Artificial intelligence clinical trials and critical appraisal: a necessity

Review Health Care Sciences & Services

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

Anne A. H. de Hond et al.

Summary: This article collects guidance on the development, evaluation, and implementation of AI-based prediction models in healthcare, finding that there is more detailed guidance for early stages such as data preparation, AIPM development, and AIPM validation, while later stages such as software development, impact assessment, and implementation have received less attention.

NPJ DIGITAL MEDICINE (2022)

添加到收藏夹

Article Health Care Sciences & Services

Effectiveness of automated alerting system compared to usual care for the management of sepsis

Zhongheng Zhang et al.

Summary: This study demonstrates the beneficial effect of using an automated alerting system in the management of sepsis, especially for patients in the emergency department and ward. Machine learning-based prediction methods are more effective in reducing mortality compared to rule-based methods.

NPJ DIGITAL MEDICINE (2022)

添加到收藏夹

Review Medicine, General & Internal

Randomized Clinical Trials of Machine Learning Interventions in Health Care A Systematic Review

Deborah Plana et al.

Summary: Despite the potential of machine learning in improving patient care, there are barriers to its clinical adoption. This systematic review examines the design, reporting standards, risk of bias, and inclusivity of randomized clinical trials (RCTs) for medical machine learning interventions. The findings reveal a lack of adherence to reporting standards, high variability in risk of bias, and a low representation of participants from underrepresented minority groups in published RCTs.

JAMA NETWORK OPEN (2022)

添加到收藏夹

Article Medical Informatics

A framework for validating AI in precision medicine: considerations from the European ITFoC consortium

Rosy Tsopra et al.

Summary: AI technologies have the potential to revolutionize healthcare systems, but their implementation is limited due to a lack of reliable validation procedures. The European ITFoC consortium has designed a framework for the clinical validation of AI technologies for predicting treatment response in oncology, based on seven key steps including intended use, target population, evaluation timing, datasets, data safety, performance metrics, and explainability. This framework forms the basis of a validation platform for assessing and comparing AI algorithms for predicting treatment response in triple-negative breast cancer with real-world data.

BMC MEDICAL INFORMATICS AND DECISION MAKING (2021)

添加到收藏夹

Article Medicine, General & Internal

Designing and undertaking randomised implementation trials: guide for researchers

Luke Wolfenden et al.

Summary: Implementation science focuses on improving the quality of randomized trials of implementation strategies and addressing current limitations. This study aims to provide guidance on the development, conduct, and reporting of these trials to advance rigorous conduct and reporting in the field.

BMJ-BRITISH MEDICAL JOURNAL (2021)

添加到收藏夹

Article Health Care Sciences & Services

What is needed to mainstream artificial intelligence in health care?

Ian A. Scott et al.

Summary: Although AI has become mainstream in many industries, its adoption in healthcare lags behind. AI has the potential to optimize care processes, maximize workforce capacity, and improve patient outcomes. The current obstacles to wider AI adoption in healthcare and the pre-requisites for its successful development need to be defined.

AUSTRALIAN HEALTH REVIEW (2021)

添加到收藏夹

Article Biochemistry & Molecular Biology