Developing, Purchasing, Implementing and Monitoring AI Tools in Radiology: Practical Considerations. A Multi-Society Statement From the ACR, CAR, ESR, RANZCR & RSNA.

Artificial Intelligence (AI) carries the potential for unprecedented disruption in radiology, with possible positive and negative consequences. The integration of AI in radiology holds the potential to revolutionize healthcare practices by advancing diagnosis, quantification, and management of multiple medical conditions. Nevertheless, the ever­growing availability of AI tools in radiology highlights an increasing need to critically evaluate claims for its utility and to differentiate safe product offerings from potentially harmful, or fundamentally unhelpful ones. This multi­society paper, presenting the views of Radiology Societies in the USA, Canada, Europe, Australia, and New Zealand, defines the potential practical problems and ethical issues surrounding the incorporation of AI into radiological practice. In addition to delineating the main points of concern that developers, regulators, and purchasers of AI tools should consider prior to their introduction into clinical practice, this statement also suggests methods to monitor their stability and safety in clinical use, and their suitability for possible autonomous function. This statement is intended to serve as a useful summary of the practical issues which should be considered by all parties involved in the development of radiology AI resources, and their implementation as clinical tools.

Radiology in the Era of Value-based Healthcare: A Multi-Society Expert Statement from the ACR, CAR, ESR, IS3R, RANZCR, and RSNA.

Background The Value-Based Healthcare (VBH) concept is designed to improve individual healthcare outcomes without increasing expenditure, and is increasingly being used to determine resourcing of and reimbursement for medical services. Radiology is a major contributor to patient and societal healthcare at many levels. Despite this, some VBH models do not acknowledge radiology's central role; this may have future negative consequences for resource allocation. Methods, findings and interpretation This multi-society paper, representing the views of Radiology Societies in Europe, the USA, Canada, Australia, and New Zealand, describes the place of radiology in VBH models and the health-care value contributions of radiology. Potential steps to objectify and quantify the value contributed by radiology to healthcare are outlined. Published under a CC BY 4.0 license.

Marfan syndrome resulting from a rare pathogenic FBN1 variant, ascertained through a proband with IgG4-related arteriopathy.

A 57-year-old man with a family history of aortic aneurysm was found, during assessment of unexplained fever, to have an infrarenal aortic aneurysm requiring immediate repair. Dilatation of popliteal and iliac arteries was also present. Progressive aortic root dilatation with aortic regurgitation was documented from 70 years leading to valve-sparing aortic root replacement at 77 years, at which time genetic studies identified a likely pathogenic FBN1 missense variant c.6916C > T (p.Arg2306Cys) in exon 56. The proband's lenses were normally positioned and the Marfan syndrome (MFS) systemic score was 0/20. Cascade genetic testing identified 15 other family members with the FBN1 variant, several of whom had unsuspected aortic root dilatation; none had ectopia lentis or MFS systemic score ≥ 7. Segregation analysis resulted in reclassification of the FBN1 variant as pathogenic. The combination of thoracic aortic aneurysm and dissection (TAAD) and a pathogenic FBN1 variant in multiple family members allowed a diagnosis of MFS using the revised Ghent criteria. At 82 years, the proband's presenting abdominal aortic aneurysm was diagnosed retrospectively to have resulted from IgG4-related inflammatory aortopathy.

Radiology in the Era of Value-Based Healthcare: A Multi-Society Expert Statement From the ACR, CAR, ESR, IS3R, RANZCR, and RSNA.

BACKGROUND: The Value-Based Healthcare (VBH) concept is designed to improve individual healthcare outcomes without increasing expenditure, and is increasingly being used to determine resourcing of and reimbursement for medical services. Radiology is a major contributor to patient and societal healthcare at many levels. Despite this, some VBH models do not acknowledge radiology's central role; this may have future negative consequences for resource allocation. METHODS, FINDINGS AND INTERPRETATION: This multi-society paper, representing the views of Radiology Societies in Europe, the USA, Canada, Australia, and New Zealand, describes the place of radiology in VBH models and the health-care value contributions of radiology. Potential steps to objectify and quantify the value contributed by radiology to healthcare are outlined.

Demonstrating the use of population level data to investigate trends in the rate, radiation dose and cost of Computed Tomography across clinical groups: Are there any areas of concern?

INTRODUCTION: Increases in computed tomography (CT) use may not always reflect clinical need or improve outcomes. This study aimed to demonstrate how population level data can be used to identify variations in care between patient groups, by analysing system-level changes in CT use around the diagnosis of new conditions. METHODS: Retrospective repeated cross-sectional observational study using West Australian linked administrative records, including 504,723 adults diagnosed with different conditions in 2006, 2012 and 2015. For 90 days pre/post diagnosis, CT use (any and 2+ scans), effective dose (mSv), lifetime attributable risk (LAR) of cancer incidence and mortality from CT, and costs were assessed. RESULTS: CT use increased from 209.4 per 1000 new diagnoses in 2006 to 258.0 in 2015; increases were observed for all conditions except neoplasms. Healthcare system costs increased for all conditions but neoplasms and mental disorders. Effective dose increased substantially for respiratory (+2.5 mSv, +23.1%, P < 0.001) and circulatory conditions (+2.1 mSv, +15.4%, P < 0.001). The LAR of cancer incidence and mortality from CT increased for endocrine (incidence +23.4%, mortality +18.0%) and respiratory disorders (+21.7%, +23.3%). Mortality LAR increased for circulatory (+12.1%) and nervous system (+11.0%) disorders. The LAR of cancer incidence and mortality reduced for musculoskeletal system disorders, despite an increase in repeated CT in this group. CONCLUSIONS: Use and costs increased for most conditions except neoplasms and mental and behavioural disorders. More strategic CT use may have occurred in musculoskeletal conditions, while use and radiation burden increased for respiratory, circulatory and nervous system conditions. Using this high-level approach we flag areas requiring deeper investigation into appropriateness and value of care.

Developing, purchasing, implementing and monitoring AI tools in radiology: practical considerations. A multi-society statement from the ACR, CAR, ESR, RANZCR & RSNA.

Artificial Intelligence (AI) carries the potential for unprecedented disruption in radiology, with possible positive and negative consequences. The integration of AI in radiology holds the potential to revolutionize healthcare practices by advancing diagnosis, quantification, and management of multiple medical conditions. Nevertheless, the ever-growing availability of AI tools in radiology highlights an increasing need to critically evaluate claims for its utility and to differentiate safe product offerings from potentially harmful, or fundamentally unhelpful ones.This multi-society paper, presenting the views of Radiology Societies in the USA, Canada, Europe, Australia, and New Zealand, defines the potential practical problems and ethical issues surrounding the incorporation of AI into radiological practice. In addition to delineating the main points of concern that developers, regulators, and purchasers of AI tools should consider prior to their introduction into clinical practice, this statement also suggests methods to monitor their stability and safety in clinical use, and their suitability for possible autonomous function. This statement is intended to serve as a useful summary of the practical issues which should be considered by all parties involved in the development of radiology AI resources, and their implementation as clinical tools.Key points • The incorporation of artificial intelligence (AI) in radiological practice demands increased monitoring of its utility and safety.• Cooperation between developers, clinicians, and regulators will allow all involved to address ethical issues and monitor AI performance.• AI can fulfil its promise to advance patient well-being if all steps from development to integration in healthcare are rigorously evaluated.

Developing, Purchasing, Implementing and Monitoring AI Tools in Radiology: Practical Considerations. A Multi-Society Statement from the ACR, CAR, ESR, RANZCR and RSNA.

Artificial Intelligence (AI) carries the potential for unprecedented disruption in radiology, with possible positive and negative consequences. The integration of AI in radiology holds the potential to revolutionize healthcare practices by advancing diagnosis, quantification, and management of multiple medical conditions. Nevertheless, the ever-growing availability of AI tools in radiology highlights an increasing need to critically evaluate claims for its utility and to differentiate safe product offerings from potentially harmful, or fundamentally unhelpful ones. This multi-society paper, presenting the views of Radiology Societies in the USA, Canada, Europe, Australia, and New Zealand, defines the potential practical problems and ethical issues surrounding the incorporation of AI into radiological practice. In addition to delineating the main points of concern that developers, regulators, and purchasers of AI tools should consider prior to their introduction into clinical practice, this statement also suggests methods to monitor their stability and safety in clinical use, and their suitability for possible autonomous function. This statement is intended to serve as a useful summary of the practical issues which should be considered by all parties involved in the development of radiology AI resources, and their implementation as clinical tools. This article is simultaneously published in Insights into Imaging (DOI 10.1186/s13244-023-01541-3), Journal of Medical Imaging and Radiation Oncology (DOI 10.1111/1754-9485.13612), Canadian Association of Radiologists Journal (DOI 10.1177/08465371231222229), Journal of the American College of Radiology (DOI 10.1016/j.jacr.2023.12.005), and Radiology: Artificial Intelligence (DOI 10.1148/ryai.230513). Keywords: Artificial Intelligence, Radiology, Automation, Machine Learning Published under a CC BY 4.0 license. ©The Author(s) 2024. Editor's Note: The RSNA Board of Directors has endorsed this article. It has not undergone review or editing by this journal.

Developing, purchasing, implementing and monitoring AI tools in radiology: Practical considerations. A multi-society statement from the ACR, CAR, ESR, RANZCR & RSNA.

Artificial Intelligence (AI) carries the potential for unprecedented disruption in radiology, with possible positive and negative consequences. The integration of AI in radiology holds the potential to revolutionize healthcare practices by advancing diagnosis, quantification, and management of multiple medical conditions. Nevertheless, the ever-growing availability of AI tools in radiology highlights an increasing need to critically evaluate claims for its utility and to differentiate safe product offerings from potentially harmful, or fundamentally unhelpful ones. This multi-society paper, presenting the views of Radiology Societies in the USA, Canada, Europe, Australia, and New Zealand, defines the potential practical problems and ethical issues surrounding the incorporation of AI into radiological practice. In addition to delineating the main points of concern that developers, regulators, and purchasers of AI tools should consider prior to their introduction into clinical practice, this statement also suggests methods to monitor their stability and safety in clinical use, and their suitability for possible autonomous function. This statement is intended to serve as a useful summary of the practical issues which should be considered by all parties involved in the development of radiology AI resources, and their implementation as clinical tools.

Developing, Purchasing, Implementing and Monitoring AI Tools in Radiology: Practical Considerations. A Multi-Society Statement From the ACR, CAR, ESR, RANZCR & RSNA.

Artificial intelligence (AI) carries the potential for unprecedented disruption in radiology, with possible positive and negative consequences. The integration of AI in radiology holds the potential to revolutionize healthcare practices by advancing diagnosis, quantification, and management of multiple medical conditions. Nevertheless, the ever-growing availability of AI tools in radiology highlights an increasing need to critically evaluate claims for its utility and to differentiate safe product offerings from potentially harmful, or fundamentally unhelpful ones. This multi-society paper, presenting the views of Radiology Societies in the USA, Canada, Europe, Australia, and New Zealand, defines the potential practical problems and ethical issues surrounding the incorporation of AI into radiological practice. In addition to delineating the main points of concern that developers, regulators, and purchasers of AI tools should consider prior to their introduction into clinical practice, this statement also suggests methods to monitor their stability and safety in clinical use, and their suitability for possible autonomous function. This statement is intended to serve as a useful summary of the practical issues which should be considered by all parties involved in the development of radiology AI resources, and their implementation as clinical tools. KEY POINTS.

Patterns of computed tomography utilisation in injury management: latent classes approach using linked administrative data in Western Australia.

PURPOSE: Whilst computed tomography (CT) imaging has been a vital component of injury management, its increasing use has raised concern regarding ionising radiation exposure. This study aims to identify latent classes (underlying patterns) of CT use over a 3-year period following the incidence of injury and factors predicting the observed patterns. METHOD: A retrospective observational cohort study was conducted in 21,544 individuals aged 18 + years presenting to emergency departments (ED) of four tertiary public hospitals with new injury in Western Australia. Mixture modelling approach was used to identify latent classes of CT use over a 3-year period post injury. RESULTS: Amongst injured people with at least one CT scan, three latent classes of CT use were identified including a: temporarily high CT use (46.4%); consistently high CT use (2.6%); and low CT use class (51.1%). Being 65 + years or older, having 3 + comorbidities, history with 3 + hospitalisations and history of CT use before injury were associated with consistently high use of CT. Injury to the head, neck, thorax or abdomen, being admitted to hospital after the injury and arriving to ED by ambulance were predictors for the temporarily high use class. Living in areas of higher socio-economic disadvantage was a unique factor associated with the low CT use class. CONCLUSIONS: Instead of assuming a single pattern of CT use for all patients with injury, the advanced latent class modelling approach has provided more nuanced understanding of the underlying patterns of CT use that may be useful for developing targeted interventions.

Artificial Intelligence Education for the Health Workforce: Expert Survey of Approaches and Needs.

BACKGROUND: The preparation of the current and future health workforce for the possibility of using artificial intelligence (AI) in health care is a growing concern as AI applications emerge in various care settings and specializations. At present, there is no obvious consensus among educators about what needs to be learned or how this learning may be supported or assessed. OBJECTIVE: Our study aims to explore health care education experts' ideas and plans for preparing the health workforce to work with AI and identify critical gaps in curriculum and educational resources across a national health care system. METHODS: A survey canvassed expert views on AI education for the health workforce in terms of educational strategies, subject matter priorities, meaningful learning activities, desired attitudes, and skills. A total of 39 senior people from different health workforce subgroups across Australia provided ratings and free-text responses in late 2020. RESULTS: The responses highlighted the importance of education on ethical implications, suitability of large data sets for use in AI clinical applications, principles of machine learning, and specific diagnosis and treatment applications of AI as well as alterations to cognitive load during clinical work and the interaction between humans and machines in clinical settings. Respondents also outlined barriers to implementation, such as lack of governance structures and processes, resource constraints, and cultural adjustment. CONCLUSIONS: Further work around the world of the kind reported in this survey can assist educators and education authorities who are responsible for preparing the health workforce to minimize the risks and realize the benefits of implementing AI in health care.

Association of regulatory body actions and subsequent media coverage with use of services in a fee-for-service system: a longitudinal cohort study of CT scanning in Australia.

OBJECTIVE: The professional service review (PSR) is an Australian Government agency aiming to reduce inappropriate practices funded via Medicare, Australia's public insurer. Our objective was to examine changes in CT following the 2008-2009 PSR annual report, which noted excessive CT use. DESIGN: Interrupted time series analysis examined trends in CT use following the 2008-2009 PSR report, estimating both change in the immediate rate of CT and the slope of the trend in usage postintervention. SETTING: Medicare-funded imaging (most out-of-hospital imaging) in Australia. PARTICIPANTS: Patients receiving Medicare-funded CT and other imaging. INTERVENTION: The 2008-2009 PSR report highlighted concerns regarding excessive CT use. Two providers were financially penalised for CT overuse with these cases detailed in the PSR report and highlighted in an associated Report to the Professions, distributed to 50 000 providers. Media articles on radiation risks followed. OUTCOMES: Quarterly rates of out-of-hospital CT, MRI (as a comparator), and all other Medicare-funded diagnostic imaging examinations 2001-2019. RESULTS: CT scanning increased from 4663.5 per 100 000 person-years in 2001 to 14 506 in 2019 (211% increase), with substantial variation by type and anatomical region. The 2008-2009 PSR report was followed by an immediate reduction in CT scanning of 237.7 CTs per 100 000 people per quarter (95% CI -333.4 to -141.9) though growth in use soon continued at the preintervention rate. The degree of change in utilisation following the report differed between states/territories and by scan type, both in terms of the immediate change and the slope. For other diagnostic imaging modalities, there was an increase in the slope, while for MRI there was no change in either parameter. CONCLUSION: Actions consisting of financial disincentives for service overtesting and provider/public education components may limit excessive use of diagnostic imaging in fee-for-service systems, however, effects observed here were only short lived.

Trend in CT utilisation and its impact on length of stay, readmission and hospital mortality in Western Australia tertiary hospitals: an analysis of linked administrative data 2003-2015.

OBJECTIVE: High use of CT scanning has raised concern due to the potential ionising radiation exposure. This study examined trends of CT during admission to tertiary hospitals and its associations with length of stay (LOS), readmission and mortality. DESIGN: Retrospective observational study from 2003 to 2015. SETTING: West Australian linked administrative records at individual level. PARTICIPANTS: 2 375 787 episodes of tertiary hospital admission in adults aged 18+ years. MAIN OUTCOME MEASURES: LOS, 30-day readmissions and mortality stratified by CT use status (any, multiple (CTs to multiple areas during episode), and repeat (repeated CT to the same area)). METHODS: Multivariable regression models were used to calculate adjusted rate of CT use status. The significance of changes since 2003 in the outcomes (LOS, 30-day readmission and mortality) was compared among patients with specific CT imaging status relative to those without. RESULTS: Between 2003 and 2015, while the rate of CT increased 3.4% annually, the rate of repeat CTs significantly decreased -1.8% annually and multiple CT showed no change. Compared with 2003 while LOS had a greater decrease in those with any CT, 30-day readmissions had a greater increase among those with any CT, while the probability of mortality remained unchanged between the any CT/no CT groups. A similar result was observed in patients with multiple and repeat CT scanning, except for a significant increase in mortality in the recent years in the repeat CT group. CONCLUSION: The observed pattern of increase in CT utilisation is likely to be activity-based funding policy-driven based on the discordance between LOS and readmissions. Meanwhile, the repeat CT reduction aligns with a more selective strategy of use based on clinical severity. Future research should incorporate in-hospital and out-of-hospital CT to better understand overall CT trends and potential shifts between settings over time.

The use of computed tomography in the management of injury in tertiary emergency departments in Western Australia: Evidence of overtesting?

BACKGROUND: This study investigated trends in computed tomography (CT) utilization across different triage categories of injury presentations to tertiary emergency departments (EDs) and associations with diagnostic yield measured by injury severity, hospitalization and length of stay (LOS), and mortality. METHODS: A total of 411,155 injury-related ED presentations extracted from linked records from Western Australia from 2004 to 2015 were included in the retrospective study. The use of CT scanning and diagnostic yield measured by rate of diagnosis with severe injury, hospitalizations and LOS, and mortality were captured annually for injury-related ED presentations. Multivariable regression models were used to calculate the annual adjusted rate of CT scanning for injury presentations and hospitalizations across triage categories, diagnosis with severe injury, LOS, and mortality. The significance of changes observed was compared among patients with CT imaging relative to those without CT. RESULTS: While the number of ED presentations with injury increased by 65% from 2004 to 2015, the use of CT scanning in these presentations increased by 176%. The largest increase in CT use was among ED presentations triaged as semi-/nonurgent (+256%). Injury presentations with CT, compared to those without, had a higher rate of diagnosis with moderate/severe injury and hospitalization but no difference in LOS and mortality. The probability/rate observed in the outcomes of interest had a greater decrease over time in those with CT scanning compared with those without CT scanning across triage categories. CONCLUSIONS: The reduction in diagnostic yield in terms of injury severity and hospitalization found in our study might indicate a shift toward overtesting using CT in ED for injury or a higher use of CT to assist in the management of injuries. This helps health care policymakers consider whether the current increase in CT use meets the desired levels of quality and efficient care.

In vitro evaluation of a manganese chloride phantom-based MRI technique for quantitative determination of lumbar intervertebral disc composition and condition.

The application of MRI as a non-invasive, quantitative tool for diagnosing lumbar intervertebral disc degeneration is currently an area of active research. The objective of this study was to examine, in vitro, the efficacy of a manganese chloride phantom-based MRI technique for quantitatively assessing lumbar disc composition and degenerative condition. Sixteen human lumbar discs were imaged ex vivo using T2-weighted MRI, and assigned a quantitative grade based on the relative signal intensities of nine phantoms containing serial concentrations of manganese chloride. Discs were then graded macroscopically for degenerative condition, and water and uronic acid (glycosaminoglycan) contents were determined. MRI ranking exhibited significant and strong negative correlation with nucleus pulposus uronic acid content (r = -0.78). MRI grades were significantly higher for degenerate discs. The technique described presents immediate potential for in vitro studies requiring robust, minimally invasive and quantitative determination of lumbar disc composition and condition. Additionally, the technique may have potential as a clinical tool for diagnosing lumbar disc degeneration as it provides a standardised series of reference phantoms facilitating cross-platform consistency, requires short scan times and simple T2-weighted signal intensity measurements.

Esophageal adenocarcinoma and hypertrophic osteoarthropathy with improvement following resection of esophageal cancer.

Esophageal cancer usually presents late and carries a grave prognosis. Early diagnosis dramatically improves outcomes, and hence recognition of the varied presenting features, including hypertrophic osteoarthropathy (HOA), may be important. Hypertrophic osteoarthropathy is a rare manifestation of esophageal adenocarcinoma and indeed may herald the presence of this neoplasm, as it did in the case we present. A 59-year-old man presented with signs and symptoms suggestive of inflammatory wrist, ankle, and knee arthritis, and imaging revealed changes of HOA. He reported dysphagia mainly to solids, and endoscopic biopsy showed adenocarcinoma. Cancer resection treatment led to clinical resolution of his musculoskeletal symptoms. This case highlights the importance of recognition of HOA as a feature of this increasingly common cancer.

Increasing use of CT requested by emergency department physicians in tertiary hospitals in Western Australia 2003-2015: an analysis of linked administrative data.

OBJECTIVE: This study aimed to examine trends in number of CT scans requested by tertiary emergency department (ED) physicians in Western Australia (WA) from 2003 to 2015 across broad demographic and presentation characteristics, anatomical areas and presented symptoms. DESIGN: An observational cross-sectional study over study period from 2003 to 2015. SETTING: Linked administrative health service data at individual level from WA. PARTICIPANTS: A total of 1 666 884 tertiary hospital ED presentations of people aged 18 years or older were included in this study MAIN OUTCOME MEASURE: Number of CT scans requested by tertiary ED physicians in an ED presentation. METHODS: Poisson regression models were used to assess variation and trends in number of CT scans requested by ED physicians across demographic characteristics, clinical presentation characteristics and anatomical areas. RESULTS: Over the entire study duration, 71 per 1000 ED episodes had a CT requested by tertiary ED physicians. Between 2003 and 2015, the rate of CT scanning almost doubled from 58 to 105 per 1000 ED presentations. After adjusted for all observed characteristics, the rate of CT scans showed a downward trend from 2009 to 2011 and subsequent increase. Males, older individuals, those attending ED as a result of pain, those with neurological symptoms or injury or with higher priority triage code were the most likely to have CT requested by tertiary ED physicians. CONCLUSIONS: Noticeable changes in the number of CTs requested by tertiary ED physicians corresponded to the time frame of major health reforms happening within WA and nationally.

Factors driving CT utilisation in tertiary hospitals: a decomposition analysis using linked administrative data in Western Australia.

OBJECTIVES: While CT scanning plays a significant role in healthcare, its increasing use has raised concerns about inappropriate use. This study investigated factors driving the changing use of CT among people admitted to tertiary hospitals in Western Australia (WA). DESIGN AND SETTING: A repeated cross-sectional study of CT use in WA in 2003-2005 and 2013-2015 using linked administrative heath data at the individual patient level. PARTICIPANTS: A total of 2 375 787 tertiary hospital admissions of people aged 18 years or older. MAIN OUTCOME MEASURE: Rate of CT scanning per 1000 hospital admissions. METHODS: A multivariable decomposition model was used to quantify the contribution of changes in patient characteristics and changes in the probability of having a CT over the study period. RESULTS: The rate of CT scanning increased by 112 CT scans per 1000 admissions over the study period. Changes in the distribution of the observed patient characteristics were accounted for 62.7% of the growth in CT use. However, among unplanned admissions, changes in the distribution of patient characteristics only explained 17% of the growth in CT use, the remainder being explained by changes in the probability of having a CT scan. While the relative probability of having a CT scan generally increased over time across most observed characteristics, it reduced in young adults (-2.8%), people living in the rural/remote areas (-0.8%) and people transferred from secondary hospitals (-0.8%). CONCLUSIONS: Our study highlights potential improvements in practice towards reducing medical radiation exposure in certain high risk population. Since changes in the relative probability of having a CT scan (representing changes in scope) rather than changes in the distribution of the patient characteristics (representing changes in need) explained a major proportion of the growth in CT use, this warrants more in-depth investigations in clinical practices to better inform health policies promoting appropriate use of diagnostic imaging tests.