A Perspective from a Case Conference on Comparing the Diagnostic Process: Human Diagnostic Thinking vs. Artificial Intelligence (AI) Decision Support Tools.

Taku Harada,Yasuhiro Suyama,Taro Shimizu,Takatoshi Nei,Yuki Kaji,Satoshi Watanuki,Chintaro Kosaka,Hidefumi Shimizu,Tomohiro Matsumoto

doi:10.3390/ijerph17176110

Abstract

Artificial intelligence (AI) has made great contributions to the healthcare industry. However, its effect on medical diagnosis has not been well explored. Here, we examined a trial comparing the thinking process between a computer and a master in diagnosis at a clinical conference in Japan, with a focus on general diagnosis. Consequently, not only was AI unable to exhibit its thinking process, it also failed to include the final diagnosis. The following issues were highlighted: (1) input information to AI could not be weighted in order of importance for diagnosis; (2) AI could not deal with comorbidities (see Hickam’s dictum); (3) AI was unable to consider the timeline of the illness (depending on the tool); (4) AI was unable to consider patient context; (5) AI could not obtain input information by themselves. This comparison of the thinking process uncovered a future perspective on the use of diagnostic support tools.

Highlights

Artificial intelligence (AI) has made great contributions to the healthcare industry
Based on the above information, the diagnostician focused on central adrenal insufficiency as the most likely diagnosis, such as in the pituitary gland and hypothalamic region, followed by paraneoplastic syndrome, and atypical Guillain-Barré syndrome, and considered these as the top three differential diagnoses
“dysphagia”, and Isabel© presented endocrine disorders that included adrenal insufficiency. This fact illustrates both the magnificent comprehensiveness of Isabel©’s differential diagnosis and the danger of giving misleading information, depending on the keywords that humans enter. This discussion regarding the use of decision support tools in the context of the comparison with human diagnostic thinking underscored the following issues: (1) input data could not be weighted in order of importance for diagnosis; (2) AI could not deal with comorbidities; (3) depending on the tool, AI may be unable to consider the timeline of the illness; (4) AI was unable to consider patient contexts such as social background and other information not explained by keywords; and more importantly, (5) AI could not obtain input information by themselves

Summary

Introduction

Artificial intelligence (AI) has made great contributions to the healthcare industry. Among the many decision support tools currently available, Isabel© has been identified as one that provides excellent outcomes, with diagnoses taking approximately only 1 min per patient and resulting in an improved diagnostic accuracy and a reduced error rate [7,8]. Physicians have better diagnostic accuracy than AI, but for rare diseases, AI performs better than humans [3,9].

Results

Conclusion