please mention here any official (governmental, local, or international health authorities) or regulatory norms applicable to your project topic. I personally think you should look at these documents and links: EU AI Act^{⁠1 ⁠2}, WHO AI Act^⁠3, and NCBI search on topic AI.

WR Answer 3: I have built a "thread" with the hashtags #AIGovernance - #AILegislation - #AIRegulation - #ResponsibleAI as robust and agile "governance" at the global and national levels (at a minimum), in addition to #EthicalAI constructs for healthcare providers and institutions, are critical and non-negotiable if #aiinhealthcare is going to benefit humanity. This thread has been running for almost a year and captures global, multilateral, national, and sub-national developments in AI Governance.

YR Question 4: What kind of analytics do you use to gather information about your followers, clients, partners, or collaborators? And what type of information do you collect (age, gender, geographic location, etc.)? If you don't collect this information, please explain why.

WR Answer 4: X/Twitter analytics are measured weekly as well as LinkedIn analytics. The broad goal is engagement and growth from patient advocates, clinicians, researchers, and key opinion leaders. Geographic reach (beyond North America) is aspirational. Other demographic information is not tracked.

YR Question 5: Do you use data from follower interactions and their data to improve your project? If yes, do you inform them about it?

WR Answer 5: No, I do not use any follower data. However, I am learning what posts resonate more with followers and work to ensure I am posting across different global time zones using a queuing process.

YR Question 6: Do you use any form of categorization for the users of your platform based on nationality or religious status? If yes, please explain why you do it.

WR Answer 6: No. Once I attract a few thousand followers I will post a bit in French and Spanish since I speak both languages.

YR Question 7: Who assesses the stability of your project’s content, and how often? Do you invite friends or colleagues to conduct independent evaluations? Do you have special medical supervisors for your project?

WR Answer 7: I am solely responsible for all content and make every effort to ensure sources of links are easily identifiable. I make no medical claims or recommendations to consumers or patients. I do receive comments or direct messages from editors or social media managers of reputable journals from time to time and appreciate and support the efforts of ML in Health Science.

Results

YR Question 1: According to my findings, your project currently boasts approximately 5600 original followers (Figure 1 and 2), which is a great result. Do you have any other information about the project's performance?

WR Answer 1: No and it is important to note that this number is an amalgam of direct X-followers and my LinkedIn base which has grown since starting @AI_4_Healthcare in February 2022.

YR Question 2: How many pieces of positive feedback have you received through personal communication? Please explain the information provided within this positive feedback.

WR Answer 2: Several dozen. Usually, kind words in a Direct Message on X or LinkedIn. Likes and reposts are also an indicator of engagement. I do this project for me first and foremost, I have a passion and curiosity for learning. If one new person per day is engaged and learns something as well, it's another step in the right direction.

YR Question 3: How many negative reactions did you receive, and what were the causes of dissatisfaction?

WR Answer 3: None so far. In this effort, people will likely engage or simply ignore a post and move on. If I get something wrong from a factual perspective, well that feedback should be considered a gift, and I will act upon it.

YR Question 4: Please provide information here about your partners and collaborators.

WR Answer 4: ML in Health Science. Am inspired by @CanHealthWatch and closely follow @EricTopol, @NEJM_AI, @JohnNosta, and @DukeAIHealth for content along with a few dozen other handles. This community, if I can call it that, of people around the world with a passion for AI in healthcare and medicine is well informed, passionate, and from what I’ve experienced, collaboration is in its DNA!

YR Response: Thank you!

YR Question 5: Did you need to communicate with official or unofficial authorities to register your project, pay any registration fees, cover promotion costs, manage collaborations, or secure funding?

WR Answer 5: Only paid for my X "blue" checkmark.

YR Question 6: Did you participate in any startup accelerators to secure funding for your project? If yes, what was your experience like?

WR Answer 6: No.

Discussion

YR Question 1: Share your personal experience and satisfaction with your project. Would you like to further promote and improve it? How do you plan to achieve this? Please discuss what you need, for example, human resources or funding, to fuel your journey.

WR Answer 1: This has been a hobby project that has turned into a passion of learning and upskilling. I am confident that consistent posting of interesting content with some commentary/opinion from time to time, will grow a follower base. I have moderated panels, debates, webinars etc. over the course of my career and could see myself doing the same for #aiinhealthcare discussions and providing a pharma/patient/consumer perspective if requested. To be clear, I am no AI expert ... we have enough real ones in the world and others pretending or aspiring to be. ;-)

YR Question 2: Discuss the influence of your project on healthcare system, social life, the official industry field, or your personal environment, such as friends, relatives, etc. Consider any impact on human wellbeing or sustainable development.

WR Answer 2: Family, friends, colleagues, and acquaintances more frequently turn to me for book recommendations, feed/KOLs/authors/sources to follow, or engage me in general #aiinhealthcare conversation. My life sciences clients are also, finally, asking for research summaries in this area/

YR Question 3: Please outline the limitations of your project and how you plan to address them.

WR Answer 3: Time to read and learn amidst the exponential increase of #aiinhealthcare information. I continue to use feeders, GenAI summaries, and prioritize what I deem to be the most insightful and impactful content for posting.

Conclusion

YR Question 1: Please provide a short conclusion in which you briefly discuss the impact of your project on healthcare, social, Web3, and industry fields. Explain why your project is promising and whether it is human-centered or not.

WR Answer 1: Two years in and my followers are growing from a pedantic pace to a daily trickle ... the future looks bright. This project is generating life sciences ecosystem interest and embryonic media attention. Patient group advocates and health charities are reaching out to learn more and ask questions. It is human-centred ... humans and patients need to learn about AI as it will be integrated into every facet of their healthcare experience just like previous technologies such as the stethoscope, x-ray, computers, lab test, wearable technology, etc. This WILL happen and we need to learn to be better informed consumers, advocates, caregivers, and of course, patients.

Conflict of Interest: WR and YR state that no conflict of interest exists.

Disclosure:

This manuscript was created using the Manuscript Self-Creator tool of ML in Health Science.

@AI_4_Healthcare was awarded the ML in Health Science Certificate Human-Centered Project. The asset was donated by the author to water.org. Blockchain certificate:

References

1 European Parliamentary Research Service. Artificial intelligence act. link.

2 European Parliamentary Research Service. Metaverse. Opportunities, risks and policy implications. link. Accessed March 10, 2024.

3 World Health Organisation, ed. Ethics and governance of artificial intelligence for health. Guidance on large multi-modal models; 2024. link.