I just completed this course and I must say it is quite terrific if you are willing to put in the effort to complete it thoroughly.
I have made a mind-map of AI in Medicine (from the course) and am expanding it to cover AI in Healthcare in general. What you see in the green shaded rectangle is the course content.
If anyone wants to collaborate to expand on the overall mindmap for AI in Healthcare, please dm me.
Well done on the effort and the idea! This is really beneficial and informative. Keep it up!
Nice work. Even this is beneficial for others who are visiting the community. Thank you for sharing this to us. You made this learning simple. Keep it up.
Glad you found it useful.
Please do let me know if you have any inputs on improving it further, particularly from students’ perspective.
The doctors here were never impressed, but rather annoyed, because one dares to intrude into their sacred refuge. An annoyed statement of a doctor when I replied something to his diagnosis: “You must not believe everything written on the Internet.” AI power the brave new world!Keep it up…!
Welcome to the community!
I am not sure I am reading your thought process correctly here, but I do agree that clinicians are skeptical and worried about AI, just as most other professions would be too initially.
To a certain degree that worry is not misplaced. The few questions I find that they ask are:
- If AI is to help me reach a diagnosis or treatment in particular and prognosis to an extent, who will be legally responsible for if it does not work out.
- Even if AI did help save time with the mundane tasks, like creating notes at the end-of-the day, a fair amount of time is spent validating those notes, otherwise the whole follow up could turn the wrong way and amplify.
- AI will only be as successful as the data shared and thereby the nature of the medical community, privacy, ethics etc.
So I do agree with you, there are lots of hurdles before doctors - clinicians and the entire supporting healthcare ecosystem starts to “trust” it and use it wisely and widely. But judging from the past, it will eventually happen - to different degree in different disciplines and functions.
Hi @getjaidev This is good. Really like this. Thank you for sharing.
I liked your approached about being balanced about understanding why some doctors are apprehensive about AI that makes one to approach the idea in a more acceptable way.
Although one needs to understand AI or NLP is still a word database algorithm and cannot be matched with a clinician expertise, experience and intuitive decision when a case, emergency or treatment is addressed. So AI can be adjunctive but cannot replace doctors as I know if a patient has come with a broken teeth and bleeding nose, AI cannot help at the emergency situation but doctor can.
I am not criticising AI but one needs to understand what best can be taken from AI to make healthcare more accessible to the common crowd who still do not have access to basic healthcare. Like there are still rural areas where heavy operators machinery are not available in health care setup in India, so if AI can develop a system which can integrate rural healthcare with city healthcare (eg. diagnostic interchange between rural and city via AI generated imaging and prompt prognostic impression can make a rural doctor to decide faster in sending a patient to the nearest community hospital), I think at that time doctors will be the the first person to cheer for artificial intelligence.
For me before joining community learning AI, I had a very limited constricted view about AI but now it has different view and have an understanding where AI could make difference in Health care and food industry which is basic necessity of many poor individual who lack in low economy countries.
By the way, your flow chart is quite impressive. Are you from medicine background or statistical background?
P.S. In your flow chart of prognostic models under challenges you can add Bias or Misinterpretation as it is one of the most common challenge faced by most of the biostatistic analysis where in Bias is again is divided into numerous bias like selection bias and Berksonian bias.
I also wanted to ask why have you limited your evaluation tools to randomised control trials. Case-control studies could also help you in correlation between exposure and disease. Case-control studies on aetiology of disease may help to give the direction of future trials. I am only suggesting, please do not take it offensively
Thanks a lot for your mindmap sharing!
Would you mind sharing more information about your expanding part (the right part of the mindmap)?
Is it your research on gathering the whole picture of ‘AI in Healthcare’, or experiences of yours?
Hi @apple985 ,
Good to see your post and my apologies for the slightly delayed reply.
I am actually looking at drawing up a detailed mind map on the most important facets of “AI in Healthcare” with help from people who have experience in the field. The plan was to build a visual top level idea of the space, and spawn double-click mind-maps, if you will, of specific areas.
I think this very exciting deep-dive track of “AI in Medicine” can benefit from discussions and posts on the width of the space as well. So, if you have any ideas or critique or would like to collaborate in doing so, I am very happy to engage.
Thank you and Best Regards.
A general post for all on or associated with this track of AI in Medicine. If you would like to help:
- Extend the mind-map to cover the broader topic of “AI in Healthcare” in general; and of course
- Update the mind map to the reflect the latest course outline, particularly as you go complete the course, please let me know.
I would be happy to collaborate.
Thanks and Regards.
@apple9855 @Deepti_Prasad, since you had expressed interest in the mind map