Greener Journal of Medical Sciences

Vol. 11(1), pp. 37-45, 2021

ISSN: 2276-7797

Copyright ©2021, the copyright of this article is retained by the author(s)

https://gjournals.org/GJMS

 

 

 

 

 

Primary health care physicians’ perception of artificial intelligence in health care

 

 

Medhat El-Shazly1; Maha Al-Majed2; Ahmad Al-Omar3; Marwa Abdalla4; Maria Pereira5

 

 

1 MD, Consultant of Public Health, MOH, Kuwait and Professor of Health Statistics, Medical Research Institute, Alexandria University, Egypt.

2 MSc, Specialist of Public Health, Head of Planning and Health System Research Department, MOH, Kuwait

3 MRCGP, Specialist, Health System Research Department, MOH, Kuwait

4 MSc, Department of Planning and Follow-up, MOH, Kuwait

5 Secretary, Department of Planning and Follow-up, MOH, Kuwait

 

 

ARTICLE INFO

ABSTRACT

 

Article No.: 032821030

Type: Research

 

 

Background: Artificial intelligence (AI), broadly defined as the imitation of human cognition by a machine. Despite the early success of AI in performing and assisting with clinical tasks, there is still skepticism about the potential of this technology.

Objectives: The aim of this study is to highlight primary health care (PHC) physician’s perception towards AI.

Subjects and Methods: The study design is a cross-sectional multicenter using predesigned self-administered questionnaire to collect data from all the physicians working in 41 PHC centers. The questionnaire included personal data as well as physician’s perception of AI regarding using in PHC, fields of application, possible risks, and their opinion if AI would replace physicians in various tasks of PHC. Simple descriptive statistics were used (mean + standard deviation for quantitative variables, and frequency with percentage distribution for categorized variables).

Results: Within the 531 physicians working in the selected settings and invited to participate in the study, 369 responded with a response rate of 69.5%. Within the participants, 37.3% were males, 42.9% were < 35 years old, 57.1% were Kuwaitis. Less than half of the participating physicians (46.2%) considered themselves familiar with AI. A higher proportion of physicians (71.7%) believed that AI had useful applications in the medical field, only 11.8% agreed that the diagnostic ability of AI is superior to the clinical experience of a human doctor and only 5.0% agreed that AI could replace their jobs. When physicians were asked about the most useful expected application of AI in the field of medicine, came at the top of the list biopharmaceutical research and development (30.5%) followed by diagnosis (24.1%). The proportion of physicians who believed that it is likely that technology will be able to fully replace human physicians was 79.5% regarding documentation about patient, 55.5% regarding establishing prognoses, 45.9% for referral of patients, and 44.0% for diagnosis.

Conclusion: Despite the agreement among participants regarding the impact of AI on the health care system, physicians still have skepticism about the potential of this technology and demonstrate a more negative attitude.

 

Accepted:  03/03/2021

Published: 08/04/2021

 

*Corresponding Author

Prof. Dr. Medhat El-Shazly

E-mail: medshaz@yahoo.com

Phone: +96566612524

 

Keywords: PHC; physicians; Perception; AI

 

 

 


 

 

Introduction:

 

Artificial intelligence (AI) in healthcare is the use of complex algorithms and software to estimate human cognition in the analysis of complicated medical data. Specifically, AI is the ability for computer algorithms to approximate conclusions without direct human input. (Russell S and Norvig, 2020; Esteva, 2019) Simply, it is the ability of computer systems to perform tasks that would usually require human levels of intelligence. (Shameer et al., 2018) A subfield of AI is machine learning, which can be used to teach a computer to analyze a vast amount of data in a rapid, accurate, and efficient manner through the use of complex computing and statistical algorithms. (Legg and Hutter, 2007)

What distinguishes AI technology from traditional technologies in health care is the ability to gain information, process it and give a well-defined output to the end-user. (Bakkar et al., 2018; Trivedi et al., 2018) The primary aim of health-related AI applications is to analyze relationships between prevention or treatment techniques and patient outcomes. AI programs have been developed and applied to practices such as diagnostic processes, treatment protocol development, drug development, personalized medicine, and patient monitoring and care. Additionally, hospitals are looking to AI solutions to increase cost saving, improve patient satisfaction, and satisfy their staffing and workforce needs. (Kent, 2018)

Despite the early success of AI in performing and assisting with clinical tasks, there is still skepticism about the potential of this technology. Although the scientific literature supporting AI in clinical medicine is accumulating quickly, few previous studies have described how physicians perceive the advent of this technology (Sarwar et al., 2019) It is expected that AI will be used extensively in the medical field in the future leading to fundamental changes in the role of physicians and the way they practice medicine. The use of AI for the development of decision support systems is an emerging area that may help to refine algorithms, but more research is needed on acceptability, feasibility and ethics. (Jha, 2016) However, there is no research, to our knowledge, on the opinions and perception of the application of AI programs in the medical field among Kuwaiti physicians. The purpose of this study is to highlight primary health care physicians’ perception of AI.

 

 

SUBJECTS AND METHODS:

 

Primary health care (PHC) in Kuwait is provided through 104 centers distributed within 5 administrative health regions corresponding to the number of populations in each region. Two health regions were selected randomly to conduct this study (Capital and Hawalli). All health centers in the selected health regions participated in the study (41 centers). The studied population consisted of all physicians working in these centers (531 during the period of study (1 August to 30 September 2019)

A cross-sectional multicenter approach was used in this survey. A predesigned self-administered questionnaire was used to collect data from the physicians. The questionnaire was derived from other published studies dealing with the same topic as well as from our own experience. It was sent out to all physicians working in all PHC centers in the selected health regions who were asked to participate in the study. Beside participants’ personal data, the questionnaire included four sections aiming to evaluate their perception of AI. The first section of the survey enquired about physician’s attitude toward the medical application of AI. The second included the medical fields in which AI could be applied. The third one entailed the possible problems regarding the application of AI in medicine and the fourth section asked the physicians if they believe that AI would replace physicians in the various tasks of PHC. Participants who believed that replacement of a given task was likely were asked to give an estimate of the timescale required for replacement. Participants were asked to respond to every question to complete the survey.

All the necessary approvals for carrying out the research were obtained. The Ethical Committee of the Kuwaiti Ministry of Health approved the research. A pilot study was carried out on a small sample of physicians to test the clarity, applicability and suitability of the used questionnaire, and to identify the difficulties that could be faced during the application. The Statistical Package for Social Sciences (SPSS-27) was used for data processing. Simple descriptive statistics were used (mean ± standard deviation for quantitative variables, and frequency with percentage distribution for categorized variables).

 

 

RESULTS:

 

Within 531 physicians working in the selected settings and invited to participate in the study, 369 responded with a response rate of 69.5%. Twelve questionnaires were excluded as they were not complete. The final analysis included 357 participants.

 

General characteristics of the participating physicians:

 

Within the participants, 37.3% were males, 42.9% were < 35 years old (mean = 42.4 ± 9.3, Min – Max: 27.0 – 69.0), 57.1% were Kuwaitis, 42.3% had a Bachelor degree. General practitioners (GP), family physicians (FP), diabetologist, maternity physicians and dentist constituted 47.9%, 25.8%, 3.9%, 5.0%, and 17.4 of the studied population. The years of working experience ranged from 1 to 46 with a median of 9 years. Half of participants worked for 30-39 hours /week with a median of 36 hours. The number of patients seen / day ranged from 12 to 100 with a median of 25 patients, whereas 38.7% of participants examined ≥40 patients / day.

 


 

Table 1: General characteristics of the participating physicians

Variable

 

No.

%

Gender:

Male

134

37.3

 

Female

223

62.5

Age:

<35

153

42.9

 

35-44

109

30.5

 

45-54

53

16.2

 

≥55

37

10.4

Nationality:

Kuwaiti

204

57.1

 

Arabic

136

37.8

 

Others

18

5.0

Education:

Bach

151

42.3

 

Master

117

32.8

 

PhD / Board

89

24.9

Job:

GP

171

47.9

 

FP

92

25.8

 

Diabetologist

14

3.9

 

Maternity

18

5.0

 

Dentist

62

17.4

Years of experience:

<5

73

20.4

 

5-9

106

29.7

 

10-14

73

20.4

 

>=15

105

29.4

Working hours / week:

<30

20

5.6

 

30-39

183

51.3

 

40-49

129

36.1

 

>=50

25

7.0

Patients seen / day:

< 20

71

19.9

 

20-29

114

31.9

 

30-39

34

9.5

 

≥40

138

38.7

Total

 

357

100.0

 


 

 

Physicians’ attitude towards AI in health care setting:

 

Less than half of participating physicians (46.2%) considered themselves familiar with AI. A higher proportion of physicians (71.7%) believed that AI had useful applications in the medical field and 58.5% believed that AI will improve the public image of physicians whereas 51.3% of them agree that AI could provide more time for physicians for training and research. On the other hand, 27.2% of physicians agreed that they would always use AI when making medical decisions in the future. Only 11.8% agreed that the diagnostic ability of AI is superior to the clinical experience of a human doctor and only 5.0% agreed that AI could replace their jobs completely. (Table 2).


 

 

Table 2: Participants’ attitude towards artificial intelligence in health care settings

Attitude statement

Disagree

Neutral

Agree

No.

%

No.

%

No.

%

AI has useful applications in the medical field

30

8.4

71

19.9

256

71.7

AI will improve the public image of physicians

54

15.1

95

26.6

209

58.5

AI could provide more time for physicians for training and research

58

16.2

116

32.5

183

51.3

You have good familiarity with AI

93

26.1

99

27.7

165

46.2

You would always use AI when making medical decisions in the future

139

38.9

121

33.9

97

27.2

The diagnostic ability of AI is superior to the clinical experience of a human doctor

245

68.6

70

19.6

42

11.8

AI could replace your job completely

306

85.7

33

9.2

18

5.0

 

 

 


Advantage of AI:

 

When physicians were asked about the advantage of AI, 71.7% agreed that it speeds up processes in health care, 59.1% agreed that it helps reducing medical errors, 51.3% agreed that it can deliver vast amounts of clinically relevant high-quality data in real time. However, only 43.4% and 20.4% supported the idea that AI has no emotional exhaustion neither space-time constraint respectively (Fig 1).


 

Figure (1): Participants’ opinion regarding the advantage of artificial intelligence

 

 

 


Diagnostic judgment of AI:

 

When physicians were asked if their medical judgment and an artificial intelligence’s judgments differ, which will they follow, 85.2% chose doctor’ opinion, and only 3.9% chose AI opinion.

 

Training on application of AI:

 

Regarding the training on applications of AI, 40.9% preferred practical individual supervised training, 39.5% preferred lectures and workshops. Lower percentage of participating physicians chose videos and tutorial online training (9.8%), and information and instruction booklets (7.6%) (Table 3).

 

Fields of applications of AI in medicine:

 

When physicians were asked about the most useful expected application of AI in the field of medicine, came at the top of the list biopharmaceutical research and development (30.5%) followed by diagnosis (24.1%) and medical assistance in underserved areas (18.5%). Less frequent application were mentioned by physicians as direct treatment (including surgery) (11.2%), making treatment decisions (9.0%) and development of social insurance program (6.7%) (Table 3)

 

Commercialization of AI:

 

Physicians’ opinion regarding the first sector of health care that will commercialize AI were university hospitals (33.1%) and governmental PHC centers (32.5%) followed by specialized clinics (spine, knee, obstetrics and gynecology, etc) (22.7%) and primary care in private clinics (11.7%) (Table 3).


 

 

Table 3: Participants’ opinion regarding artificial intelligence judgment, training and applications in medicine

Statement

No.

%

If your medical judgment and an artificial intelligence’s judgment differ, which will you follow?

 

 

-       Doctor’s opinion

304

85.2

-       Artificial intelligence’s opinion

14

3.9

-       Patients’ choice

39

10.9

What is the most appropriate method of training on applications of artificial intelligence?  

 

 

-       Lectures and workshops

141

39.5

-       Practical individual supervised training

146

40.9

-       Information and instruction booklets

27

7.6

-       Videos and tutorial online training

36

9.8

-       None of the above

8

2.2

Expected Applications in Medicine

In which field of medicine do you think artificial intelligence will be most useful?  

 

 

-       Biopharmaceutical research and development

109

30.5

-       Making a diagnosis

86

24.1

-       Providing medical assistance in underserved areas

66

18.5

-       Direct treatment (including surgery)

40

11.2

-       Making treatment decisions

32

9.0

-       Development of social insurance program

24

6.7

Which sector of health care do you think will be the first to commercialize artificial intelligence?  

 

 

-       Governmental primary care such as public health centers

116

32.5

-       Primary care in private clinics

42

11.8

-       Specialized clinics (spine, knee, obstetrics and gynecology, etc)

81

22.7

-       University hospitals

118

33.1

Total

357

100.0

 

 


Possible risk of AI application in medicine:

 

When physicians were asked about the possible risks of application of AI in medicine, the most frequent answer was inflexibility to be applied to every patient (65.3%). This was followed by inability to be used to provide opinions in unpredicted situations due to inadequate information (59.1%), inability to sympathize and consider the emotional well-being of the patient (55.2%) and difficulty of application to controversial subjects (48.7%). At the bottom of the list came its development by a specialist with little clinical experience in medical practice (24.6%) (Figure 2).


 

 

Figure (2): Participants’ opinion regarding possible risk of AI application in medicine

 

 

Table 4:  Participants’ opinion regarding the likelihood of AI replacing physicians on certain tasks

Task

Unlikely

Likely

No

%

No

%

Provide documentation (e.g., update medical records) about patients.

73

20.4

284

79.5

Analyze patient information to establish prognoses

159

44.5

198

55.5

Evaluate when to refer patients to other health professionals

193

54.1

164

45.9

Analyze patient information to reach diagnoses

200

56.0

157

44.0

Formulate personalized treatment plans for patients.

225

63.1

132

36.9

Provide empathetic care to patients

287

80.4

70

19.6

 


 

 

Almost four fifths of participants (79.5%) believed that it is likely that technology will be able to fully replace human physicians when it comes to providing documentation about patient. More than half of them (59.8%) believed that this would occur in the next ten years. As regards analyzing patient information to establish prognoses, 55.5% of participants stated that it is likely that AI can replace human physicians. Among them 40.9% believed that this will occur within 10 years. Less than half of the participating physicians thought that AI can replace human physicians regarding evaluation of time of referral of patients to other health professionals (45.9%) and 39.6% of them believed that it would happened within 10 years. Nearly the same proportion of physicians thought that it is likely for AI to replace physicians regarding analysis of patient information to reach diagnoses (44.0%), and 31.2 of them thought that could be happen within 10 years.

The least proportions were encountered regarding formulating personalized treatment plans for patients (36.9%) and providing empathetic care to patients (19.6%) Within those physicians 34.8% and 38.5% thought that replacement could occur within 10 years respectively. (Table 4, 5).


 

Table 5: Timescale (years) given by participants who believed that replacement of a given task by AI is likely.

Task

Years

1 - 10

11 - 25

>25

No.

%

No.

%

No.

%

Provide documentation (e.g., update medical records) about patients. (n = 284)

167

59.8

64

22.5

53

18.7

Analyze patient information to establish prognoses (n = 198)

81

40.9

49

24.7

68

34.4

Evaluate when to refer patients to other health professionals (n = 164)

65

39.6

46

28.0

53

32.3

Analyze patient information to reach diagnoses (n = 157)

49

31.2

50

31.8

58

37.0

Formulate personalized treatment plans for patients. (n = 132)

46

34.8

46

34.1

41

31.1

Provide empathetic care to patients (n = 70)

27

38.5

24

34.3

19

27.1

 

 


 

DISCUSSION:

 

While recent studies have shown that AI holds the potential to improve performance with regard to health care, the pathway toward widespread adoption of new technologies is complex and physician’ perceptions are likely to play a significant role in the pace of technology adoption in clinical practice. The results of the present study suggest that the perception of AI by PHC doctors was low whereas less than half of participating physicians (46.2%) considered themselves familiar with AI. However, 71.7% believed that AI had useful applications in the health care field. These results are comparable with a similar study that was conducted in Korea where the authors reported that familiarity with AI was very low, only 6.0% believed that they were familiar with AI in a good degree and many participants answered that AI is useful in health care and medical field (73.4%). (Oh, 2019)

As regards the advantage of artificial intelligence, 71.7% agreed that it speeds up processes in health care, 59.1% agreed that it helps reducing medical errors, 51.3% agreed that it can support very fast plenty of relevant high-quality information. However, only 43.4% and 20.4% support the idea that AI is protective against emotional stress and space-time constraint respectively. In another study, the physicians agreed that the advantages of AI include supporting good relevant data in short time (62.3%), rapid acceleration of processes in health care (19.1%), and lower down the number of medical errors (9.6%). (Oh et al., 2019)

Only 11.8% stated that the diagnostic precision of AI is superior to the clinical opinion of a human doctor. A higher proportion was reported in a previous study whereas 44% of participants agreed that “AI is superior to a doctor’s experience”. (Oh et al., 2019) On the other hand, 27.2% of physicians agreed that, in the future, they will always use AI for their final medical decisions, a figure that was relatively low when compared with similar study as 42% of participants stated that to reach a medical decision AI should be used. (Oh et al., 2019). When physicians were asked that in case of difference between AI judgment and doctor’s opinion during medical practice, 85.2% chose doctor’ opinion, and only 3.9% chose AI opinion. This goes with a previous study as 79% of participants would follow the doctor’s opinion. (Oh et al., 2019)

Only 5.0% of participants in the present study agreed that AI could fully replace their jobs. The proportion of physicians who believed that AI would replace physicians showed variations in literatures. It may be as high as in a Korean study where 35.4% of physicians reported that AI could replace a doctor, (Oh et al., 2019) or very low as in Doraiswamy et al (2020) study who reported that though half of participants believed that AI would substantially change physicians’ jobs, only 3.8% felt that it would make their jobs no longer in use. In accordance with our results regarding difficulties in replacing doctors. Krittanawong (2018) stated that AI cannot replace doctors as it cannot conduct specific conversation with patients to reach a point of trust. Also, in spite of the valuable data given by AI, still interpretation of data, medical examination and history should be conducted by physicians. (Krittanawong, 2018; Inkster et al., 2018)

In another study, only 3.8% of physicians believed that future AI would make their jobs out of date and the majority (83%) of respondents were sure that AI would never have the ability to provide empathic care as well as or better than the average. Another 47% predicted that their jobs would be moderately changed by AI over the next 25 years. (Doraiswamy et al., 2020)

When physicians were asked about the most useful expected application of AI in the field of medicine, came at the top of the list biopharmaceutical research and development (30.5%) followed by diagnosis (24.1%) and medical assistance in areas with less services. (18.5%) Less frequent applications were mentioned by physicians as direct treatment (including surgery) (11.2%), and making treatment decisions. These results are relatively lower than that obtained in other study whereas respondents felt that in medical field, AI would be most useful in the area of diagnosis (83.4%) as well as formulation of the treatment plan (53.8%). (Oh et al., 2019) According to Doraiswamy et al. (2020), most participants considered it unlikely that AI would fully replace doctors in the field of psychology particularly getting medical history during interview (58%), determination of the department to which the patient should be referred (55%), planning of personalized treatment plans for each patient (53%). On the other side, most of them (83%) believed that the task of documentation will be likely replaced by AI.

In this study physicians believed that commercialization of AI will be through university hospitals (33.15) followed by governmental primary centers (32.5%) which goes in accordance this Oh et al. (2019) study who reported that most participants (66.2%) thought also university hospital would be the first.

In the present study, physicians ‘opinion regarding the potential risks of AI when applied in medicine revealed that the most frequent answer was inflexibility to be applied to every patient (65.3%). This was followed by inability to give definite decision in case of inadequate information (59.1%), neglecting patient’ emotional well-being (55.2%) and difficulty of application to controversial subjects (48.7%). At the bottom of the list came when AI is developed by a person with little clinical experience in medicine or health (24.6%). Comparable with our results, Oh et al. (2019) reported that the possible risks with AI are that AI is deficient in providing a decision in case of inadequate information (29.3%) and that it would not be applicable to every patient (34.1%). Precision medicine is “an emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person” With this approach physicians are allowed to decide treatment plan and prevention strategies for patients. (James et al., 2017) The mental status examination, evaluation of dangerous behavior and formulation of a personalized treatment plan, were also felt to be tasks that a future AI technology would be unlikely to perform as well. This could be as doctors may be overvaluing their skills and/or underestimating the rapid pace of progress in intelligent technologies. (Topol, 2019)

Blease et al. (2018) reported that the most GPs were not sure that AI technology could perform physicians’ tasks as well as or better than humans. However, they think that technology would be able to replace physicians administrative work related to documentation.

The present study revealed that the proportion of physicians believed that it is likely that technology will be able to replace human physicians were 79.5% regarding providing documentation about patient, 55.5% regarding analyzing patient information to establish prognoses, just less than half of participants regards evaluation of time of referral of patients to other health professionals, analysis of patient information to reach diagnoses. The least proportions were encountered regarding formulating personalized treatment plans for patients (36.9%) and providing empathetic care to patients (19.6%) Most of participants in the present study who believed that AI is likely to replace physicians in certain tasks, indicated that this would happen within ten years. Our results go in accordance with a previous study conducted by Blease et al., (2018), who found that most GPs believed it unlikely that technology will ever be able to fully replace physicians when it comes to diagnosing patients (68%), referring patients to other specialists (61%), formulating personalized treatment plans (61%), and delivering empathic care (94%). On the other hand, 53% considered it likely that technology will be fully capable of replacing physicians in prognostic process and 80% believed it likely to fully replace humans to undertake documentation. Blease et al., (2018) In another previous study, most physicians expected that AI would be helpful with diagnoses and in planning treatment by providing the latest clinically relevant data. However, doctors were skeptical that future technologies could perform most tasks as well as or better than human doctors.. This is may be due to the fact that doctors may not feel confident in their knowledge of AI and may find it difficult to separate marketing hype from ground truth. (Doraiswamy et al., 2020)

According to some AI experts, machine level intelligence will be capable of replacing the basic functions of medical professionals by 2050. (Walsh, 2017) Some physicians working in biomedical informatics and related fields predicted that developments in AI are poised to revolutionize the delivery of health care, (Hinton, 2018; Naylor, 2018; Mandl and Bourgeois, 2017) with some suggesting advancements may eventually obviate the need for physicians altogether. (Obermeyer and Lee, 2017) Physicians may have multiple, possibly conflicting, beliefs about the impact of technology in primary care and research is needed to explore these issues further. (Blease et al., (2018)

Some technologists argue that advancements in AI may one day obviate the need of physicians altogether. (Darcy et al., 2017; Naylor, 2018) Others forecasting that the role of doctors can never be fully replaced, and that the future of medicine will likely become a team between humans and machines. (Verghese et al., 2018; Obermeyer and Lee, 2017) This view is also shared by authors of a more recent working paper from the Office of Economic Cooperation and Development which looked at skills data from over 30 countries. (Nedelkoska and Quintini, 2018) Our results together with that in literatures provide powerful insights into how AI could be optimally deployed to work with physicians, rather than replace them, to enhance health care.

Our study had a limitation regarding generalization of the results as our sample of PHC physicians may not be representative of all Kuwaiti physicians and participation was voluntary. Another limitation is the self-reporting nature of the study with the risk of recall bias.

 

 

References:

 

Bakkar N, Kovalik T, Lorenzini I, Spangler S, Lacoste A, Sponaugle K, Ferrante P, Argentinis F, Sattler R, Bowser R. (2018). Artificial intelligence in neurodegenerative disease research: use of IBM Watson to identify additional RNA-binding proteins altered in amyotrophic lateral sclerosis. Acta Neuropathol; 135(2) :227-47.

Blease C, Bernstein MH, Gaab J, Kaptchuk YJ, Kossowsky J, Mandl KD, Davis RB. DesRoches CM. (2018). Computerization and the future of primary care: A survey of general practitioners in the UK. PLOS ONE | https://doi.org/10.1371/journal.pone.0207418

Darcy AM, Louie AK, Roberts LW. (2016). Machine learning and the profession of medicine. JAMA, 315(6), 551–2.

Doraiswamy M, Blease C, Bodner K. (2020). Artificial intelligence and the future of psychiatry: Insights from a global physician survey (2020). Artificial Intelligence in Medicine; 102. https://doi.org/10.1016/j.artmed.2019.101753

Esteva A, Robicquet A, Ramsundar B, Kuleshov V, DePristo M, Chou K, Cui C, Corrado G, Thrun S, Dean J. (2019). A guide to deep learning in healthcare. Nat Med;25(1): 4-29.

Hinton G.  (2018). Deep learning–A technology with the potential to transform health care. Jama. ; 320(11): 1101–2.

Inkster B, Sarda S, Subramanian V. (2018). An empathy-driven, conversational artificial intelligence agent (Wysa) for digital mental well-being: real-world data evaluation mixed-methods study. JMIR Mhealth Uhealth;6(11):e12106. [doi: 10.2196/12106]

James R. National Institute of on Minority Health and Health Disparities. (2017). National Institutes of Health Precision medicine initiative URL: https://www.nimhd.nih.gov/about/legislative-info/clips/pmi.html [accessed 2019-02-05] [WebCite Cache ID 75yNuNGg5]

Jha S (2016). Will computer replace radiologist? https://www.medscape.com/viewarticle/863127

Kent J ( (2018). Providers Embrace Predictive Analytics for Clinical, Financial Benefits". Health IT Analytics. https://healthitanalytics.com/news/providers-embrace-predictive-analytics-for-clinical-financial-benefits

Krittanawong C. (2018).. The rise of artificial intelligence and the uncertain future for physicians. Eur J Intern Med; 48: e13-e14. [doi: 10.1016/j.ejim.2017.06.017] [Medline: 28651747]

Legg S, Hutter M. (2007). A Collection of Definitions of Intelligence (Technical report). IDSIA. https://arXiv:0706.3639. Bibcode:2007arXiv0706.3639L. 07-07.

Mandl KD, Bourgeois FT. (2017). The Evolution of patient diagnosis: From art to digital data-driven science. Jama; 318(19): 1859–60

Naylor CD. (2018). On the prospects for a deep learning health care system. JAMA, 320(11): 1099-100.

Nedelkoska L, Quintini G. (2018). Automation, skills use and training. OECD Social, Employment and Migration Working Papers, 2018. OECD Publishing, Paris, 202, 1- 124.

Obermeyer Z., Lee T. H. (2017). Lost in thought – the limits of the human mind and the future of medicine. New England Journal of Medicine, 377(13), 1209-1211.

Oh S, Kim JH, Choi S, Lee HJ, Hong J, Kwon SH. (2019). Physician Confidence in Artificial Intelligence: An Online Mobile Survey. J Med Internet Res; 21(3): e12422) doi:10.2196/12422

Russell S, Norvig P. (2020). Artificial Intelligence: A Modern Approach (4th ed.). http://aima.cs.berkeley.edu/.

Sarwar S, Dent A, Faust K, Richer M, Djuric U, Van Ommeren R, Diamandis P. (2019). Physician perspectives on integration of artificial intelligence into diagnostic pathology. NPG Digit Med; 2: 28 https://www.nature.com/articles/s41746-019-0106-0

Shameer K, Johnson KW, Glicksberg BS, Dudley JT, Sengupta PP. (2018). Machine learning in cardiovascular medicine: are we there yet? Heart; 104(14): 1156-64.

Topol, E. (2019). Deep medicine: how artificial intelligence can make healthcare human again. Basic Books. Hachette UK

Trivedi H, Mesterhazy J, Laguna B, Vu T, Sohn JH. (2018). Automatic determination of the need for intravenous contrast in musculoskeletal MRI examinations using IBM Watson's natural language processing algorithm. J Digit Imaging;31(2):245-251.

Verghese A, Shah NH, Harrington RA. (2018). What this computer needs is a Physician: Humanism and AI. JAMA, 2, 319-21.

Walsh T. (2017). Android dreams: the past, present, and future of artificial intelligence. UK: Hurst.

 


 

 

Cite this Article: El-Shazly M; Al-Majed M; Al-Omar A; Abdalla M; Pereira M (2021). Primary health care physicians’ perception of artificial intelligence in health care. Greener Journal of Medical Sciences, 11(1): 37-45.