Table 1.
Inclusion and exclusion criteria.
Fig 1.
PRISMA 2020 flow diagram of study selection.
Flow diagram showing the identification, screening, eligibility assessment, and inclusion of studies. A total of 800 records were identified, including 725 from databases and 75 from other methods. After removal of 302 records before screening, 498 records were screened, 126 full-text reports were assessed for eligibility, and 45 studies were included in the review. Abbreviations: AI, artificial intelligence; IEEE, Institute of Electrical and Electronics Engineers..
Fig 2.
Publication year and geographical distribution.
Figure (a) presents the distribution of included studies by year of publication. Figure (b) illustrates the geographical distribution of the study locations, with the bar heights representing the number of publications originating from each country. Abbreviations: USA, United States of America; UK, United Kingdom.
Table 2.
Study design/validation categories and settings of included studies.
Table 3.
Technical pathways, characteristics, and validation context / study maturity of AI-based hand hygiene compliance monitoring.
Table 4.
Framework for interpreting reported performance metrics according to validation context and evidence maturity.
Fig 3.
General framework of AI-based hand hygiene monitoring systems in healthcare settings.
General framework of AI-based hand hygiene monitoring systems in healthcare settings, showing data inputs, analytic methods, monitoring outputs, and implementation considerations.Abbreviations: AI, artificial intelligence; RGB, red-green-blue; RGB-D, red-green-blue plus depth; IMU, inertial measurement unit; RF, radio frequency; IoT, Internet of Things.
Fig 4.
Comparative technical landscape of AI hand hygiene monitoring approaches represented in the reviewed studies.
Comparison of the main AI-based hand hygiene monitoring approaches reported in the reviewed studies, including computer vision-based, wearable sensing, radar/radio frequency-based, and IoT-integrated systems. Abbreviations: AI, artificial intelligence; RGB, red-green-blue; RGB-D, red-green-blue plus depth; CNN, convolutional neural network; 3D CNN, three-dimensional convolutional neural network; IMU, inertial measurement unit; ML, machine learning; LSTM, long short-term memory; RF, radio frequency; mmWave, millimeter wave; IoT, Internet of Things.f Things.
Table 5.
Cross-cutting implementation dimensions of AI-based hand hygiene monitoring systems reported in included studies.