Week 2
Brainstorm:
- Iām interested in the development of AI the medical field
- What are the laws in regard to AI in medical practice?
- Can AI help nurses when taking care of patients?
- Does AI, in any way, help discrimination and gender gap in healthcare?
- Can a doctor be sued for using AI to treat someone?
Possible Topics:
- Ā AI and Medical malpractice/ Healthcare
- How is AI complicating the medical industry?
- Ā AI and Women in Stem/ gender issues/ development through time
- Ā AI and Environmental Issues (Deforestation, Carbon emissions)
- Ā AI and Political corruption
Two PSAās I like:
- āInvisible killerā Public Awareness Campaign on Air pollution
YouTube. (2018, May 7).āInvisible killerā Public Awareness Campaign on Air pollution. YouTube. ‘Invisible Killer’ Public Awareness Campaign on Air Pollution (youtube.com)
- āTogether, we can find a childā- Wireless AMBER alerts PSA
YouTube. (2011, September 22). āTogether, we can find a childā- Wireless AMBER alerts PSA. YouTube. “Together, we can find a child” – Wireless AMBER Alerts PSA (youtube.com)
Article on AI and topics that interest me:
-How does AI help physicians and nurses
Armitage H. (2024). How AI improves physician and nurse collaboration. Stanford Medicine News Center. How AI improves physician and nurse collaboration | News Center | Stanford Medicine
-AI and Nursing
Pailaha A. D. (2023). The Impact and Issues of Artificial Intelligence in Nursing Science and Healthcare Settings. SAGE open nursing, 9, 23779608231196847. https://doi.org/10.1177/23779608231196847
-AI In healthcareĀ
Bajwa, J., Munir, U., Nori, A., & Williams, B. (2021). Artificial intelligence in healthcare: transforming the practice of medicine. Future healthcare journal, 8(2), e188āe194. https://doi.org/10.7861/fhj.2021-0095
Beginning APA references:
Biever, C. (2023, July 27). The Easy Intelligence Test that AI Chatbots Fail. Nature. Celeste-Biever-AI-Chatbots-Fail-Nature
Crawford, K. (2021). Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press. pp. 1-7. KCrawford-Atlas-of-AI-Intro
Halpern, S. (2023, December 8). The Year A.I. Ate the Internet. The New Yorker. https://openlab.citytech.cuny.edu/profscanlan-english1121-spring2024/files/2024/02/Halpern-AI-at-the-Internet-NewYorker-2023.pdf
Lapore, J. (2023, March 27). The Data Delusion. The New Yorker. The-Data-Delusion-Jill-Lapore-NewYorker
Week 3:Ā Body text, generative AI versions, and comparison essay Ā
The Three Versions of My PSA:Ā
Version 1: My own writing:
Title: AI development helps medical specialists provide better patient care
How can AI help doctors and nurses document patient information and decrease the patient decline rate?
There are great expectations as to how the development of Generative AI can help doctors in the future, nevertheless, at the moment it is mainly being used to facilitate day to day tasks such as patient documentation.Ā Taking notes of patient files by hand can take a doctor around 2 hours; but with the help of GAI, this task takes about 20 minutes. However, there has been a development recently called the deterioration model, and although it has not been perfected, it provides an immense amount of support towards patient care. This model uses the patient’s information to predict the future health status of the patient; then, it sends a notice about the prediction to doctors and nurses so that they can know which patients need more care. Although it is only able to accurately point out about 20% of the patients that will get worse, this model has helped decrease deterioration of patients by about 10.4%. I encourage nurses and doctors to integrate this model into their work as itās already very useful, and there are great hopes that it will develop to great extents and help prevent patientās conditions from worsening.Ā
Ā Ā Ā If nurses and doctors integrate the AI Deterioration Model into their work and GAI systems such as ChatGPT to organize patient data, then they will provide more accurate care to patients, and greatly benefit from faster data storage.
Audience: doctors and nurses
ChatGPT audience #1: Hospital administration (Hospitals)
Prompt: Rewrite the following public service announcement for hospital administration: (input my PSA)
**Public Service Announcement: Advancements in Generative AI for Patient Care**
As we look to the future, the potential of Generative AI (GAI) in supporting our healthcare professionals is promising. Currently, GAI primarily assists with routine tasks like patient documentation. For example, what once took doctors around two hours to complete manually can now be done in approximately 20 minutes with GAI, significantly enhancing efficiency.
A recent development, known as the deterioration model, is also making waves in patient care. While still being refined, this model leverages patient data to predict potential health declines, alerting doctors and nurses to those who may require closer attention. Although it currently identifies only about 20% of patients likely to deteriorate, it has already contributed to a 10.4% reduction in patient decline.
We strongly encourage all healthcare professionals to incorporate this model into their practices. Its current capabilities are already beneficial, and as it evolves, it holds great promise for preventing further health complications. Together, letās embrace these innovations to enhance patient outcomes.
APA Citation:
ChatGPT audience #2: Patients
Prompt: Now rewrite the PSA but for patients as an audience
**Public Service Announcement: Enhancements in Patient Care Through Generative AI**
Weāre excited to share how advancements in Generative AI (GAI) are improving your healthcare experience. Right now, GAI is helping doctors streamline daily tasks, like taking notes on patient files, which used to take around two hours. With GAI, this process is now completed in about 20 minutes, allowing your healthcare providers to spend more time focusing on your care.
One promising development is the deterioration model, which uses your health information to predict future health changes. This model alerts your doctors and nurses when a patient may need extra attention. While it currently identifies about 20% of patients who might experience worsening conditions, it has already helped reduce patient decline by 10.4%.
We encourage you to engage with your healthcare team about these innovations. GAI is here to support better patient care now and, in the future, and together, we can work towards maintaining your health and well-being.
APA Citation:
OpenAI. (2024). ChatGPT (oct 27 version) [Large language model]. https://chat.openai.com/chat
Comparison Essay:
As I started working on my PSA, the focus was on providing useful information as to how the development in the AI field could help nurses complete their tasks and provide better care to patients; However, when searching for information I became more interested in the topic of how the development of AI has become very helpful to the medical industry as a whole. Firstly, I focused on how Generative AI such as ChatGPT has helped healthcare workers, especially doctors, to reduce the time needed to document patient information from 1 hour to about 20 minutes. I wanted to bring this forward to doctors and nurses as an audience since I believe it will help them greatly since they are usually very busy. Secondly, I focused on a new development of AI Called the Deterioration Model which according to Armitage (2024) has helped decrease patient deterioration by about 10.4%. This model does so by looking at patient data such as vitals and alerting nurses and doctors of possible worsening conditions. I believe that along with doctors and nurses, this program will be greatly helpful to improve patient care, specially so if further development occurs. One thing I did worry about my PSA was that it seemed too straightforward and strictly professional so I worried that the audience would be disinterested. Ā Nevertheless, I decided to leave it as so since it is directed to medical workers.Ā
For the First Chat, I was directing the PSA towards hospital administration. In this chat, I really liked that the words were very fluid. For example, with my call to action. Instead of ending like my PSA, āand there are great hopes that it will develop to great extents and help prevent patientās conditions from worsening.ā The chat ended with, āIts current capabilities are already beneficial, and as it evolves, it holds great promise for preventing further health complications. Together, letās embrace these innovations to enhance patient outcomes.ā Which seems more welcoming and provides a more open and positive feeling. I will incorporate part of this ending into my final PSA.
As for the Second Chat, it was even more helpful because as it was directed towards patients. It helped me see what I was pondering (whether to make the PSA more professional or simpler). Ā I loved the Chatās beginning, āWeāre excited to share how advancements in Generative AI (GAI) are improving your healthcare experience.ā Which seems more welcoming when compared with my PSA, āThere are great expectations as to how the development of Generative AI can help doctors in the future,ā which seems too straightforward and boring. It just doesnāt gather peopleās attention as much as a positive beginning. I may or may not decide to incorporate the beginning of the second Chat into my final PSA and mix it with my own writing.
This article by Lohr (2023) was very helpful on providing information in regard to noting patients’ information:
Lohr, S. (2023). A.I. May Someday Work Medical Miracles. For Now, it helps Do Paperwork. The New York Times.
https://www.nytimes.com/2023/06/26/technology/ai-health-care-documentation.html?unlocked_article_code=1.UU4.QHvl.cLea2-3XChko&smid=url-share
This article by Armitage (2024) was very helpful, especially in the part of the deterioration model:
Armitage H. (2024). How AI improves physician and nurse collaboration. Stanford Medicine News Center. How AI improves physician and nurse collaboration | News Center | Stanford Medicine
Additional sources:
Umapathy, V. R., Rajinikanth B, S., Samuel Raj, R. D., Yadav, S., Munavarah, S. A., Anandapandian, P. A., Mary, A. V., Padmavathy, K., & R, A. (2023). Perspective of Artificial Intelligence in Disease Diagnosis: A Review of Current and Future Endeavours in the Medical Field. Cureus, 15(9), e45684. https://doi.org/10.7759/cureus.45684
Partovi, S. (2024). Bridging gaps in healthcare: Three key takeaways from the 2024 Future Health Index. Philips Global.Ā https://www.philips.com/a-w/about/news/archive/blogs/innovation-matters/2024/bridging-gaps-in-healthcare-three-key-takeaways-from-the-2024-future-health-index.html?utm_source=search&utm_medium=paid&utm_campaign=fhi-search-artificial-intelligence&utm_content=rsa-ad-1&utm_term=artificial-intelligence-us&gad_source=1&gbraid=0AAAAA-EHpPlHCt40ruosMsqAW3hRgq2Kz&gclid=Cj0KCQiA_qG5BhDTARIsAA0UHSLXrHQYJ4PDtl8YUCvc5gRhj9DHkfzGgNfJIE2FtTr2fxZh38-VHroaAsmpEALw_wcB
Word count (468)
Week 4: Final PSA
Title:
AI development helps medical specialists provide better patient care
Key Question:
HowĀ can AI help doctor and nurses document patient information and decrease the patient decline rate?
PSA:
There are great expectations as to how the development of Generative AI can help doctors in the future, nevertheless, at the moment it is mainly being used to facilitate day to day tasks such as patient documentation.Ā Taking notes of patient files by hand can take a doctor around 2 hours; but with the help of GAI, this task takes about 20 minutes. However, there has been a development recently called the deterioration model, and although it has not been perfected, it provides an immense amount of support towards patient care. This model uses the patient’s information to predict the future health status of the patient; then, it sends a notice about the prediction to doctors and nurses so that they can know which patients need more care. Although it is only able to accurately point out about 20% of the patients that will get worse, this model has helped decrease deterioration of patients by about 10.4%. We strongly encourage all healthcare professionals to incorporate this model into their practices. Its current capabilities are already beneficial, and as it evolves, it holds great promise for preventing further health complications. Together, letās embrace these innovations to enhance patient outcomes.
Call to Action: Ā
Ā If nurses and doctors integrate the AI Deterioration Model into their work and GAI systems such as ChatGPT to organize patient data, then they will provide more accurate care to patients, and greatly benefit from faster data storage. Additionally, AI is great at Diagnosis and Humans are great at diagnosis but together they can achieve almost perfect diagnosis. Just imagine the possibilities that can be achieved with this union!! So many people can be saved and cured faster!!
Refer to the articles below for more detailed information:
-Lohr, S. (2023). A.I. May Someday Work Medical Miracles. For Now, It helps Do Paperwork. The New York Times.
https://www.nytimes.com/2023/06/26/technology/ai-health-care-documentation.html?unlocked_article_code=1.UU4.QHvl.cLea2-3XChko&smid=url-share
Armitage H. (2024). How AI improves physician and nurse collaboration. Stanford Medicine News Center. How AI improves physician and nurse collaboration | News Center | Stanford Medicine
Reverberi, C., Rigon, T., Solari, A. et al. Experimental evidence of effective humanāAI collaboration in medical decision-making. Sci Rep 12, 14952 (2022). https://doi.org/10.1038/s41598-022-18751-2
Umapathy, V. R., Rajinikanth B, S., Samuel Raj, R. D., Yadav, S., Munavarah, S. A., Anandapandian, P. A., Mary, A. V., Padmavathy, K., & R, A. (2023). Perspective of Artificial Intelligence in Disease Diagnosis: A Review of Current and Future Endeavours in the Medical Field. Cureus, 15(9), e45684. https://doi.org/10.7759/cureus.45684
Reverberi, C., Rigon, T., Solari, A. et al. Experimental evidence of effective humanāAI collaboration in medical decision-making. Sci Rep 12, 14952 (2022). https://doi.org/10.1038/s41598-022-18751-2
Partovi, S. (2024). Bridging gaps in healthcare: Three key takeaways from the 2024 Future Health Index. Philips Global.Ā https://www.philips.com/a-w/about/news/archive/blogs/innovation-matters/2024/bridging-gaps-in-healthcare-three-key-takeaways-from-the-2024-future-health-index.html?utm_source=search&utm_medium=paid&utm_campaign=fhi-search-artificial-intelligence&utm_content=rsa-ad-1&utm_term=artificial-intelligence-us&gad_source=1&gbraid=0AAAAA-EHpPlHCt40ruosMsqAW3hRgq2Kz&gclid=Cj0KCQiA_qG5BhDTARIsAA0UHSLXrHQYJ4PDtl8YUCvc5gRhj9DHkfzGgNfJIE2FtTr2fxZh38-VHroaAsmpEALw_wcB
Reflection:
Although it was quite extensive, I found my PSA project to be easy to complete as long as I followed the steps provided by the professor. I did have a little difficulty when looking for information I wished to have but overall, it was a great project to work on. I work better on long study and work times, so I tried to complete a whole part at once, and since I know what works for me, it was easy to go along with the schedule. In this reflection I will thoroughly contemplate my experience with the project. Focusing on the things I learned, difficulties and simplicities of the project, the changes I would make if provided with more time, and how my thoughts on AI have transformed.
During the time I worked on my project I have learned a lot about how I work on projects. First, I am not able to focus and do a great job on my projects if I do it a little at a time; However, when I focus for a long time on finishing a whole part, it works out greatly. Second, Iām great at looking for sources on the internet and at curating said works. APA citations have become a second nature for me, and I get the information I wish for quite easily because I know how to work terms around. Third, I only use a few tools when working on projects. Especially for hardware since I only used my laptop and phone. As for Software, I used a few more, such as Google docs, WordPress (Open Lab), ChatGPT, and google search. Lastly, I learned that ChatGPT is quite helpful. I liked my own writing on the PSA, and I loved parts that ChatGPT added. As I switched the audience, the Chat made the PSA warmer and more positive which sounded a lot like me, so I was surprised. In my PSA, I didnāt sound like myself as much as I would have liked but combining it with parts of ChatGPT, it sounded so much like me.Ā
There were few difficulties and simplicities were many when working on this project. For difficulties, there were only 2: Finding one source with information I hoped was there and working with ChatGPT. I have never used any LLM AI before, so I was not very comfortable using it. Thankfully I watched a video by Kniberg which helped me with that. After a lot of trial and error, I came to realize that simple prompts work the best if they are specific. As for simplicities, everything else such as brainstorming and making the first draft was easy; I just needed to follow the steps well and everything ran smoothly.Ā
If I had another week, I would work harder on finding a source that provided more information on how the collaboration between humans and AI can increase the rate of correct diagnosis. Additionally, if we were moving forward to posting the PSA somewhere, I would focus on making an appealing design with well decorated borders or background where I would later input my PSA.Ā
Have my views on Generative AI changed as I obtained more in-depth information? Yes, for the most part it has changed drastically. In the past, I have considered using such AI software to be cheating academically. Especially if it is used to cheat on exams. However, through this PSA project, I have come to realize that GAI can also be used for good. Like writing your information and doing your research but asking AI to write it in different ways for a variety of audiences so now I may integrate AI when doing other works. Although, I do prefer trying to do everything by myself since it helps with the growth of my critical thinking and writing skills. Moreover, for other uses of AI, I have not yet decided on a thought. I do believe that by studying more about AI I can teach different conclusions to the one I have stated. But for now, Iāll stick to liking AI if it is used as a tool rather than a replacement. But disliking it because of the probability of it providing wrong information.Ā
(700 words)
Final Citations:
Armitage H. (2024). How AI improves physician and nurse collaboration. Stanford Medicine News Center. How AI improves physician and nurse collaboration | News Center | Stanford Medicine
Biever, C. (2023, July 27). The Easy Intelligence Test that AI Chatbots Fail. Nature. Celeste-Biever-AI-Chatbots-Fail-Nature
Crawford, K. (2021). Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press. pp. 1-7. KCrawford-Atlas-of-AI-Intro
Halpern, S. (2023, December 8). The Year A.I. Ate the Internet. The New Yorker. https://openlab.citytech.cuny.edu/profscanlan-english1121-spring2024/files/2024/02/Halpern-AI-at-the-Internet-NewYorker-2023.pdf
Lapore, J. (2023, March 27). The Data Delusion. The New Yorker. The-Data-Delusion-Jill-Lapore-NewYorker
Lohr, S. (2023). A.I. May Someday Work Medical Miracles. For Now, it helps Do Paperwork. The New York Times. https://www.nytimes.com/2023/06/26/technology/ai-health-care-documentation.html?unlocked_article_code=1.UU4.QHvl.cLea2-3XChko&smid=url-share
OpenAI. (2024). ChatGPT (oct 27 version) [Large language model]. https://chat.openai.com/chat
Partovi, S. (2024). Bridging gaps in healthcare: Three key takeaways from the 2024 Future Health Index. Philips Global.Ā https://www.philips.com/a-w/about/news/archive/blogs/innovation-matters/2024/bridging-gaps-in-healthcare-three-key-takeaways-from-the-2024-future-health-index.html?utm_source=search&utm_medium=paid&utm_campaign=fhi-search-artificial-intelligence&utm_content=rsa-ad-1&utm_term=artificial-intelligence-us&gad_source=1&gbraid=0AAAAA-EHpPlHCt40ruosMsqAW3hRgq2Kz&gclid=Cj0KCQiA_qG5BhDTARIsAA0UHSLXrHQYJ4PDtl8YUCvc5gRhj9DHkfzGgNfJIE2FtTr2fxZh38-VHroaAsmpEALw_wcB
Reverberi, C., Rigon, T., Solari, A. et al. Experimental evidence of effective humanāAI collaboration in medical decision-making. Sci Rep 12, 14952 (2022). https://doi.org/10.1038/s41598-022-18751-2
Umapathy, V. R., Rajinikanth B, S., Samuel Raj, R. D., Yadav, S., Munavarah, S. A., Anandapandian, P. A., Mary, A. V., Padmavathy, K., & R, A. (2023). Perspective of Artificial Intelligence in Disease Diagnosis: A Review of Current and Future Endeavours in the Medical Field. Cureus, 15(9), e45684. https://doi.org/10.7759/cureus.45684