Firstly, the webinar which was about AI and ethics in health was very interesting. The case study was a useful example of AI in healthcare. Moreover, the participants have overwhelmingly ideas so they made the topic discussion smoothly.

However, this AI may not be justified by some participants for particular reasons such as privacy, unemployed situation, and bias of the dataset. I have learned a lot and I think most of the participants are concerned about the privacy, accuracy, and unemployed situation of radiologists in case of replacement of AI. One interesting idea is the equality or accessibility of the AI system. If there are limitations of accessibility, the system would be not standardize nationwide.

To sum up, we also need double-checking by an expert or radiologist to prevent mistakes or errors of the AI.

In case, the image cannot be viewed, please go to this link: https://prnt.sc/12qc32d

My wrap-up for the 1st week topic “AI and Ethics in health care”
I can’t upload my picture to the page, please follow this link https://prnt.sc/12r8hd2
Thank you!

This week I have learned about AI and Ethics in Health. AI system is a modern technology that disrupts the way things work, in this case, clinicians in diagnosis. In this case, the discussion is about using AI to detect and differentiate abnormal CT scan images of COVID-19 patients from others, which is a good example to weigh between risk and benefit from this technology.
The benefits are
– Decrease workload.
– Improve accuracy in diagnosis.
– Reduce time in managing patients.
The concerns are
– Unemployment
– Unnecessary test (in this case CT scan, which the patients exposed to higher radiation)
– Lack of understanding the logic of AI (“Black Box”)
From previous studies, common ethical themes of AI (and my opinion) are
– Privacy and security -> in this example the data collected from patients should require consent from the patients to use, and make sure that the data is used appropriately and kept securely.
– Trust in AI applications -> the developer of the system should declare the intention of using data and privacy, the regulation that this project follows, and limitations that might occur if used in a different situation.
– Accountability and responsibility -> while the health care provider should be responsible for the consequences of the decision made by AI and other information gained from other clinical information, AI developer should be responsible for improving accuracy, decreasing error, and help users (especially health care provider) to understand how the AI makes a decision.
– Bias -> to decrease bias, more data should be input to train AI. In this situation, AI users should be aware of the difference between the system that generates from different settings.

1. Job Loss and Wealth Inequality.
2. AI is Imperfect — What if it Makes a Mistake?
3. Should AI Systems Be Allowed to Kill?
4. Rogue AIs.
5. Singularity and Keeping Control Over AIs.
6. AI Bias.

The image is not show.
You can see the image by follow THIS link.

https://prnt.sc/12sglso

From the article, I would like to summarise into pros and cons of implementing AI in healthcare.

Pros:
1. Faster testing and report
2. Higher throughput
3. More accuracy (depend on data and methods)
4. Increase patient service coverage
5. Remote diagnosis (from telemedicine)
6. Reduce cost (from professional hiring)

Cons:
1. Require technological knowledge
2. Privacy concern
3. Proprietary matter (the data from the subject belong to who?)
4. AI BlackBox (what is the formula?)
5. Selection bias, data from some group of people might not work well with another group
6. Increase cost in computational and storage instrument
7. Require professional personnel to approve the result

In conclusion, I think using healthcare AI is in development and it could be the future method of diagnosis but we need to be aware of potential privacy concerns.

The seminar began with explaining the course objectives and summarizing the contents of the article: AI system for COVID-19 Diagnosis that assigned to students to read and watch the VDO clip presented by Prof. Reidar K. Lie, on Responsible Research and Innovation.
All students who participated this webinar meeting were requested to provide the answers and explanations about the benefits, risks, and concerns in applying AI system for COVID-19 Diagnosis as followings:
– Benefits: improved diagnosis (detection, accuracy), Reduce workload, Reduce cost
– Risks: Bias, Quality of data set (Size, sampling) that use in developing the AI, Confidentiality, and privacy
– Concerns: Unemployed, Ethics
During the discussion, all participants were requested to vote and provide the explanation that would this project be justified as risk-benefit VS. precaution (Better safe than sorry)? 75% of the vote justified this AI system as benefit but there were some concerns as mentioned, and about 80% of students think that this AI system can be applied to other diseases.

Please see infographic in the following link.
http://prnt.sc/12vcaou

Here is my exotic summary!
http://prnt.sc/12yb1x5

This week I have learn about AI and ethics in health.
I think this technology will help both of Physician andRadiological Technologist less work load for screening chest X-ray films on the COVID-19 epidemic situation.
As we discussion about this topic
the benefit of AI are Less workloads, fast diagnostic and report and more accuracy
and risk/concern are misdiagnosis, bias of database, unemployed and data privacy

This week I have learn about AI and ethics in health.
I think this technology will help both of Physician andRadiological Technologist less work load for screening chest X-ray films on the COVID-19 epidemic situation.
As we discussion about this topic
the benefit of AI are Less workloads, fast diagnostic and report and more accuracy
and risk/concern are misdiagnosis, bias of database, unemployed and data privacy

Here’s my wrap-up assignment for this week’s topic discussion. https://prnt.sc/13mlu3b