|
 
 |
| ORIGINAL ARTICLE |
|
| Year : 2020 | Volume
: 1
| Issue : 1 | Page : 13-19 |
|
Artificial intelligence in healthcare; its knowledge, practice, and perception among medical personnel in the developing economy
Ebbi Donald Robinson
Department of Radiology, Rivers State University Teaching Hospital, Port Harcourt, Rivers State, Nigeria
| Date of Submission | 15-May-2020 |
| Date of Decision | 28-May-2020 |
| Date of Acceptance | 04-Jun-2020 |
| Date of Web Publication | 25-Sep-2020 |
Correspondence Address:
Dr. Ebbi Donald Robinson
Department of Radiology, Rivers State University Teaching Hospital, Port Harcourt, Rivers State
Nigeria
 Source of Support: None, Conflict of Interest: None
DOI: 10.4103/JRMT.JRMT_1_20
Background: The world is remodeling with the emergence of the most revolutionary technological advancement called artificial intelligence (AI). The study is to evaluate the knowledge, practice, and perception of AI in healthcare among care providers. Methodology: A descriptive, prospective design was adopted over a period of 8 months using a well-structured, self-administered questionnaire administered to medical practitioners. A total of 510 questionnaires out of 525 administered were retrieved immediately after completion. The responses were collated and analyzed using SPSS version 21.0 in line with the study objectives. The data were presented orderly in frequency tables and charts. Results: Male constituted 59.02%, while the female was 40.98%. The mean (±standard deviation) age of the respondents is 45.88 ± 10.26 years. 21.18% of the respondents are consultants, 42.35% are resident doctors while 9.22% are medical officers and 27.25% constituting interns. 94.31% of the respondents know about AI, and the internet has been the main source (89.32%). Majorities know that AI is applied in radiology (96.50%) and surgery (94.05%). Four hundred and thirty-seven (91.81%) responded that there was no form of AI in their facilities and are willing to be trained. Two hundred and sixty-eight (55.83%) agreed that AI would improve healthcare services, while 43.66% of the respondents thought that it will eliminate the human factor. Poor knowledge of information technology, absence of legislation, and promotion of self-medication were the anticipated challenges. Conclusion: Little is known about AI use in pathology, drug-dispensing, and nursing care. There is no AI practice in this environment, but physicians are willing to embrace AI. The fear of eliminating the human factor and cause job loss remained a concern. Training and legislation to regulate AI use are recommended to improve the knowledge, practice, and perception of AI among medical personnel in the developing world.
Keywords: Artificial intelligence, developing world, medical personnel, Rivers State University Teaching Hospital
How to cite this article:
Robinson ED. Artificial intelligence in healthcare; its knowledge, practice, and perception among medical personnel in the developing economy. J Radiat Med Trop 2020;1:13-9 |
How to cite this URL:
Robinson ED. Artificial intelligence in healthcare; its knowledge, practice, and perception among medical personnel in the developing economy. J Radiat Med Trop [serial online] 2020 [cited 2023 Mar 26];1:13-9. Available from: http://www.jrmt.org/text.asp?2020/1/1/13/296106 |
| Introduction | |  |
Artificial intelligence (AI) is one of the most revolutionary technological advancements to emerge in recent times that is remodeling and shaping the world. AI has the propensity of making impossibilities of the past possible. With the advent of AI technology, there will be continual discovery of new possibilities and opportunities. The field of AI dates back to 1956 at a workshop at Dartmouth College where the terminology AI was first postulated by John McCarthy.[1]
At present, in every aspect of human endeavor, AI has played a substantial role in the perception, processing, and interpretation of information.[2] AI, also known as machine intelligence, is the ability of machines to demonstrate intelligence likened to that exhibited by humans.[3] No wonder, it is also often referred to as machines' ability to demonstrate cognitive function which is associated with human thought.[3] With AI, machines now interpret, analyze, and process complex data systems in a variety of applications, including web search, self-driving vehicles, and computer vision and most especially in healthcare.[2]
It is the process whereby human intelligence process is stimulated by machines to have capability of understanding and interpreting human speech as well as vision.[4],[5] The initial AI applications were not able to match the human intelligence, but with improvement in the application of AI with the advent of the deep learning algorithm, AI has surpassed the human intelligence and has become more proficient in certain applications that has to do with specific activities.[2] With this development, it is predicted by research that, in the coming decade, AI will perform automated tasks such as language translation, performing of surgeries, as well as writing books.[2]
AI methods that are commonly used today are classified into two: handcrafted-engineered feature and deep learning. Handcrafted-engineered feature is the use of mathematical equations that is qualified with computer programs.[2] Deep learning has progressively gained recognition because of its ability to learn features automatically from available data without prior definition of the concept by human expertise.[2]
Ian Goldin, a Professor of Globalization and Development at Oxford University, disclosed to BBC that AI will lead to millions of people losing their jobs resulting to a decline in economy in developing countries like Africa.[6] In that report, he likened AI to a disruptive technology that will impact negatively on emerging economy by truncating the traditional pattern of economic growth, wherein robots will be used to perform low wage jobs.[6] It was further reported that workforce and responsibility that is based on laid down rules, repetitive, and does not need human dexterity or human empathy can be replaced by AI; therefore, it can replace millions of jobs in car industry sooner than expected.[6]
A study conducted by BBC at Oxford revealed that AI could lead to 40% of jobs loss in Europe within 10 years whereas up to half of jobs will be lost in the USA, and to a greater degree in the developing parts of the world.[6] Technological advancement has contributes to immensely to global economic dominance and power by the advanced economies. This will eventually make the developing economy to be more and more dependents on the developed economy.[6] A research published by the Accenture Institute for High Performance revealed that by 2035, AI will lead to very rapid yearly economic growth in the developed world.[7],[8] According to a New York-based innovation business visionary Jack Hidary, he said that AI will be of great advantage by improving the quality of lives of the world's poor economy.[7],[8]
A Vanguard publication in 2019[5] documented that since the inception of the concept of AI, most developed countries have keyed into its application; however, its concept is still not very familiar with many Nigerians. This report is similar to that obtained in other developing economy. An information and communication technology (ICT) expert mentioned in that report that this is the apt time for Nigeria as a nation need to key into the concept of AI like other developed countries and utilize the opportunities therein.[5] Adding that, most people in Nigeria still perceive AI as a fiction of science that will rule the world and replace the jobs done by humans;[5] however, contrary to that perception, the ICT expert documented with optimism that AI will not replace humans.[5]
The application of AI in the health industry has been of immense contribution to the quality of healthcare. It is documented that AI application is involved drug discovery, medical diagnosis, routine patient monitoring, and risk management.[2] Notwithstanding there are some challenges regarding the implementation of this technology, it is believed that AI application will save more lives and improve lives by revolutionizing the inefficient healthcare delivery systems.[8]
It is employed in various fields of medicine including ophthalmology, surgery, pathology, dermatology, and ophthalmology, as well as hospital management.[2] According to a recent study in Children's National Medical Center in Washington, a robot successfully performed an open surgery under the supervision of a team of surgeons, and the most interesting aspect is the acceptance of the team of surgeons that the surgery performed by the robot was better than that performed by human surgeons.[9] Studies have revealed that AI is capable to precisely making diagnose as good as a well trained medical personnel for cancer of the skin,[10] monitor patients at high risk of cancer,[11] and perform ophthalmologic diagnosis within 30 seconds.[12]
AI is seen as the best way to understand the pattern of disease spread, as well as provide support to patients.[13] The future of AI is advancing to distinguish the cause of infection, find the cure for disease, envisage a possible outbreak, and also predict the possible area of the outbreak, all with very negligible human input. These were evident by the use of AI technology in the management of the Ebola virus in the past.[7],[8]
The management of autism has been very challenging for both parents and care providers, especially as it regards communication.[14] Autism is a disorder associated with developmental delays associated with impairment of the ability to form normal social relationship and skills. Therefore, care providers use behavioral therapy to manage the communication difficulties associated with the disorder. Research has now found that, with the help of AI, these children with autism can now develop skills that can help them effectively communicate.[14] It was contained in the publication that a computer model known as socially assistive robotics is designed with AI technology to provide personalized verbal and visual support to children with autism.[14] The study also documented that a computer model was about 90% accurate in acknowledging the appropriate time to provide children with information and to get feedback from the child about a task.[14]
The potential of AI is on a continual growth pathway as the day goes by.[15] This is evident by the recent report by a Chinese Tech Company which documented that using a computed tomographic scan having an AI algorithm can detect the deadly coronavirus within 20 s as against humans that will need up to 20 min to make the diagnosis of the disease.[15] According to the report, the model developed has about 96% precision rate.
AI is the trending technology worldwide. Notwithstanding this advancement, little about AI is applied in developing economy. There is also a paucity of documentation about the application of AI in healthcare in our environment. It is therefore pertinent to evaluate the knowledge, practice, and perception of AI in the healthcare system of developing countries like Nigeria.
| Methodology | | |
A descriptive, prospective design was adopted for the study over a period of 8 months in two tertiary institutions in Port Harcourt, South-South Nigeria. The two tertiary institutions include the Rivers State University Teaching Hospital and the University of Port Harcourt Teaching Hospital. Medical practitioners working in the hospital participated in the study, irrespective of their department and carder.
A well-structured, self-administered questionnaire developed in English was the instrument of a data collection. The questionnaire was first pretested among healthcare providers to ascertain its validity and reliability. The pretesting was aimed at ensuring that the instrument of data collection was well structured, logical in sequence, simple to understand the respondents, and devoid of ambiguity. To ensure the questionnaire stand the test of validity, it was also sent to other experts in the field of AI and ICT for modifications and validation.
The questionnaire was administered manually to participants during clinical meetings, mortality meetings, forum discussion, and gathering of health personnel and retrieved immediately after proper completion.
Data analysis
Of the 525 questionnaires distributed, 510 were retrieved giving a response rate of 97.14%. The responses were collated from the questionnaires and analyzed using SPSS version 21.0 in line with the knowledge, practice, and perception of AI in healthcare by medical personnel. The data were presented orderly in frequency tables as percentages, charts, and graphs.
| Results | | |
A total of 510 questionnaires out of the 525 distributed were retrieved. Males constituted 59.02% while female 40.98% [Figure 1] and [Table 1]. The mean (±standard deviation) age of the respondents was 45.88 ± 10.26 years, while that of male and female respondents was 46.19 ± 10.71 and 45.44 ± 9.59 years, respectively [Table 1]. Those within the age classification of 20–29 years constituted 5.29% whereas age group 40–49 years was the largest group constituting 32.35% as shown in [Figure 2] and [Table 1]. 27.65% of the respondents are single, 70.20% are married, while 1.18% of the respondents are divorced, and 0.58% separated [Table 2]. 21.18% of the respondents are consultants, 42.35% are resident doctors, while 9.22% are medical officers and 27.25% constituting house officers [Table 2].
[Table 3] shows that 481 (94.31%) respondents have heard about AI while 17 (3.33%) do not know about it. As shown in [Table 4], 4.64% and 7.18% of males and females, respectively, do not know about AI. All the medical consultants are aware of AI, while 95.383% of resident doctors know about AI [Table 4]. Meanwhile, a majority (54.36%) of the respondents do not know about deep machine learning [Table 3]. Two hundred and twenty-nine (47.22%) respondents know that AI is also known as machine intelligence while 47.84% do not have idea about it [Table 3]. Approximately half (49.9%) of the respondents agree that AI is the ability of machines to function as humans in learning and problem-solving, whereas 2.83% do not agree and 47.27% do not have idea about it. As also shown in [Table 3], 49.9% of the respondents agree that AI is intelligence demonstrated by machines in contrast to the natural intelligence displayed by humans. 3.61% do not agree with the statement while 46.49% do not know [Table 3]. | Table 4: Knowledge of artificial intelligence among sex and carder of respondents
Click here to view |
In [Figure 3], it is documented that 89.32% had the knowledge of AI through the internet whereas 4.72% got informed about AI in conferences and 2.46% from the clinical meeting [Figure 3]. Among the respondents who know about AI, 96.50% knew that AI is applied in radiology, 94.05% surgery, 10.06% ENT, 11.90% pathology, 9.24% hospital management, and 5.34% ophthalmology. No respondents know about the application of AI in drug-dispensing and nursing care [Figure 4]. | Figure 4: Aspects of healthcare where artificial intelligence is applied
Click here to view |
[Table 5] reveals that 91.81% of the respondents had no AI being practiced in their facility while 8.19% do not know whether there is or not. 85.08% of the respondents documented that there is no AI practice in their environment. 95.83% and 91.42% have not had any formal or informal training, respectively, regarding the application of AI. However, only 2.08% have had some form of awareness and sensitization about AI practice and almost all the respondents (99.58%) are willing to embrace AI, if training is provided.
Concerning the perception of AI, 55.83% of the respondents agreed that AI would improve healthcare services in their facility while 32.50% opined that it will not improve healthcare services [Table 6]. 43.66% of the respondents think that AI will eliminate the human factor in workplace, whereas 82.29% document that it will lead to job loss. Majority of the respondents (97.92%) envisage various types of challenge with the practice of AI in developing economy [Table 6].
According to [Table 7], respondents envisaged some challenges that include poor knowledge of information technology (82.74%), no legislation and control over the software (58.84%), poor internet connectivity (71.93%), and the promotion of self-medication (39.29%). | Table 7: Envisaged challenges of artificial intelligence practice in developing economy
Click here to view |
| Discussion | | |
AI is the ability of machines to demonstrate intelligence like humans and also demonstrate cognitive function which is associated with human thought.[3] The advent of AI has changed the world's perception and processing of information and has also surpassed the human intelligence.[2] AI is revolutionizing the healthcare delivery system with respect to doctor–patient interaction, clinical diagnosis, and treatment of patients. Its application in the health industry has been of immense contribution to improve the quality of healthcare delivery in drug discovery, diagnosis, and routine patient management.[2]
A majority knew about AI but do not have in-depth knowledge. This was evident by the fact that most of the respondents do not understand that AI is also known as deep machine learning. The index study shows that internet is the major source of information about AI, followed by conferences. These suggest that AI is barely discussed and not giving the attention it deserved in clinical meeting among medical personnel.
Most of the respondents know that AI is applied in radiology and surgery, and very few know about the application of AI in other aspects of medicine and healthcare such as pathology, hospital management, drug-dispensing, as well as nursing care. There is no form of AI practice in any of the facilities within the environment of study, and most of the respondents have not had any form of training both formally or informally about this revolutionizing technology in their facilities. Adequate sensitization regarding AI is also not done in the facilities. Almost all the respondents are, however, willing to embrace and apply AI in their practice if giving the necessary training.
The study revealed that AI would improve healthcare services in their facility. However, there are the fears that it has the propensity of eliminating the human factor. This concern was also shared in the article “Artificial intelligence and the future of humans” by Anderson and Rainie.[16] The article documented a great concern regarding the long-term effect of AI on human beings. Although the article did not expressly specify that AI will eliminate the human factor, the article documented uncertainly of the human element in the near future.
The resultant effect of job loss to AI is a great concern expressed. This was in agreement with the report of Ian Goldin[6] that AI will lead to millions losing their jobs with resultant economic decline in developing countries like Africa. However, on the contrary, Bernard[17] documented that AI will change the world order in respect to the type of jobs performed by humans and will not necessarily cause outright job loss. Another publication by the Bangkokpost[18] titled “the impact of artificial intelligence on humans” reported that AI will actually cause job evolution. The publication[18] also reported a forecast which predicted that 7 million already existing jobs in the UK will be replaced by AI by 2037 while 7.2 million jobs will be newly created by AI. This gives an overall indication that there will be a loss in conventional jobs with the evolution of new jobs. The disparity of these reports and the present study may be attributed to lack of in-depth knowledge of AI and information technology.
There is the wiliness to embrace the application of AI in practice, although majority envisage various forms of challenges. These challenges documented in the index study are poor knowledge of information technology as there is no information technology inputted in the curriculum of medical education of all cadres in this environment. The report by Arinya[18] acknowledges in context the illusion of individuals to this technology advancement and recommended the need for personal development of soft skills to obtain a more satisfying work in this new era of technological evolution. The report recommended that training should start in schools to enable individuals understand the machine language and work alongside the machines.
The study revealed that there was a need for proper legislation over the control of AI machines and software applications. Most radiologists believe that the application of AI in radiology is meant to assist a radiologist; however; it will impact negatively on the profession of radiology when the software/technology are handled by a nonradiologist. The absence of regulations will also promote self-medication. This skepticism of absence of rules and regulation was also addressed in the report of Bernard.[17] In that report,[17] Nick Bostrom of Oxford University advised the need for a scalable control over AI. It was mentioned that now is the time for government and business institutions to make regulations and formulate rules that will control the use of AI. The report[17] further stated that if this is not done now, there will be significant negative impact in the near future.
| Conclusion | | |
Notwithstanding majority know about AI, only very few have in-depth knowledge about AI and the internet is the major source of information. Majority are aware of the application of AI in radiology and surgery, but only few know about its application in pathology, hospital management, drug-dispensing, as well as nursing care. There is no form of AI practice in this environment; however, medical personnel are willing to embrace AI in their practice. It is believed that AI will improve healthcare services, but there are fears that it may eliminate the human factor and cause job loss. Poor knowledge of information technology and absence of proper legislation and regulations are identified concerns.
The study recommends that information technology should be encompassed in medical education curriculum with training and re-training of healthcare providers in the use and application of AI. There is need for proper legislation to control the use of AI software and machines to allay the fear of clinicians, with regard to quackery and self-medication. These will improve the knowledge, practice, and perception of AI in healthcare industry of the developing world.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
| References | | |
| 1. | Crevier D. AI: The Tumultuous Search for Artificial Intelligence. New York, NY: Basic Books; 1993.  |
| 2. | Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts HJ. Artificial intelligence in radiology. Nat Rev Cancer 2018;18:500-10. |
| 3. | |
| 4. | Russell SJ, Norvig P. Artificial Intelligence: A Modern Approach. 3 rd ed. Upper Saddle River, New Jersey: Prentice Hall; 2009. |
| 5. | |
| 6. | |
| 7. | |
| 8. | |
| 9. | |
| 10. | |
| 11. | Dina B. Microsoft Develops AI to Help Cancer Doctors Find the Right Treatments. Bloomberg; 20 September, 2016. [Last accessed on 2017 May 11]. |
| 12. | Gallagher J. Artificial Intelligence as Good as Cancer Doctors. BBC News; 26 January, 2017. [Last retrieved 2017 Jan 26]. |
| 13. | Agrebi S, Larbi A. Use of artificial intelligence in infectious diseases. Artificial Intelligence Precision Health 2020:415-38. doi:10.1016/b978-0-12-817338-1.00018-5. |
| 14. | |
| 15. | |
| 16. | |
| 17. | |
| 18. | |
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7]
|
Search Pubmed for