Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: a scoping review

Continuous prediction and clinical alarm management of late-onset sepsis in preterm infants using vital signs from a patient monitor

BACKGROUND AND OBJECTIVE

Continuous prediction of late-onset sepsis (LOS) could be helpful for improving clinical outcomes in neonatal intensive care units (NICU). This study aimed to develop an artificial intelligence (AI) model for assisting the bedside clinicians in successfully identifying infants at risk for LOS using non-invasive vital signs monitoring.

METHODS

In a retrospective study from the NICU of the Máxima Medical Center in Veldhoven, the Netherlands, a total of 492 preterm infants less than 32 weeks gestation were included between July 2016 and December 2018. Data on heart rate (HR), respiratory rate (RR), and oxygen saturation (SpO2) at 1 Hz were extracted from the patient monitor. We developed multiple AI models using 102 extracted features or raw time series to provide hourly LOS risk prediction. Shapley values were used to explain the model. For the best performing model, the effect of different vital signs and also the input type of signals on model performance was tested. To further assess the performance of applying the best performing model in a real-world clinical setting, we performed a simulation using four different alarm policies on continuous real-time predictions starting from three days after birth.

RESULTS

A total of 51 LOS patients and 68 controls were finally included according to the patient inclusion and exclusion criteria. When tested by seven-fold cross-validations, the mean (standard deviation) area under the receiver operating characteristic curve (AUC) six hours before CRASH was 0.875 (0.072) for the best performing model, compared to the other six models with AUC ranging from 0.782 (0.089) to 0.846 (0.083). The best performing model performed only slightly worse than the model learning from raw physiological waveforms (0.886 [0.068]), successfully detecting 96.1 % of LOS patients before CRASH. When setting the expected alarm window to 24 h and using a multi-threshold alarm policy, the sensitivity metric was 71.6 %, while the positive predictive value was 9.9 %, resulting in an average of 1.15 alarms per day per patient.

CONCLUSIONS

The proposed AI model, which learns from routinely collected vital signs, has the potential to assist clinicians in the early detection of LOS. Combined with interpretability and clinical alarm management, this model could be better translated into medical practice for future clinical implementation.

Artificial intelligence applied to magnetic resonance imaging reliably detects the presence, but not the location, of meniscus tears: a systematic review and meta-analysis

OBJECTIVES

To review and compare the accuracy of convolutional neural networks (CNN) for the diagnosis of meniscal tears in the current literature and analyze the decision-making processes utilized by these CNN algorithms.

MATERIALS AND METHODS

PubMed, MEDLINE, EMBASE, and Cochrane databases up to December 2022 were searched in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement. Risk of analysis was used for all identified articles. Predictive performance values, including sensitivity and specificity, were extracted for quantitative analysis. The meta-analysis was divided between AI prediction models identifying the presence of meniscus tears and the location of meniscus tears.

RESULTS

Eleven articles were included in the final review, with a total of 13,467 patients and 57,551 images. Heterogeneity was statistically significantly large for the sensitivity of the tear identification analysis (I2 = 79%). A higher level of accuracy was observed in identifying the presence of a meniscal tear over locating tears in specific regions of the meniscus (AUC, 0.939 vs 0.905). Pooled sensitivity and specificity were 0.87 (95% confidence interval (CI) 0.80-0.91) and 0.89 (95% CI 0.83-0.93) for meniscus tear identification and 0.88 (95% CI 0.82-0.91) and 0.84 (95% CI 0.81-0.85) for locating the tears.

CONCLUSIONS

AI prediction models achieved favorable performance in the diagnosis, but not location, of meniscus tears. Further studies on the clinical utilities of deep learning should include standardized reporting, external validation, and full reports of the predictive performances of these models, with a view to localizing tears more accurately.

CLINICAL RELEVANCE STATEMENT

Meniscus tears are hard to diagnose in the knee magnetic resonance images. AI prediction models may play an important role in improving the diagnostic accuracy of clinicians and radiologists.

KEY POINTS

• Artificial intelligence (AI) provides great potential in improving the diagnosis of meniscus tears. • The pooled diagnostic performance for artificial intelligence (AI) in identifying meniscus tears was better (sensitivity 87%, specificity 89%) than locating the tears (sensitivity 88%, specificity 84%). • AI is good at confirming the diagnosis of meniscus tears, but future work is required to guide the management of the disease.

Internet-based Surveillance Systems and Infectious Diseases Prediction: An Updated Review of the Last 10 Years and Lessons from the COVID-19 Pandemic

The last decade has seen major advances and growth in internet-based surveillance for infectious diseases through advanced computational capacity, growing adoption of smart devices, increased availability of Artificial Intelligence (AI), alongside environmental pressures including climate and land use change contributing to increased threat and spread of pandemics and emerging infectious diseases. With the increasing burden of infectious diseases and the COVID-19 pandemic, the need for developing novel technologies and integrating internet-based data approaches to improving infectious disease surveillance is greater than ever. In this systematic review, we searched the scientific literature for research on internet-based or digital surveillance for influenza, dengue fever and COVID-19 from 2013 to 2023. We have provided an overview of recent internet-based surveillance research for emerging infectious diseases (EID), describing changes in the digital landscape, with recommendations for future research directed at public health policymakers, healthcare providers, and government health departments to enhance traditional surveillance for detecting, monitoring, reporting, and responding to influenza, dengue, and COVID-19.

Comparative Analysis of Machine-Learning Model Performance in Image Analysis: The Impact of Dataset Diversity and Size

BACKGROUND

This study presents an analysis of machine-learning model performance in image analysis, with a specific focus on videolaryngoscopy procedures. The research aimed to explore how dataset diversity and size affect the performance of machine-learning models, an issue vital to the advancement of clinical artificial intelligence tools.

METHODS

A total of 377 videolaryngoscopy videos from YouTube were used to create 6 varied datasets, each differing in patient diversity and image count. The study also incorporates data augmentation techniques to enhance these datasets further. Two machine-learning models, YOLOv5-Small and YOLOv8-Small, were trained and evaluated on metrics such as F1 score (a statistical measure that combines the precision and recall of the model into a single metric, reflecting its overall accuracy), precision, recall, mAP@50, and mAP@50-95.

RESULTS

The findings indicate a significant impact of dataset configuration on model performance, especially the balance between diversity and quantity. The Multi-25 × 10 dataset, featuring 25 images from 10 different patients, demonstrates superior performance, highlighting the value of a well-balanced dataset. The study also finds that the effects of data augmentation vary across different types of datasets.

CONCLUSIONS

Overall, this study emphasizes the critical role of dataset structure in the performance of machine-learning models in medical image analysis. It underscores the necessity of striking an optimal balance between dataset size and diversity, thereby illuminating the complexities inherent in data-driven machine-learning development.

Nursing Variables Predicting Readmissions in Patients With a High Risk: A Scoping Review

Unplanned readmission endangers patient safety and increases unnecessary healthcare expenditure. Identifying nursing variables that predict patient readmissions can aid nurses in providing timely nursing interventions that help patients avoid readmission after discharge. We aimed to provide an overview of the nursing variables predicting readmission of patients with a high risk. The authors searched five databases-PubMed, CINAHL, EMBASE, Cochrane Library, and Scopus-for publications from inception to April 2023. Search terms included "readmission" and "nursing records." Eight studies were included for review. Nursing variables were classified into three categories-specifically, nursing assessment, nursing diagnosis, and nursing intervention. The nursing assessment category comprised 75% of the nursing variables; the proportions of the nursing diagnosis (25%) and nursing intervention categories (12.5%) were relatively low. Although most variables of the nursing assessment category focused on the patients' physical aspect, emotional and social aspects were also considered. This study demonstrated how nursing care contributes to patients' adverse outcomes. The findings can assist nurses in identifying the essential nursing assessment, diagnosis, and interventions, which should be provided from the time of patients' admission. This can mitigate preventable readmissions of patients with a high risk and facilitate their safe transition from an acute care setting to the community.

Evaluating accuracy and fairness of clinical decision support algorithms when health care resources are limited

OBJECTIVE

Guidance on how to evaluate accuracy and algorithmic fairness across subgroups is missing for clinical models that flag patients for an intervention but when health care resources to administer that intervention are limited. We aimed to propose a framework of metrics that would fit this specific use case.

METHODS

We evaluated the following metrics and applied them to a Veterans Health Administration clinical model that flags patients for intervention who are at risk of overdose or a suicidal event among outpatients who were prescribed opioids (N = 405,817): Receiver - Operating Characteristic and area under the curve, precision - recall curve, calibration - reliability curve, false positive rate, false negative rate, and false omission rate. In addition, we developed a new approach to visualize false positives and false negatives that we named 'per true positive bars.' We demonstrate the utility of these metrics to our use case for three cohorts of patients at the highest risk (top 0.5 %, 1.0 %, and 5.0 %) by evaluating algorithmic fairness across the following age groups: <=30, 31-50, 51-65, and >65 years old.

RESULTS

Metrics that allowed us to assess group differences more clearly were the false positive rate, false negative rate, false omission rate, and the new 'per true positive bars'. Metrics with limited utility to our use case were the Receiver - Operating Characteristic and area under the curve, the calibration - reliability curve, and the precision - recall curve.

CONCLUSION

There is no "one size fits all" approach to model performance monitoring and bias analysis. Our work informs future researchers and clinicians who seek to evaluate accuracy and fairness of predictive models that identify patients to intervene on in the context of limited health care resources. In terms of ease of interpretation and utility for our use case, the new 'per true positive bars' may be the most intuitive to a range of stakeholders and facilitates choosing a threshold that allows weighing false positives against false negatives, which is especially important when predicting severe adverse events.

Don't be misled: 3 misconceptions about external validation of clinical prediction models

Clinical prediction models provide risks of health outcomes that can inform patients and support medical decisions. However, most models never make it to actual implementation in practice. A commonly heard reason for this lack of implementation is that prediction models are often not externally validated. While we generally encourage external validation, we argue that an external validation is often neither sufficient nor required as an essential step before implementation. As such, any available external validation should not be perceived as a license for model implementation. We clarify this argument by discussing 3 common misconceptions about external validation. We argue that there is not one type of recommended validation design, not always a necessity for external validation, and sometimes a need for multiple external validations. The insights from this paper can help readers to consider, design, interpret, and appreciate external validation studies.

AI-Assisted Decision-Making in Long-Term Care: Qualitative Study on Prerequisites for Responsible Innovation

BACKGROUND

Although the use of artificial intelligence (AI)-based technologies, such as AI-based decision support systems (AI-DSSs), can help sustain and improve the quality and efficiency of care, their deployment creates ethical and social challenges. In recent years, a growing prevalence of high-level guidelines and frameworks for responsible AI innovation has been observed. However, few studies have specified the responsible embedding of AI-based technologies, such as AI-DSSs, in specific contexts, such as the nursing process in long-term care (LTC) for older adults.

OBJECTIVE

Prerequisites for responsible AI-assisted decision-making in nursing practice were explored from the perspectives of nurses and other professional stakeholders in LTC.

METHODS

Semistructured interviews were conducted with 24 care professionals in Dutch LTC, including nurses, care coordinators, data specialists, and care centralists. A total of 2 imaginary scenarios about AI-DSSs were developed beforehand and used to enable participants articulate their expectations regarding the opportunities and risks of AI-assisted decision-making. In addition, 6 high-level principles for responsible AI were used as probing themes to evoke further consideration of the risks associated with using AI-DSSs in LTC. Furthermore, the participants were asked to brainstorm possible strategies and actions in the design, implementation, and use of AI-DSSs to address or mitigate these risks. A thematic analysis was performed to identify the opportunities and risks of AI-assisted decision-making in nursing practice and the associated prerequisites for responsible innovation in this area.

RESULTS

The stance of care professionals on the use of AI-DSSs is not a matter of purely positive or negative expectations but rather a nuanced interplay of positive and negative elements that lead to a weighed perception of the prerequisites for responsible AI-assisted decision-making. Both opportunities and risks were identified in relation to the early identification of care needs, guidance in devising care strategies, shared decision-making, and the workload of and work experience of caregivers. To optimally balance the opportunities and risks of AI-assisted decision-making, seven categories of prerequisites for responsible AI-assisted decision-making in nursing practice were identified: (1) regular deliberation on data collection; (2) a balanced proactive nature of AI-DSSs; (3) incremental advancements aligned with trust and experience; (4) customization for all user groups, including clients and caregivers; (5) measures to counteract bias and narrow perspectives; (6) human-centric learning loops; and (7) the routinization of using AI-DSSs.

CONCLUSIONS

The opportunities of AI-assisted decision-making in nursing practice could turn into drawbacks depending on the specific shaping of the design and deployment of AI-DSSs. Therefore, we recommend considering the responsible use of AI-DSSs as a balancing act. Moreover, considering the interrelatedness of the identified prerequisites, we call for various actors, including developers and users of AI-DSSs, to cohesively address the different factors important to the responsible embedding of AI-DSSs in practice.

A systematic literature review of predicting patient discharges using statistical methods and machine learning

Discharge planning is integral to patient flow as delays can lead to hospital-wide congestion. Because a structured discharge plan can reduce hospital length of stay while enhancing patient satisfaction, this topic has caught the interest of many healthcare professionals and researchers. Predicting discharge outcomes, such as destination and time, is crucial in discharge planning by helping healthcare providers anticipate patient needs and resource requirements. This article examines the literature on the prediction of various discharge outcomes. Our review discovered papers that explore the use of prediction models to forecast the time, volume, and destination of discharged patients. Of the 101 reviewed papers, 49.5% looked at the prediction with machine learning tools, and 50.5% focused on prediction with statistical methods. The fact that knowing discharge outcomes in advance affects operational, tactical, medical, and administrative aspects is a frequent theme in the papers studied. Furthermore, conducting system-wide optimization, predicting the time and destination of patients after discharge, and addressing the primary causes of discharge delay in the process are among the recommendations for further research in this field.

Protein Biomarkers in Lung Cancer Screening: Technical Considerations and Feasibility Assessment

Lung cancer remains the leading cause of cancer-related deaths worldwide, mainly due to late diagnosis and the presence of metastases. Several countries around the world have adopted nation-wide LDCT-based lung cancer screening that will benefit patients, shifting the stage at diagnosis to earlier stages with more therapeutic options. Biomarkers can help to optimize the screening process, as well as refine the TNM stratification of lung cancer patients, providing information regarding prognostics and recommending management strategies. Moreover, novel adjuvant strategies will clearly benefit from previous knowledge of the potential aggressiveness and biological traits of a given early-stage surgically resected tumor. This review focuses on proteins as promising biomarkers in the context of lung cancer screening. Despite great efforts, there are still no successful examples of biomarkers in lung cancer that have reached the clinics to be used in early detection and early management. Thus, the field of biomarkers in early lung cancer remains an evident unmet need. A more specific objective of this review is to present an up-to-date technical assessment of the potential use of protein biomarkers in early lung cancer detection and management. We provide an overview regarding the benefits, challenges, pitfalls and constraints in the development process of protein-based biomarkers. Additionally, we examine how a number of emerging protein analytical technologies may contribute to the optimization of novel robust biomarkers for screening and effective management of lung cancer.

Designing medical artificial intelligence systems for global use: focus on interoperability, scalability, and accessibility

Advances in artificial intelligence (AI) and machine learning systems promise faster, more efficient, and more personalized care. While many of these models are built on the premise of improving access to the timely screening, diagnosis, and treatment of cardiovascular disease, their validity and accessibility across diverse and international cohorts remain unknown. In this mini-review article, we summarize key obstacles in the effort to design AI systems that will be scalable, accessible, and accurate across distinct geographical and temporal settings. We discuss representativeness, interoperability, quality assurance, and the importance of vendor-agnostic data types that will be available to end-users across the globe. These topics illustrate how the timely integration of these principles into AI development is crucial to maximizing the global benefits of AI in cardiology.

Use of kidney failure risk equation as a tool to evaluate referrals from primary care to specialist nephrology care

BACKGROUND

With rising costs and burden of chronic kidney disease (CKD), timely referral of patients to a kidney specialist is crucial. Currently, Kidney Health Australia (KHA) uses a 'heat map' based on severity and not future risk of kidney failure, whereas the kidney failure risk equation (KFRE) score predicts future risk of progression.

AIMS

Evaluate whether a KFRE score assists with timing of CKD referrals.

METHODS

Retrospective cohort of 2137 adult patients, referred to tertiary hospital outpatient nephrologist between 2012 and 2020, were analysed. Referrals were analysed for concordance with the KHA referral guidelines and, with the KFRE score, a recommended practice.

RESULTS

Of 2137 patients, 626 (29%) did not have urine albumin-to-creatinine ratio (UACR) measurement at referral. For those who had a UACR, the number who met KFRE preferred referral criteria was 36% less than KHA criteria. If the recommended KFRE score was used, then fewer older patients (≥40 years) needed referral. Positively, many diabetes patients were referred, even if their risk of kidney failure was low, and 29% had a KFRE over 3%. For patients evaluated meeting KFRE criteria, a larger proportion (76%) remained in follow-up, with only 8% being discharged.

CONCLUSIONS

KFRE could reduce referrals and be a useful tool to assist timely referrals. Using KFRE for triage may allow those patients with very low risk of future kidney failure not be referred, remaining longer in primary care, saving health resources and reducing patients' stress and wait times. Using KFRE encourages albuminuria measurement.

A practical guide to the implementation of AI in orthopaedic research, Part 6: How to evaluate the performance of AI research?

Artificial intelligence's (AI) accelerating progress demands rigorous evaluation standards to ensure safe, effective integration into healthcare's high-stakes decisions. As AI increasingly enables prediction, analysis and judgement capabilities relevant to medicine, proper evaluation and interpretation are indispensable. Erroneous AI could endanger patients; thus, developing, validating and deploying medical AI demands adhering to strict, transparent standards centred on safety, ethics and responsible oversight. Core considerations include assessing performance on diverse real-world data, collaborating with domain experts, confirming model reliability and limitations, and advancing interpretability. Thoughtful selection of evaluation metrics suited to the clinical context along with testing on diverse data sets representing different populations improves generalisability. Partnering software engineers, data scientists and medical practitioners ground assessment in real needs. Journals must uphold reporting standards matching AI's societal impacts. With rigorous, holistic evaluation frameworks, AI can progress towards expanding healthcare access and quality.

Level of Evidence

Level V.

A Neuronal Network-Based Score Predicting Survival in Patients Undergoing Aortic Valve Intervention: The ABC-AS Score

Background: Despite being the most commonly performed valvular intervention, risk prediction for aortic valve replacement in patients with severe aortic stenosis by currently used risk scores remains challenging. The study aim was to develop a biomarker-based risk score by means of a neuronal network. Methods: In this multicenter study, 3595 patients were divided into test and validation cohorts (70% to 30%) by random allocation. Input variables to develop the ABC-AS score were age, the cardiac biomarker high-sensitivity troponin T, and a patient history of cardiac decompensation. The validation cohort was used to verify the scores' value and for comparison with the Society of Thoracic Surgery Predictive Risk of Operative Mortality score. Results: Receiver operating curves demonstrated an improvement in prediction by using the ABC-AS score compared to the Society of Thoracic Surgery Predictive Risk of Operative Mortality (STS prom) score. Although the difference in predicting cardiovascular mortality was most notable at 30-day follow-up (area under the curve of 0.922 versus 0.678), ABC-AS also performed better in overall follow-up (0.839 versus 0.699). Furthermore, univariate analysis of ABC-AS tertiles yielded highly significant differences for all-cause (p < 0.0001) and cardiovascular mortality (p < 0.0001). Head-to-head comparison between both risk scores in a multivariable cox regression model underlined the potential of the ABC-AS score (HR per z-unit 2.633 (95% CI 2.156-3.216), p < 0.0001), while the STS prom score failed to reach statistical significance (p = 0.226). Conclusions: The newly developed ABC-AS score is an improved risk stratification tool to predict cardiovascular outcomes for patients undergoing aortic valve intervention.

DREAMER: a computational framework to evaluate readiness of datasets for machine learning

BACKGROUND

Machine learning (ML) has emerged as the predominant computational paradigm for analyzing large-scale datasets across diverse domains. The assessment of dataset quality stands as a pivotal precursor to the successful deployment of ML models. In this study, we introduce DREAMER (Data REAdiness for MachinE learning Research), an algorithmic framework leveraging supervised and unsupervised machine learning techniques to autonomously evaluate the suitability of tabular datasets for ML model development. DREAMER is openly accessible as a tool on GitHub and Docker, facilitating its adoption and further refinement within the research community..

RESULTS

The proposed model in this study was applied to three distinct tabular datasets, resulting in notable enhancements in their quality with respect to readiness for ML tasks, as assessed through established data quality metrics. Our findings demonstrate the efficacy of the framework in substantially augmenting the original dataset quality, achieved through the elimination of extraneous features and rows. This refinement yielded improved accuracy across both supervised and unsupervised learning methodologies.

CONCLUSION

Our software presents an automated framework for data readiness, aimed at enhancing the integrity of raw datasets to facilitate robust utilization within ML pipelines. Through our proposed framework, we streamline the original dataset, resulting in enhanced accuracy and efficiency within the associated ML algorithms.

Development of Machine Learning Models for the Identification of Elevated Ketone Bodies During Hyperglycemia in Patients with Type 1 Diabetes

Aims: Diabetic ketoacidosis (DKA) is a serious life-threatening condition caused by a lack of insulin, which leads to elevated plasma glucose and metabolic acidosis. Early identification of developing DKA is important to start treatment and minimize complications and risk of death. The aim of the present study is to develop and test prediction model(s) that gives an alarm about their risk of developing elevated ketone bodies during hyperglycemia. Methods: We analyzed data from 138 type 1 diabetes patients with measurements of ketone bodies and continuous glucose monitoring (CGM) data from over 30,000 days of wear time. We utilized a supervised binary classification machine learning approach to identify elevated levels of ketone bodies (≥0.6 mmol/L). Data material was randomly divided at patient level in 70%/30% (training/test) dataset. Logistic regression (LR) and random forest (RF) classifier were compared. Results: Among included patients, 913 ketone samples were eligible for modeling, including 273 event samples with ketone levels ≥0.6 mmol/L. An area under the receiver operating characteristic curve from the RF classifier was 0.836 (confidence interval [CI] 90%, 0.783-0.886) and 0.710 (CI 90%, 0.646-0.77) for the LR classifier. Conclusions: The novel approach for identifying elevated ketone levels in patients with type 1 diabetes utilized in this study indicates that CGM could be a valuable resource for the early prediction of patients at risk of developing DKA. Future studies are needed to validate the results.

Suitability of the Current Health Technology Assessment of Innovative Artificial Intelligence-Based Medical Devices: Scoping Literature Review

BACKGROUND

Artificial intelligence (AI)-based medical devices have garnered attention due to their ability to revolutionize medicine. Their health technology assessment framework is lacking.

OBJECTIVE

This study aims to analyze the suitability of each health technology assessment (HTA) domain for the assessment of AI-based medical devices.

METHODS

We conducted a scoping literature review following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) methodology. We searched databases (PubMed, Embase, and Cochrane Library), gray literature, and HTA agency websites.

RESULTS

A total of 10.1% (78/775) of the references were included. Data quality and integration are vital aspects to consider when describing and assessing the technical characteristics of AI-based medical devices during an HTA process. When it comes to implementing specialized HTA for AI-based medical devices, several practical challenges and potential barriers could be highlighted and should be taken into account (AI technological evolution timeline, data requirements, complexity and transparency, clinical validation and safety requirements, regulatory and ethical considerations, and economic evaluation).

CONCLUSIONS

The adaptation of the HTA process through a methodological framework for AI-based medical devices enhances the comparability of results across different evaluations and jurisdictions. By defining the necessary expertise, the framework supports the development of a skilled workforce capable of conducting robust and reliable HTAs of AI-based medical devices. A comprehensive adapted HTA framework for AI-based medical devices can provide valuable insights into the effectiveness, cost-effectiveness, and societal impact of AI-based medical devices, guiding their responsible implementation and maximizing their benefits for patients and health care systems.

PRERISK: A Personalized, Artificial Intelligence-Based and Statistically-Based Stroke Recurrence Predictor for Recurrent Stroke

BACKGROUND

Predicting stroke recurrence for individual patients is difficult, but individualized prediction may improve stroke survivors' engagement in self-care. We developed PRERISK: a statistical and machine learning classifier to predict individual risk of stroke recurrence.

METHODS

We analyzed clinical and socioeconomic data from a prospectively collected public health care-based data set of 41 975 patients admitted with stroke diagnosis in 88 public health centers over 6 years (2014-2020) in Catalonia-Spain. A new stroke diagnosis at least 24 hours after the index event was considered as a recurrent stroke, which was considered as our outcome of interest. We trained several supervised machine learning models to provide individualized risk over time and compared them with a Cox regression model. Models were trained to predict early, late, and long-term recurrence risk, within 90, 91 to 365, and >365 days, respectively. C statistics and area under the receiver operating characteristic curve were used to assess the accuracy of the models.

RESULTS

Overall, 16.21% (5932 of 36 114) of patients had stroke recurrence during a median follow-up of 2.69 years. The most powerful predictors of stroke recurrence were time from previous stroke, Barthel Index, atrial fibrillation, dyslipidemia, age, diabetes, and sex, which were used to create a simplified model with similar performance, together with modifiable vascular risk factors (glycemia, body mass index, high blood pressure, cholesterol, tobacco dependence, and alcohol abuse). The areas under the receiver operating characteristic curve were 0.76 (95% CI, 0.74-0.77), 0.60 (95% CI, 0.58-0.61), and 0.71 (95% CI, 0.69-0.72) for early, late, and long-term recurrence risk, respectively. The areas under the receiver operating characteristic curve of the Cox risk class probability were 0.73 (95% CI, 0.72-0.75), 0.59 (95% CI, 0.57-0.61), and 0.67 (95% CI, 0.66-0.70); machine learning approaches (random forest and AdaBoost) showed statistically significant improvement (P<0.05) over the Cox model for the 3 recurrence time periods. Stroke recurrence curves can be simulated for each patient under different degrees of control of modifiable factors.

CONCLUSIONS

PRERISK is a novel approach that provides a personalized and fairly accurate risk prediction of stroke recurrence over time. The model has the potential to incorporate dynamic control of risk factors.

Frameworks for procurement, integration, monitoring, and evaluation of artificial intelligence tools in clinical settings: A systematic review

Research on the applications of artificial intelligence (AI) tools in medicine has increased exponentially over the last few years but its implementation in clinical practice has not seen a commensurate increase with a lack of consensus on implementing and maintaining such tools. This systematic review aims to summarize frameworks focusing on procuring, implementing, monitoring, and evaluating AI tools in clinical practice. A comprehensive literature search, following PRSIMA guidelines was performed on MEDLINE, Wiley Cochrane, Scopus, and EBSCO databases, to identify and include articles recommending practices, frameworks or guidelines for AI procurement, integration, monitoring, and evaluation. From the included articles, data regarding study aim, use of a framework, rationale of the framework, details regarding AI implementation involving procurement, integration, monitoring, and evaluation were extracted. The extracted details were then mapped on to the Donabedian Plan, Do, Study, Act cycle domains. The search yielded 17,537 unique articles, out of which 47 were evaluated for inclusion based on their full texts and 25 articles were included in the review. Common themes extracted included transparency, feasibility of operation within existing workflows, integrating into existing workflows, validation of the tool using predefined performance indicators and improving the algorithm and/or adjusting the tool to improve performance. Among the four domains (Plan, Do, Study, Act) the most common domain was Plan (84%, n = 21), followed by Study (60%, n = 15), Do (52%, n = 13), & Act (24%, n = 6). Among 172 authors, only 1 (0.6%) was from a low-income country (LIC) and 2 (1.2%) were from lower-middle-income countries (LMICs). Healthcare professionals cite the implementation of AI tools within clinical settings as challenging owing to low levels of evidence focusing on integration in the Do and Act domains. The current healthcare AI landscape calls for increased data sharing and knowledge translation to facilitate common goals and reap maximum clinical benefit.

Automatic assessment of atherosclerotic plaque features by intracoronary imaging: a scoping review

Background

The diagnostic performance and clinical validity of automatic intracoronary imaging (ICI) tools for atherosclerotic plaque assessment have not been systematically investigated so far.

Methods

We performed a scoping review including studies on automatic tools for automatic plaque components assessment by means of optical coherence tomography (OCT) or intravascular imaging (IVUS). We summarized study characteristics and reported the specifics and diagnostic performance of developed tools.

Results

Overall, 42 OCT and 26 IVUS studies fulfilling the eligibility criteria were found, with the majority published in the last 5 years (86% of the OCT and 73% of the IVUS studies). A convolutional neural network deep-learning method was applied in 71% of OCT- and 34% of IVUS-studies. Calcium was the most frequent plaque feature analyzed (26/42 of OCT and 12/26 of IVUS studies), and both modalities showed high discriminatory performance in testing sets [range of area under the curve (AUC): 0.91-0.99 for OCT and 0.89-0.98 for IVUS]. Lipid component was investigated only in OCT studies (n = 26, AUC: 0.82-0.86). Fibrous cap thickness or thin-cap fibroatheroma were mainly investigated in OCT studies (n = 8, AUC: 0.82-0.94). Plaque burden was mainly assessed in IVUS studies (n = 15, testing set AUC reported in one study: 0.70).

Conclusion

A limited number of automatic machine learning-derived tools for ICI analysis is currently available. The majority have been developed for calcium detection for either OCT or IVUS images. The reporting of the development and validation process of automated intracoronary imaging analyses is heterogeneous and lacks critical information.

Systematic Review Registration

Open Science Framework (OSF), https://osf.io/nps2b/.Graphical AbstractCentral Illustration.

Patient and Staff Experience of Remote Patient Monitoring-What to Measure and How: Systematic Review

BACKGROUND

Patient and staff experience is a vital factor to consider in the evaluation of remote patient monitoring (RPM) interventions. However, no comprehensive overview of available RPM patient and staff experience-measuring methods and tools exists.

OBJECTIVE

This review aimed at obtaining a comprehensive set of experience constructs and corresponding measuring instruments used in contemporary RPM research and at proposing an initial set of guidelines for improving methodological standardization in this domain.

METHODS

Full-text papers reporting on instances of patient or staff experience measuring in RPM interventions, written in English, and published after January 1, 2011, were considered for eligibility. By "RPM interventions," we referred to interventions including sensor-based patient monitoring used for clinical decision-making; papers reporting on other kinds of interventions were therefore excluded. Papers describing primary care interventions, involving participants under 18 years of age, or focusing on attitudes or technologies rather than specific interventions were also excluded. We searched 2 electronic databases, Medline (PubMed) and EMBASE, on February 12, 2021.We explored and structured the obtained corpus of data through correspondence analysis, a multivariate statistical technique.

RESULTS

In total, 158 papers were included, covering RPM interventions in a variety of domains. From these studies, we reported 546 experience-measuring instances in RPM, covering the use of 160 unique experience-measuring instruments to measure 120 unique experience constructs. We found that the research landscape has seen a sizeable growth in the past decade, that it is affected by a relative lack of focus on the experience of staff, and that the overall corpus of collected experience measures can be organized in 4 main categories (service system related, care related, usage and adherence related, and health outcome related). In the light of the collected findings, we provided a set of 6 actionable recommendations to RPM patient and staff experience evaluators, in terms of both what to measure and how to measure it. Overall, we suggested that RPM researchers and practitioners include experience measuring as part of integrated, interdisciplinary data strategies for continuous RPM evaluation.

CONCLUSIONS

At present, there is a lack of consensus and standardization in the methods used to measure patient and staff experience in RPM, leading to a critical knowledge gap in our understanding of the impact of RPM interventions. This review offers targeted support for RPM experience evaluators by providing a structured, comprehensive overview of contemporary patient and staff experience measures and a set of practical guidelines for improving research quality and standardization in this domain.

Managing Risk and Quality of AI in Healthcare: Are Hospitals Ready for Implementation?

Artificial intelligence (AI) provides a unique opportunity to help meet the demands of the future healthcare system. However, hospitals may not be well equipped to handle safe and effective development and/or procurement of AI systems. Furthermore, upcoming regulations such as the EU AI Act may enforce the need to establish new management systems, quality assurance and control mechanisms, novel to healthcare organizations. This paper discusses challenges in AI implementation, particularly potential gaps in current management systems (MS), by reviewing the harmonized standard for AI MS, ISO 42001, as part of a gap analysis of a tertiary acute hospital with ongoing AI activities. Examination of the industry agnostic ISO 42001 reveals a technical debt within healthcare, aligning with previous research on digitalization and AI implementation. To successfully implement AI with quality assurance in mind, emphasis should be put on the foundation and structure of the healthcare organizations, including both workforce and data infrastructure.

Fairness and bias correction in machine learning for depression prediction across four study populations

A significant level of stigma and inequality exists in mental healthcare, especially in under-served populations. Inequalities are reflected in the data collected for scientific purposes. When not properly accounted for, machine learning (ML) models learned from data can reinforce these structural inequalities or biases. Here, we present a systematic study of bias in ML models designed to predict depression in four different case studies covering different countries and populations. We find that standard ML approaches regularly present biased behaviors. We also show that mitigation techniques, both standard and our own post-hoc method, can be effective in reducing the level of unfair bias. There is no one best ML model for depression prediction that provides equality of outcomes. This emphasizes the importance of analyzing fairness during model selection and transparent reporting about the impact of debiasing interventions. Finally, we also identify positive habits and open challenges that practitioners could follow to enhance fairness in their models.

Artificial-Intelligence-Enhanced Analysis of In Vivo Confocal Microscopy in Corneal Diseases: A Review

Artificial intelligence (AI) has seen significant progress in medical diagnostics, particularly in image and video analysis. This review focuses on the application of AI in analyzing in vivo confocal microscopy (IVCM) images for corneal diseases. The cornea, as an exposed and delicate part of the body, necessitates the precise diagnoses of various conditions. Convolutional neural networks (CNNs), a key component of deep learning, are a powerful tool for image data analysis. This review highlights AI applications in diagnosing keratitis, dry eye disease, and diabetic corneal neuropathy. It discusses the potential of AI in detecting infectious agents, analyzing corneal nerve morphology, and identifying the subtle changes in nerve fiber characteristics in diabetic corneal neuropathy. However, challenges still remain, including limited datasets, overfitting, low-quality images, and unrepresentative training datasets. This review explores augmentation techniques and the importance of feature engineering to address these challenges. Despite the progress made, challenges are still present, such as the "black-box" nature of AI models and the need for explainable AI (XAI). Expanding datasets, fostering collaborative efforts, and developing user-friendly AI tools are crucial for enhancing the acceptance and integration of AI into clinical practice.

Large language models and multimodal foundation models for precision oncology

The technological progress in artificial intelligence (AI) has massively accelerated since 2022, with far-reaching implications for oncology and cancer research. Large language models (LLMs) now perform at human-level competency in text processing. Notably, both text and image processing networks are increasingly based on transformer neural networks. This convergence enables the development of multimodal AI models that take diverse types of data as an input simultaneously, marking a qualitative shift from specialized niche models which were prevalent in the 2010s. This editorial summarizes these developments, which are expected to impact precision oncology in the coming years.

Development and Validation of an Interpretable Conformal Predictor to Predict Sepsis Mortality Risk: Retrospective Cohort Study

BACKGROUND

Early and reliable identification of patients with sepsis who are at high risk of mortality is important to improve clinical outcomes. However, 3 major barriers to artificial intelligence (AI) models, including the lack of interpretability, the difficulty in generalizability, and the risk of automation bias, hinder the widespread adoption of AI models for use in clinical practice.

OBJECTIVE

This study aimed to develop and validate (internally and externally) a conformal predictor of sepsis mortality risk in patients who are critically ill, leveraging AI-assisted prediction modeling. The proposed approach enables explaining the model output and assessing its confidence level.

METHODS

We retrospectively extracted data on adult patients with sepsis from a database collected in a teaching hospital at Beth Israel Deaconess Medical Center for model training and internal validation. A large multicenter critical care database from the Philips eICU Research Institute was used for external validation. A total of 103 clinical features were extracted from the first day after admission. We developed an AI model using gradient-boosting machines to predict the mortality risk of sepsis and used Mondrian conformal prediction to estimate the prediction uncertainty. The Shapley additive explanation method was used to explain the model.

RESULTS

A total of 16,746 (80%) patients from Beth Israel Deaconess Medical Center were used to train the model. When tested on the internal validation population of 4187 (20%) patients, the model achieved an area under the receiver operating characteristic curve of 0.858 (95% CI 0.845-0.871), which was reduced to 0.800 (95% CI 0.789-0.811) when externally validated on 10,362 patients from the Philips eICU database. At a specified confidence level of 90% for the internal validation cohort the percentage of error predictions (n=438) out of all predictions (n=4187) was 10.5%, with 1229 (29.4%) predictions requiring clinician review. In contrast, the AI model without conformal prediction made 1449 (34.6%) errors. When externally validated, more predictions (n=4004, 38.6%) were flagged for clinician review due to interdatabase heterogeneity. Nevertheless, the model still produced significantly lower error rates compared to the point predictions by AI (n=1221, 11.8% vs n=4540, 43.8%). The most important predictors identified in this predictive model were Acute Physiology Score III, age, urine output, vasopressors, and pulmonary infection. Clinically relevant risk factors contributing to a single patient were also examined to show how the risk arose.

CONCLUSIONS

By combining model explanation and conformal prediction, AI-based systems can be better translated into medical practice for clinical decision-making.

Modest performance of text mining to extract health outcomes may be almost sufficient for high-quality prognostic model development

BACKGROUND

Across medicine, prognostic models are used to estimate patient risk of certain future health outcomes (e.g., cardiovascular or mortality risk). To develop (or train) prognostic models, historic patient-level training data is needed containing both the predictive factors (i.e., features) and the relevant health outcomes (i.e., labels). Sometimes, when the health outcomes are not recorded in structured data, these are first extracted from textual notes using text mining techniques. Because there exist many studies utilizing text mining to obtain outcome data for prognostic model development, our aim is to study the impact of the text mining quality on downstream prognostic model performance.

METHODS

We conducted a simulation study charting the relationship between text mining quality and prognostic model performance using an illustrative case study about in-hospital mortality prediction in intensive care unit patients. We repeatedly developed and evaluated a prognostic model for in-hospital mortality, using outcome data extracted by multiple text mining models of varying quality.

RESULTS

Interestingly, we found in our case study that a relatively low-quality text mining model (F1 score ≈ 0.50) could already be used to train a prognostic model with quite good discrimination (area under the receiver operating characteristic curve of around 0.80). The calibration of the risks estimated by the prognostic model seemed unreliable across the majority of settings, even when text mining models were of relatively high quality (F1 ≈ 0.80).

DISCUSSION

Developing prognostic models on text-extracted outcomes using imperfect text mining models seems promising. However, it is likely that prognostic models developed using this approach may not produce well-calibrated risk estimates, and require recalibration in (possibly a smaller amount of) manually extracted outcome data.

Trust and acceptability of data-driven clinical recommendations in everyday practice: A scoping review

BACKGROUND

Increasing attention is being given to the analysis of large health datasets to derive new clinical decision support systems (CDSS). However, few data-driven CDSS are being adopted into clinical practice. Trust in these tools is believed to be fundamental for acceptance and uptake but to date little attention has been given to defining or evaluating trust in clinical settings.

OBJECTIVES

A scoping review was conducted to explore how and where acceptability and trustworthiness of data-driven CDSS have been assessed from the health professional's perspective.

METHODS

Medline, Embase, PsycInfo, Web of Science, Scopus, ACM Digital, IEEE Xplore and Google Scholar were searched in March 2022 using terms expanded from: "data-driven" AND "clinical decision support" AND "acceptability". Included studies focused on healthcare practitioner-facing data-driven CDSS, relating directly to clinical care. They included trust or a proxy as an outcome, or in the discussion. The preferred reporting items for systematic reviews and meta-analyses extension for scoping reviews (PRISMA-ScR) is followed in the reporting of this review.

RESULTS

3291 papers were screened, with 85 primary research studies eligible for inclusion. Studies covered a diverse range of clinical specialisms and intended contexts, but hypothetical systems (24) outnumbered those in clinical use (18). Twenty-five studies measured trust, via a wide variety of quantitative, qualitative and mixed methods. A further 24 discussed themes of trust without it being explicitly evaluated, and from these, themes of transparency, explainability, and supporting evidence were identified as factors influencing healthcare practitioner trust in data-driven CDSS.

CONCLUSION

There is a growing body of research on data-driven CDSS, but few studies have explored stakeholder perceptions in depth, with limited focused research on trustworthiness. Further research on healthcare practitioner acceptance, including requirements for transparency and explainability, should inform clinical implementation.

N-of-1 medicine

ABSTRACT

The fields of precision and personalised medicine have led to promising advances in tailoring treatment to individual patients. Examples include genome/molecular alteration-guided drug selection, single-patient gene therapy design and synergy-based drug combination development, and these approaches can yield substantially diverse recommendations. Therefore, it is important to define each domain and delineate their commonalities and differences in an effort to develop novel clinical trial designs, streamline workflow development, rethink regulatory considerations, create value in healthcare and economics assessments, and other factors. These and other segments are essential to recognise the diversity within these domains to accelerate their respective workflows towards practice-changing healthcare. To emphasise these points, this article elaborates on the concept of digital health and digital medicine-enabled N-of-1 medicine, which individualises combination regimen and dosing using a patient's own data. We will conclude with recommendations for consideration when developing novel workflows based on emerging digital-based platforms.

Leveraging artificial intelligence for perioperative cancer risk assessment of oral potentially malignant disorders

Oral potentially malignant disorders (OPMDs) are mucosal conditions with an inherent disposition to develop oral squamous cell carcinoma. Surgical management is the most preferred strategy to prevent malignant transformation in OPMDs, and surgical approaches to treatment include conventional scalpel excision, laser surgery, cryotherapy, and photodynamic therapy. However, in reality, since all patients with OPMDs will not develop oral squamous cell carcinoma in their lifetime, there is a need to stratify patients according to their risk of malignant transformation to streamline surgical intervention for patients with the highest risks. Artificial intelligence (AI) has the potential to integrate disparate factors influencing malignant transformation for robust, precise, and personalized cancer risk stratification of OPMD patients than current methods to determine the need for surgical resection, excision, or re-excision. Therefore, this article overviews existing AI models and tools, presents a clinical implementation pathway, and discusses necessary refinements to aid the clinical application of AI-based platforms for cancer risk stratification of OPMDs in surgical practice.

Time-domain heart rate variability features for automatic congestive heart failure prediction

AIMS

Heart failure is a serious condition that often goes undiagnosed in primary care due to the lack of reliable diagnostic tools and the similarity of its symptoms with other diseases. Non-invasive monitoring of heart rate variability (HRV), which reflects the activity of the autonomic nervous system, could offer a novel and accurate way to detect and manage heart failure patients. This study aimed to assess the feasibility of using machine learning techniques on HRV data as a non-invasive biomarker to classify healthy adults and those with heart failure.

METHODS AND RESULTS

We used digitized electrocardiogram recordings from 54 adults with normal sinus rhythm and 44 adults categorized into New York Heart Association classes 1, 2, and 3, suffering from congestive heart failure. All recordings were sourced from the PhysioNet database. Following data pre-processing, we performed time-domain HRV analysis on all individual recordings, including root mean square of the successive difference in adjacent RR interval (RRi) (RMSSD), the standard deviation of RRi (SDNN, the NN stands for natural or sinus intervals), the standard deviation of the successive differences between successive RRi (SDSD), the number or percentage of RRi longer than 50 ms (NN50 and pNN50), and the average value of RRi [mean RR interval (mRRi)]. In our experimental classification performance evaluation, on the computed HRV parameters, we optimized hyperparameters and performed five-fold cross-validation using four machine learning classification algorithms: support vector machine, k-nearest neighbour (KNN), naïve Bayes, and decision tree (DT). We evaluated the prediction accuracy of these models using performance criteria, namely, precision, recall, specificity, F1 score, and overall accuracy. For added insight, we also presented receiver operating characteristic (ROC) plots and area under the ROC curve (AUC) values. The overall best performance accuracy of 77% was achieved when KNN and DT were trained on computed HRV parameters with a 5 min time window. KNN obtained an AUC of 0.77, while DT attained 0.78. Additionally, in the classification of severe congestive heart failure, KNN and DT had the best accuracy of 91%, with KNN achieving an AUC of 0.88 and DT obtaining 0.92.

CONCLUSIONS

The results show that HRV can accurately predict severe congestive heart failure. The findings of this study could inform the use of machine learning approaches on non-invasive HRV, to screen congestive heart failure individuals in primary care.

Deep learning and machine learning predictive models for neurological function after interventional embolization of intracranial aneurysms

Objective

The objective of this study is to develop a model to predicts the postoperative Hunt-Hess grade in patients with intracranial aneurysms by integrating radiomics and deep learning technologies, using preoperative CTA imaging data. Thereby assisting clinical decision-making and improving the assessment and prognosis of postoperative neurological function.

Methods

This retrospective study encompassed 101 patients who underwent aneurysm embolization surgery. 851 radiomic features were extracted from CTA images. 512 deep learning features are extracted from last layer of ResNet50 deep convolutional neural network model. The feature screening process pipeline encompassed intraclass correlation coefficient analysis, principal component analysis, U test, spearman correlation analysis, minimum redundancy maximum relevance algorithm and Lasso regression, to identify features most correlated with postoperative Hunt-Hess grading. In the model construction phase, three distinct models were constructed: radiomics feature-based model (RSM), deep learning feature-based model (DLM), and deep learning-radiomics feature fusion model (DLRSCM). The study also calculated the radiomics score and combined it with clinical data to construct a Nomogram for predictive modeling. DLM, RSM and DLRSCM model was constructed by 9 base algorithms and 1 ensemble learning algorithm - Stacking ensemble model. Model performance was evaluated based on the area under the Receiver Operating Characteristic (ROC) curve (AUC), Matthews Correlation Coefficient (MCC), calibration curves, and decision curves analysis.

Results

5 significant radiomic feature and 4 significant deep learning features were obtained through the feature selection process. These features were utilized for model construction. Bootstrap resampling method was used for internal validation of the models. In terms of model evaluation, the DLM model, the stacking ensemble algorithm results achieved an AUC of 0.959 and MCC of 0.815. In the RSM model, the stacking ensemble model AUC was 0.935 and MCC was 0.793. The stacking ensemble model in DLRSCM outperformed others, with an AUC of 0.968 and MCC of 0.820. Results indicated that the ANN performed optimally among all base models, while the stacked ensemble learning model exhibited the highest predictive performance.

Conclusion

This study demonstrates that the combination of radiomics and deep learning is an effective approach to predict the postoperative Hunt-Hess grade in patients with intracranial aneurysms. This holds significant value in the early identification of postoperative neurological complications and in enhancing clinical decision-making.

Predicting anti-TNF treatment response in rheumatoid arthritis: An artificial intelligence-driven model using cytokine profile and routine clinical practice parameters

Introduction

Rheumatoid arthritis (RA) is a heterogeneous disease in which therapeutic strategies used have evolved dramatically. Despite significant progress in treatment strategies such as the development of anti-TNF drugs, it is still not possible to differentiate those patients who will respond from who will not. This can lead to effective-treatment delays and unnecessary costs. The aim of this study was to utilize a profile of the patient's characteristics, clinical parameters, immune status (cytokine profile) and artificial intelligence to assess the feasibility of developing a tool that could allow us to predict which patients will respond to treatment with anti-TNF drugs.

Methods

This study included 38 patients with RA from the RA-Paz cohort. Clinical activity was measured at baseline and after 6 months of treatment. The cytokines measured before the start of anti-TNF treatment were IL-1, IL-12, IL-10, IL-2, IL-4, IFNg, TNFa, and IL-6. Statistical analyses were performed using the Wilcoxon-Rank-Sum Test and the Benjamini-Hochberg method. The predictive model viability was explored using the 5-fold cross-validation scheme in order to train the logistic regression models.

Results

Statistically significant differences were found in parameters such as IL-6, IL-2, CRP and DAS-ESR. The predictive model performed to an acceptable level in correctly classifying patients (ROC-AUC 0.804167 to 0.891667), suggesting that it would be possible to develop a clinical classification tool.

Conclusions

Using a combination of parameters such as IL-6, IL-2, CRP and DAS-ESR, it was possible to develop a predictive model that can acceptably discriminate between remitters and non-remitters. However, this model needs to be replicated in a larger cohort to confirm these findings.

The ménage à trois of healthcare: the actors in after-AI era under patient consent

Introduction

Artificial intelligence has become an increasingly powerful technological instrument in recent years, revolutionizing many sectors, including public health. Its use in this field will inevitably change clinical practice, the patient-caregiver relationship and the concept of the diagnosis and treatment pathway, affecting the balance between the patient's right to self-determination and health, and thus leading to an evolution of the concept of informed consent. The aim was to characterize the guidelines for the use of artificial intelligence, its areas of application and the relevant legislation, to propose guiding principles for the design of optimal informed consent for its use.

Materials and methods

A classic review by keywords on the main search engines was conducted. An analysis of the guidelines and regulations issued by scientific authorities and legal bodies on the use of artificial intelligence in public health was carried out.

Results

The current areas of application of this technology were highlighted, divided into sectors, its impact on them, as well as a summary of current guidelines and legislation.

Discussion

The ethical implications of artificial intelligence in the health care system were assessed, particularly regarding the therapeutic alliance between doctor and patient, and the balance between the right to self-determination and health. Finally, given the evolution of informed consent in relation to the use of this new technology, seven guiding principles were proposed to guarantee the right to the most informed consent or dissent.

Peer review of GPT-4 technical report and systems card

The study provides a comprehensive review of OpenAI's Generative Pre-trained Transformer 4 (GPT-4) technical report, with an emphasis on applications in high-risk settings like healthcare. A diverse team, including experts in artificial intelligence (AI), natural language processing, public health, law, policy, social science, healthcare research, and bioethics, analyzed the report against established peer review guidelines. The GPT-4 report shows a significant commitment to transparent AI research, particularly in creating a systems card for risk assessment and mitigation. However, it reveals limitations such as restricted access to training data, inadequate confidence and uncertainty estimations, and concerns over privacy and intellectual property rights. Key strengths identified include the considerable time and economic investment in transparent AI research and the creation of a comprehensive systems card. On the other hand, the lack of clarity in training processes and data raises concerns about encoded biases and interests in GPT-4. The report also lacks confidence and uncertainty estimations, crucial in high-risk areas like healthcare, and fails to address potential privacy and intellectual property issues. Furthermore, this study emphasizes the need for diverse, global involvement in developing and evaluating large language models (LLMs) to ensure broad societal benefits and mitigate risks. The paper presents recommendations such as improving data transparency, developing accountability frameworks, establishing confidence standards for LLM outputs in high-risk settings, and enhancing industry research review processes. It concludes that while GPT-4's report is a step towards open discussions on LLMs, more extensive interdisciplinary reviews are essential for addressing bias, harm, and risk concerns, especially in high-risk domains. The review aims to expand the understanding of LLMs in general and highlights the need for new reflection forms on how LLMs are reviewed, the data required for effective evaluation, and addressing critical issues like bias and risk.

Antimicrobial learning systems: an implementation blueprint for artificial intelligence to tackle antimicrobial resistance

The proliferation of various forms of artificial intelligence (AI) brings many opportunities to improve health care. AI models can harness complex evolving data, inform and augment human actions, and learn from health outcomes such as morbidity and mortality. The global public health challenge of antimicrobial resistance (AMR) needs large-scale optimisation of antimicrobial use and wider infection care, which could be enabled by carefully constructed AI models. As AI models become increasingly useful and robust, health-care systems remain challenging places for their deployment. An implementation gap exists between the promise of AI models and their use in patient and population care. Here, we outline an adaptive implementation and maintenance framework for AI models to improve antimicrobial use and infection care as a learning system. The roles of AMR problem identification, law and regulation, organisational support, data processing, and AI development, assessment, maintenance, and scalability in the implementation of AMR-targeted AI models are considered.

Prediction of Complications and Prognostication in Perioperative Medicine: A Systematic Review and PROBAST Assessment of Machine Learning Tools

BACKGROUND

The utilization of artificial intelligence and machine learning as diagnostic and predictive tools in perioperative medicine holds great promise. Indeed, many studies have been performed in recent years to explore the potential. The purpose of this systematic review is to assess the current state of machine learning in perioperative medicine, its utility in prediction of complications and prognostication, and limitations related to bias and validation.

METHODS

A multidisciplinary team of clinicians and engineers conducted a systematic review using the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) protocol. Multiple databases were searched, including Scopus, Cumulative Index to Nursing and Allied Health Literature (CINAHL), the Cochrane Library, PubMed, Medline, Embase, and Web of Science. The systematic review focused on study design, type of machine learning model used, validation techniques applied, and reported model performance on prediction of complications and prognostication. This review further classified outcomes and machine learning applications using an ad hoc classification system. The Prediction model Risk Of Bias Assessment Tool (PROBAST) was used to assess risk of bias and applicability of the studies.

RESULTS

A total of 103 studies were identified. The models reported in the literature were primarily based on single-center validations (75%), with only 13% being externally validated across multiple centers. Most of the mortality models demonstrated a limited ability to discriminate and classify effectively. The PROBAST assessment indicated a high risk of systematic errors in predicted outcomes and artificial intelligence or machine learning applications.

CONCLUSIONS

The findings indicate that the development of this field is still in its early stages. This systematic review indicates that application of machine learning in perioperative medicine is still at an early stage. While many studies suggest potential utility, several key challenges must be first overcome before their introduction into clinical practice.

EDITOR’S PERSPECTIVE

Artificial intelligence-based prediction models for acute myeloid leukemia using real-life data: A DATAML registry study

We designed artificial intelligence-based prediction models (AIPM) using 52 diagnostic variables from 3687 patients included in the DATAML registry treated with intensive chemotherapy (IC, N = 3030) or azacitidine (AZA, N = 657) for an acute myeloid leukemia (AML). A neural network called multilayer perceptron (MLP) achieved a prediction accuracy for overall survival (OS) of 68.5% and 62.1% in the IC and AZA cohorts, respectively. The Boruta algorithm could select the most important variables for prediction without decreasing accuracy. Thirteen features were retained with this algorithm in the IC cohort: age, cytogenetic risk, white blood cells count, LDH, platelet count, albumin, MPO expression, mean corpuscular volume, CD117 expression, NPM1 mutation, AML status (de novo or secondary), multilineage dysplasia and ASXL1 mutation; and 7 variables in the AZA cohort: blood blasts, serum ferritin, CD56, LDH, hemoglobin, CD13 and disseminated intravascular coagulation (DIC). We believe that AIPM could help hematologists to deal with the huge amount of data available at diagnosis, enabling them to have an OS estimation and guide their treatment choice. Our registry-based AIPM could offer a large real-life dataset with original and exhaustive features and select a low number of diagnostic features with an equivalent accuracy of prediction, more appropriate to routine practice.

Systematic reviews of machine learning in healthcare: a literature review

INTRODUCTION

The increasing availability of data and computing power has made machine learning (ML) a viable approach to faster, more efficient healthcare delivery.

METHODS

A systematic literature review (SLR) of published SLRs evaluating ML applications in healthcare settings published between1 January 2010 and 27 March 2023 was conducted.

RESULTS

In total 220 SLRs covering 10,462 ML algorithms were reviewed. The main application of AI in medicine related to the clinical prediction and disease prognosis in oncology and neurology with the use of imaging data. Accuracy, specificity, and sensitivity were provided in 56%, 28%, and 25% SLRs respectively. Internal and external validation was reported in 53% and less than 1% of the cases respectively. The most common modeling approach was neural networks (2,454 ML algorithms), followed by support vector machine and random forest/decision trees (1,578 and 1,522 ML algorithms, respectively).

EXPERT OPINION

The review indicated considerable reporting gaps in terms of the ML's performance, both internal and external validation. Greater accessibility to healthcare data for developers can ensure the faster adoption of ML algorithms into clinical practice.

Exploring the Ethical, Legal, and Social Implications of ChatGPT in Urology

PURPOSE OF THE REVIEW

ChatGPT is programmed to generate responses based on pattern recognition. With this vast popularity and exponential growth, the question arises of moral issues, security and legitimacy. In this review article, we aim to analyze the ethical and legal implications of using ChatGPT in Urology and explore potential solutions addressing these concerns.

RECENT FINDINGS

There are many potential applications of ChatGPT in urology, and the extent to which it might improve healthcare may cause a profound shift in the way we deliver our services to patients and the overall healthcare system. This encompasses diagnosis and treatment planning, clinical workflow, patient education, augmenting consultations, and urological research. The ethical and legal considerations include patient autonomy and informed consent, privacy and confidentiality, bias and fairness, human oversight and accountability, trust and transparency, liability and malpractice, intellectual property rights, and regulatory framework. The application of ChatGPT in urology has shown great potential to improve patient care and assist urologists in various aspects of clinical practice, research, and education. Complying with data security and privacy regulations, and ensuring human oversight and accountability are some potential solutions to these legal and ethical concerns. Overall, the benefits and risks of using ChatGPT in urology must be weighed carefully, and a cautious approach must be taken to ensure that its use aligns with human values and advances patient care ethically and responsibly.

National Multi-Institutional Validation of a Surgical Transfusion Risk Prediction Model

BACKGROUND

Accurate estimation of surgical transfusion risk is important for many aspects of surgical planning, yet few methods for estimating are available for estimating such risk. There is a need for reliable validated methods for transfusion risk stratification to support effective perioperative planning and resource stewardship.

STUDY DESIGN

This study was conducted using the American College of Surgeons NSQIP datafile from 2019. S-PATH performance was evaluated at each contributing hospital, with and without hospital-specific model tuning. Linear regression was used to assess the relationship between hospital characteristics and area under the receiver operating characteristic (AUROC) curve.

RESULTS

A total of 1,000,927 surgical cases from 414 hospitals were evaluated. Aggregate AUROC was 0.910 (95% CI 0.904 to 0.916) without model tuning and 0.925 (95% CI 0.919 to 0.931) with model tuning. AUROC varied across individual hospitals (median 0.900, interquartile range 0.849 to 0.944), but no statistically significant relationships were found between hospital-level characteristics studied and model AUROC.

CONCLUSIONS

S-PATH demonstrated excellent discriminative performance, although there was variation across hospitals that was not well-explained by hospital-level characteristics. These results highlight the S-PATH's viability as a generalizable surgical transfusion risk prediction tool.

Discriminative fusion of moments-aligned latent representation of multimodality medical data

Fusion of multimodal medical data provides multifaceted, disease-relevant information for diagnosis or prognosis prediction modeling. Traditional fusion strategies such as feature concatenation often fail to learn hidden complementary and discriminative manifestations from high-dimensional multimodal data. To this end, we proposed a methodology for the integration of multimodality medical data by matching their moments in a latent space, where the hidden, shared information of multimodal data is gradually learned by optimization with multiple feature collinearity and correlation constrains. We first obtained the multimodal hidden representations by learning mappings between the original domain and shared latent space. Within this shared space, we utilized several relational regularizations, including data attribute preservation, feature collinearity and feature-task correlation, to encourage learning of the underlying associations inherent in multimodal data. The fused multimodal latent features were finally fed to a logistic regression classifier for diagnostic prediction. Extensive evaluations on three independent clinical datasets have demonstrated the effectiveness of the proposed method in fusing multimodal data for medical prediction modeling.

Design of an interface to communicate artificial intelligence-based prognosis for patients with advanced solid tumors: a user-centered approach

OBJECTIVES

To design an interface to support communication of machine learning (ML)-based prognosis for patients with advanced solid tumors, incorporating oncologists' needs and feedback throughout design.

MATERIALS AND METHODS

Using an interdisciplinary user-centered design approach, we performed 5 rounds of iterative design to refine an interface, involving expert review based on usability heuristics, input from a color-blind adult, and 13 individual semi-structured interviews with oncologists. Individual interviews included patient vignettes and a series of interfaces populated with representative patient data and predicted survival for each treatment decision point when a new line of therapy (LoT) was being considered. Ongoing feedback informed design decisions, and directed qualitative content analysis of interview transcripts was used to evaluate usability and identify enhancement requirements.

RESULTS

Design processes resulted in an interface with 7 sections, each addressing user-focused questions, supporting oncologists to "tell a story" as they discuss prognosis during a clinical encounter. The iteratively enhanced interface both triggered and reflected design decisions relevant when attempting to communicate ML-based prognosis, and exposed misassumptions. Clinicians requested enhancements that emphasized interpretability over explainability. Qualitative findings confirmed that previously identified issues were resolved and clarified necessary enhancements (eg, use months not days) and concerns about usability and trust (eg, address LoT received elsewhere). Appropriate use should be in the context of a conversation with an oncologist.

CONCLUSION

User-centered design, ongoing clinical input, and a visualization to communicate ML-related outcomes are important elements for designing any decision support tool enabled by artificial intelligence, particularly when communicating prognosis risk.

Artificial intelligence applied to analyzes during the pandemic: COVID-19 beds occupancy in the state of Rio Grande do Norte, Brazil

The COVID-19 pandemic is already considered one of the biggest global health crises. In Rio Grande do Norte, a Brazilian state, the RegulaRN platform was the health information system used to regulate beds for patients with COVID-19. This article explored machine learning and deep learning techniques with RegulaRN data in order to identify the best models and parameters to predict the outcome of a hospitalized patient. A total of 25,366 bed regulations for COVID-19 patients were analyzed. The data analyzed comes from the RegulaRN Platform database from April 2020 to August 2022. From these data, the nine most pertinent characteristics were selected from the twenty available, and blank or inconclusive data were excluded. This was followed by the following steps: data pre-processing, database balancing, training, and test. The results showed better performance in terms of accuracy (84.01%), precision (79.57%), and F1-score (81.00%) for the Multilayer Perceptron model with Stochastic Gradient Descent optimizer. The best results for recall (84.67%), specificity (84.67%), and ROC-AUC (91.6%) were achieved by Root Mean Squared Propagation. This study compared different computational methods of machine and deep learning whose objective was to classify bed regulation data for patients with COVID-19 from the RegulaRN Platform. The results have made it possible to identify the best model to help health professionals during the process of regulating beds for patients with COVID-19. The scientific findings of this article demonstrate that the computational methods used applied through a digital health solution, can assist in the decision-making of medical regulators and government institutions in situations of public health crisis.

Demystification of artificial intelligence for respiratory clinicians managing patients with obstructive lung diseases

INTRODUCTION

Asthma and chronic obstructive pulmonary disease (COPD) are leading causes of morbidity and mortality worldwide. Despite all available diagnostics and treatments, these conditions pose a significant individual, economic and social burden. Artificial intelligence (AI) promises to support clinical decision-making processes by optimizing diagnosis and treatment strategies of these heterogeneous and complex chronic respiratory diseases. Its capabilities extend to predicting exacerbation risk, disease progression and mortality, providing healthcare professionals with valuable insights for more effective care. Nevertheless, the knowledge gap between respiratory clinicians and data scientists remains a major constraint for wide application of AI and may hinder future progress. This narrative review aims to bridge this gap and encourage AI deployment by explaining its methodology and added value in asthma and COPD diagnosis and treatment.

AREAS COVERED

This review offers an overview of the fundamental concepts of AI and machine learning, outlines the key steps in building a model, provides examples of their applicability in asthma and COPD care, and discusses barriers to their implementation.

EXPERT OPINION

Machine learning can advance our understanding of asthma and COPD, enabling personalized therapy and better outcomes. Further research and validation are needed to ensure the development of clinically meaningful and generalizable models.