Digital Technology: Key considerations for the use of artificial intelligence in healthcare and clinical research

Abstract

Interest in artificial intelligence (AI) has grown exponentially in recent years, attracting sensational headlines and speculation. While there is considerable potential for AI to augment clinical practice, there remain numerous practical implications that must be considered when exploring AI solutions. These range from ethical concerns about algorithmic bias to legislative concerns in an uncertain regulatory environment. In the absence of established protocols and examples of best practice, there is a growing need for clear guidance both for innovators and early adopters. Broadly, there are three stages to the innovation process: invention, development and implementation. In this paper, we present key considerations for innovators at each stage and offer suggestions along the AI development pipeline, from bench to bedside.

Research on artificial intelligence in shoulder and elbow surgery is increasing

“…With increasing promise of AI tools being able to assist in lowering costs and improving quality of care, researchers, physicians, and patients are becoming more aware of and interested in AI. 2 , 10 , 12 , 15 , 23 , 24 , 26 , 31 Second, the optimism and success seen in AI research and applications in other orthopedic subspecialties such as hip and knee arthroplasty and spine surgery has likely spurred more interest in shoulder and elbow AI research. 33 , 34 , 39 For example, Karnuta et al trained and externally validated a DL model to classify one of eight femoral-sided total hip arthroplasty implants directly from anteroposterior plain radiographs, which performed with a mean speed of 0.02 seconds per image, area under the curve (AUC) of 0.99, and accuracy of nearly 98%.…”

Implementing artificial intelligence in Canadian primary care: Barriers and strategies identified through a national deliberative dialogue

“…Overall, Sittig and Singh’s sociotechnological framework is advantageous to consider both the social and technical perspectives of technology implementation projects, but based on our deliberative dialogue series, we believe that new dimensions ( e . g ., trust, bias, and equity) would be required to translate this model to be applicable to AI implementation projects in the primary care space now and in the future.As mentioned in the introduction, the recent review conducted by Lovejoy et al divides the AI innovation process into three stages (invention, development, and implementation) and outlines key considerations for innovators during each phase [ 13 ]. The considerations of the third stage, implementation, include generalizability, regulation, and deployment.…”

“…Specifically with respect to the implementation stage, they highlight three main barriers of focus (generalizability, regulation, and deployment). Generalizability refers to the need for diversity within the training set, regulation refers to the need for ongoing monitoring of an algorithm or tool throughout its lifetime to ensure continued safety, and deployment refers to the need to minimize disruption to workflow and account for interoperability between healthcare technologies [ 13 ]. Although this review is not specific to the primary care space, it provides insights into barriers that are commonly encountered in the healthcare AI space, all of which will help inform our study.…”

Role of artificial intelligence and machine learning in haematology

“…There has been interest in using AI to personalise treatment plans and truly embrace a holistic approach towards patient care 12. Traditional pharmacological methods are designed to treat the non-existent ‘average patient’, despite recognition that there is potentially substantial heterogeneity in how a population may respond to a treatment 13. One such example is warfarin anticoagulation therapy, which relies on manual interpretation of international normalized ratio (INR) results and for which the patient response to treatment can be affected by many clinical factors.…”