. 2022 Oct 28;5:162. doi: 10.1038/s41746-022-00700-y

Table 5.

Characteristics of sampled grey literature.

Name of AI tool	Description	Country	Main user group	AI use category	AI type employed
Disease outbreak intelligence platform⁶⁵	Disease outbreak prediction and real-time disease risk assessment for COVID-19	ASEAN countries⁶⁶	Public health practitioners	Population health	Natural language processing; machine learning
Medical robot assistants⁶⁷	Service delivery robots in hospitals to improve patient care	China	Non-physician healthcare workers	Healthcare delivery	Robotics
CT and X-ray diagnostics^68,69	Computer assisted diagnostic (CAD) radiology tool for COVID-19 and other conditions to aid physicians	China	Physicians	Clinical decision support	Computer vision using deep learning
Close contact catcher⁷⁰	Population surveillance identifying close contact between individuals	China	Public health practitioners	Population health	Computer vision using deep learning
Deep learning - fractional flow reserve derived from coronary CT angiography⁷¹	Automated non-invasive physiological functional assessment of coronary arteries using coronary CT angiograms as an alternative to invasive coronary angiography	China	Physicians	Clinical decision support	Computer vision using deep learning
Diabetes risk prediction tool⁷²	Predictive tool for individual users to identify their risk of diabetes, and to promote early diagnosis	China	Individuals/patients	Population health	Machine learning
Robotic COVID-19 case monitoring⁷³	Automated screening calls and follow-up calls, performed by voice robots, in order to reduce call centre workloads	China	Non-physician healthcare workers	Frontline health worker virtual assistant	Robotics
Intelligent triage⁷⁴	Platform for patients to consult with ‘AI Doctor’ in order to facilitate access to relevant medical information and possible diagnoses, and to find a suitable doctor	China	Individuals/patients	Patient virtual assistant	Natural language processing; machine learning
Intelligent hospitals⁷⁵	Integration of multiple AI services, including speech input medical records, CAD systems, and AI-driven follow up	China	Physicians; non-physician healthcare workers	Frontline health worker virtual assistant; clinical decision support	Natural language processing; computer vision using deep learning; machine learning; expert systems
Autonomous drone delivery^33,76–78	Drone delivery of medical supplies and samples to hospitals	China, Dominican Republic, Rwanda, Madagascar, Malawi, Senegal	Healthcare providers	Healthcare delivery	Robotics
Diabetic retinopathy screening^79,80	Computer Assisted Diagnostic tool diagnosing diabetic retinopathy using hospital retinal imaging to ease physician workloads	China, India	Physicians; non-physician healthcare workers	Clinical decision support	Computer vision using deep learning
RAD-AID AI radiology⁸¹	Capacity building and implementation of CAD radiology tools to ease workloads of radiologists in low-resource settings	Multiple Latin American, Asian and African Countries	Physicians; non-physician healthcare workers	Clinical decision support	Computer vision using deep learning
Automated whole slide imaging and histological diagnostics⁸²	Automated whole slide imaging using conventional microscopes and smartphone, and AI histology diagnostic assistant, for diagnostics in low-resource settings	Mexico, Tanzania, India	Physicians; non-physician healthcare workers	Frontline health worker virtual assistant; clinical decision support	Computer vision using deep learning
Automated malaria diagnostics⁸³	Web-based platform for diagnosing malaria with thick blood smear images to strengthen laboratories	Uganda	Physicians; non-physician healthcare workers	Frontline health worker virtual assistant	Computer vision using deep learning
Health chatbots^84–87	App and web-based chatbots automating triage and self-directed care for patients	Brazil, China, India, Tanzania	Individuals/patients	Patient virtual assistant	Natural language processing; machine learning
Patient retention in HIV care^88,89	Predictive tool for healthcare providers to identify HIV patients at risk of being lost to follow up, promoting proactive intervention	South Africa	Healthcare providers	Population health	Machine learning
Tailored healthcare worker training⁹⁰	Identifying recurrent errors by health workers and proposing AI tailored training modules via digital platform	Burkina Faso	Non-physician healthcare workers	Personnel management	Machine learning
Perinatal asphyxia computer aided diagnostics⁹¹	CAD tool used by healthcare workers to detect early signs of perinatal asphyxia using recordings of newborn cry sounds	Nigeria	Non-physician healthcare workers	Frontline health worker virtual assistant	Computer audition using deep learning