Digital Media Concepts/Artificial Intelligence in Healthcare
As technology continues to progress, new advancements persist in the field of medicine. Notably, the ongoing pandemic has sparked the rise of Telehealth usage. With this platform, patients have the ability to communicate with their health specialists without physically being at the clinic. Additionally, AI has paved the path for the early-detection of cancers through the use of mammograms, colonoscopies, and brain imaging to detect abnormalities at a greater performance than Human Intelligence.
Ethics and ConfidentialityEdit
However, the usage of AI in the medical field has brought upon concerns in relation to ethics and confidentiality. Artificial Intelligence has yet to be perfected, and continues to make mistakes and may place unconscious bias in a variety of situation. Thus, initial overconfidence in AI may result in a patient misdiagnoses or lead to data breaches within online Telehealth portals. Furthermore, a 2018 World Economic Forum noted that 75 million jobs are subject to be lost with the implementation of AI in the workforce, including healthcare. Furthermore, Artificial Intelligence can be used to support human decision making and specific health care implementations. Therefore, there are ethical issues related to the rights of deep learning.
- (AI and Decision Support) Discuss the concept of autonomous decision making of algorithms in Health Care in comparison of decision support system where the final decision of diagnose, treatment is performed by medically trained doctors as human decision makers. Who is responsible for decisions that caused a negative impact on the patient?
- the computer scientist who programmed the algorithm,
- the maintainer of the AI application that trained the algorithm with data, so that the AI application "learns" from the data,
- the medical staff that applies the AI algorithm in health care.
- (AI Domains of Application) Discuss different areas of AI application in health care according to the ethics and confidentiality. Compare those areas in terms ethics and confidentiality.
- (Cybersecurity) Cybersecurity and banking have also been argued to be positively impacted by AI. Privacy in health care is important to protect the data of patients. Deep learning has been proven to successfully prevent intrusions from hackers trying to gain access of a platform. Compare this application of AI with machine learning used for pattern recognition in Computer Tomography, Magnet Resonance Imaging. What are the similarities and differences according to the application of AI in health care. E.g. Artificial Intelligence is able to detect threats and disable attackers from gaining private information of patients in comparison to the detection of pattern in CT and MRI images. before they begin hacking.
- (Pattern Recognition in Patient Databases) Look for scientific publication in the health care field in which Artificial Intelligence is used on patient databases. What are pattern in patient databases that help in terms of diagnosis and early detection of diseases that trigger an in-depth analysis of the patient with additional diagnostic workflows to confirm the identified pattern in the data.
- (Visualization and prospective Evolution of a Disease) Deepfakes use Artificial Intelligence technology to edit videos and photos. Look a the example on the right and identify the real image of a person merged into a painting. Assume you have a image of skin disease and you create a prospective evolution of disease how it will look like
- if the disease remains untreated or
- treated with a specific therapy.
- Assume that these projections into the future are based on scientific evidence. How could that be use for educational purpose for medical staff and/or for visualization for patients to understand the impact of a specific suggested therapy. What are the benefits and drawbacks of such an approach?
- Compare the application of Deepfakes with application in the healthcare domain with the example of the skin disease and a prospective evolution of the skin disease in the future. Both images
- the Deepfakes on the right and
- a prospective image in skin disease in a few month if it remains untreated
- do not exist in reality. What do the approaches have in common on the technological level? What are the differences in the underlying ethics?
- (Who is feeding and what is fed into AI algorithms?) Look at the video above (by Kamala Harris) and explain the consequences on healthcare. What do you conclude and want can be done in terms of Risk Management and Quality Assurance?
Artificial Intelligence is meant to mimic Human Intelligence. Its usage has been advantageous in problem solving and creating software and machinery. In healthcare, this means that diagnostic expertise can be hidden in data and at the AI algorithm are used to uncover hidden patterns for treatment and diagnosis in the data even if an explicit rule-based approach with scientific evidence for each applied rule does not exist. For decision making this means that the medical staff are able to identify a specific pattern for treatment or further analytic procedures in areas that would not be addressed by a standard classical workflow in health care. Especially when resource are limited in healthcare and a standard procedure would not make use of a specific analysis, then this additional analysis can be triggered by pattern recognition and assignment of additional limited diagnostic resources can be applied to cases where they could benefit the patient, even if the method are not in the standard medical workflow. Explain the requirements and constraints of such an application of early detection indicators.
The technology has been used in the field of criminal justice, however, ethical issues have risen. Some algorithms have been trained on data that is flawed; the consequences of such has led to systems that exclude African Americans. Due to confirmation bias, the system is searching for criminals and collecting data based only on the information that it has been programmed to search for. In addition, Artificial Intelligence preforms based on the data that it has been given. Not only does this leave to bias in the social justice, but can also brings difficulty in the advancement of COVID-19 protection. Since the Coronavirus has only recently been uncovered, little information regarding COVID-19 risks and preventative measures. Therefore, Artificial Intelligence has failed to bring progress in moving past the pandemic, especially because its usage has produced inaccurate data. Because AI has had difficulty in combating the virus, groups of medical workers have expressed their inability to trust the deep learning programing.
AI Advancements in HealthcareEdit
Every 3 to 4 seconds, on average, a Radiologist interprets a medical image (such as an X-ray) in an 8-hour work day. Currently, Radiologists have benefited from AI and deep learning technology has provided physicians with the ability to learn form data without further interpretation from humans. In doing so, deep learning technology has allowed Radiologists to limit the time spent dissecting images, and more time to have patient-to-doctor dialogues about potential treatment plans. Furthermore, the technology is able to find abstract details unintelligible by the human eye; this data-driven approach can provide further information and can be generalized to recognize phenotypical characters of human tissue, for example.
Artificial intelligence algorithms have accurately diagnosed stratifying patients with potential risk of coronary artery disease. However, a limited studies have been conducted. Therefore, humans have still been proven to be more accurate in detecting the early sings of CAD as compared to AI technology. Wearable technologies have also been created in order to monitor cardiac behavior. Furthermore, this monitoring can contact the physician for any abnormalities within the patient.
In surgery, programed robots have been utilized in maxillofacial surgery. Since the robots have greater precision than the surgeon, less-invasive surgeries have been done solely by the work of a robot. In addition to this, the Maastricht University Medical Center in 2017 was able to use a robot driven by Artificial Intelligence in order to do a microsurgery. This consisted of used an AI-driven robot in a microsurgery intervention. The surgical robot was used to suture blood vessels in a patient affected by lymphedema. The diagnosis of this condition may often lead to the swelling or building up of fluids within the body. With the robot suturing the blood vessels instead of the human surgeon, more accurate movements helped in successfully completing the procedure.
Telemedicine usage, such as Telehealth, has increased with the rise of the recent pandemic. Artificial Intelligence has been able to help patients with chat-bot therapy. Physicians and researches have been using chatbots in order to observe patient mental health, but has yet to show signs of promise in making the communication with their specialist more easy. Psychologists have also noted that the technology needs to make some progress before the chat-box therapy methods can provide legitimate support without the continual overwatch of psychologists. Telemedicine has also brought benefit for the elderly. Tools such as sensors that detect an abnormal heart beat or even a possible patient collapse can send immediate alerts to the physician and paramedics so bring immediate care to the patient. However, concerns related to privacy have arise with this technology again. The user may have little control over the type of information being collected through their sensors, and the information might be sent to sources that are not trusted by the patient.
Ethics of AI for HealthEdit
Many critics have stated that research and government overwatch is lacking in reviewing the ethical concerns related to Artificial Intelligence. For instance, the use of Artificial Intelligence to replace patient screenings removes the physician interactions with patients. Furthermore, deep learning processing has not yet reached a point in which it can mimic human emotions. Moreover, trusting AI to address patient concerns without weightage of human emotions can impact the treatment plan of patients. Artificial Intelligence cannot resemble human compassion and wisdom. However, studies have shown that nearly 35% of jobs in the UK within the next 10 to 20 years are subject to loss due to Artificial Intelligence.
- ↑ 1.0 1.1 Hosny, Ahmed; Parmar, Chintan; Quackenbush, John; Schwartz, Lawrence H.; Aerts, Hugo J. W. L. (2018-8). "Artificial intelligence in radiology". Nature reviews. Cancer 18 (8): 500–510. doi:10.1038/s41568-018-0016-5. ISSN 1474-175X. PMID 29777175. PMC 6268174. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6268174/.
- ↑ Chu, Linda C.; Anandkumar, Anima; Shin, Hoo Chang; Fishman, Elliot K. (2020-10-01). "The Potential Dangers of Artificial Intelligence for Radiology and Radiologists". Journal of the American College of Radiology 17 (10): 1309–1311. doi:10.1016/j.jacr.2020.04.010. ISSN 1546-1440. PMID 32360451. https://www.jacr.org/article/S1546-1440(20)30403-8/abstract.
- ↑ "Pros & Cons of Artificial Intelligence in Medicine". College of Computing & Informatics. 2021-08-17. Retrieved 2022-03-03.
- ↑ Vayena, E., Blasimme, A., & Cohen, I. G. (2018). Machine learning in medicine: addressing ethical challenges. PLoS medicine, 15(11), e1002689.
- ↑ Currie, G., Hawk, K. E., Rohren, E., Vial, A., & Klein, R. (2019). Machine learning and deep learning in medical imaging: intelligent imaging. Journal of medical imaging and radiation sciences, 50(4), 477-487.
- ↑ Calderon, Ricardo (2019-01-15). "The Benefits of Artificial Intelligence in Cybersecurity". Economic Crime Forensics Capstones. https://digitalcommons.lasalle.edu/ecf_capstones/36.
- ↑ The Gale Encyclopedia of Psychology (Vol. 1. 3rd ed.) Publisher: Gale, a Cengage Company
- ↑ Naudé, Wim (2020-09-01). "Artificial intelligence vs COVID-19: limitations, constraints and pitfalls". AI & SOCIETY 35 (3): 761–765. doi:10.1007/s00146-020-00978-0. ISSN 1435-5655. PMID 32346223. PMC PMC7186767. https://doi.org/10.1007/s00146-020-00978-0.
- ↑ Chen, Wei; Sun, Qiang; Chen, Xiaomin; Xie, Gangcai; Wu, Huiqun; Xu, Chen (2021-05-26). "Deep Learning Methods for Heart Sounds Classification: A Systematic Review". Entropy 23 (6): 667. doi:10.3390/e23060667. ISSN 1099-4300. PMID 34073201. PMC 8229456. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8229456/.
- ↑ Shaheen, Mohammed Yousef (2021-09-25). "AI in Healthcare: medical and socio-economic benefits and challenges". ScienceOpen Preprints. doi:10.14293/S2199-1006.1.SOR-.PPRQNI1.v1. https://www.scienceopen.com/hosted-document?doi=10.14293/S2199-1006.1.SOR-.PPRQNI1.v1.
- ↑ "Contributed: The power of AI in surgery". MobiHealthNews. 2021-11-23. Retrieved 2022-03-04.
- ↑ Hamet, Pavel; Tremblay, Johanne (2017-04-01). "Artificial intelligence in medicine". Metabolism. Insights Into the Future of Medicine: Technologies, Concepts, and Integration 69: S36–S40. doi:10.1016/j.metabol.2017.01.011. ISSN 0026-0495. https://www.sciencedirect.com/science/article/pii/S002604951730015X.
- ↑ Pouke, Matti; Häkkilä, Jonna (2013-12). "Elderly Healthcare Monitoring Using an Avatar-Based 3D Virtual Environment". International Journal of Environmental Research and Public Health 10 (12): 7283–7298. doi:10.3390/ijerph10127283. ISSN 1661-7827. PMID 24351747. PMC 3881167. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3881167/.
- ↑ Yang, Misti (2020-3). [https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7441483/ "Painful conversations: Therapeutic chatbots and public capacities"]. Communication and the Public 5 (1-2): 35–44. doi:10.1177/2057047320950636. ISSN 2057-0473. PMC 7441483. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7441483/.
- ↑ Davenport, Thomas; Kalakota, Ravi (2019-6). "The potential for artificial intelligence in healthcare". Future Healthcare Journal 6 (2): 94–98. doi:10.7861/futurehosp.6-2-94. ISSN 2514-6645. PMID 31363513. PMC 6616181. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6616181/.