Jonathan Guo và Bin Li
Artificial intelligence (AI) is a rapidly developing computer technology that has begun to be widely used in the healthcare field to improve professional skills and clinical efficiency. In developing countries around the world, there remains an inequality between urban and rural healthcare services, with a shortage of doctors being a major cause.
Several studies have shown that the application of computer-aided or AI-assisted medical techniques can improve healthcare outcomes in rural areas of developing countries.
AI in Clinical Decision-Making
With the rapid development of technology, the amount of new information from research is increasing and growing faster. Currently, the amount of medical information doubles every three years. It is estimated that a doctor would need to read 29 hours a day to keep up with all medical information. Furthermore, big data, including data from electronic health records (EHRs), “omic” data—genomics, metabolomics, proteomics, and sociological and lifestyle data—is useless without comprehensive analysis.
The only solution to access and utilize the enormous amount of information in the healthcare field is to use artificial intelligence (AI) technology. IBM Watson is one of the leading AI-powered healthcare support systems, helping doctors make effective decisions. With its machine learning and natural language processing capabilities, this system helps doctors review patients’ electronic health records and update treatment guidelines from medical research.
A double-blind study compared the decisions of a panel of cancer experts with the decisions of the Watson cancer system’s AI. The results showed that 90% of the system’s recommendations were consistent with those of the expert panel, but the system only took 40 seconds to complete the process.
AI in electronic health records
In 2009, the US Department of Health began encouraging the adoption of EHRs. However, implementation has faced numerous challenges, with key barriers including low satisfaction among healthcare staff with EHR systems, particularly issues with interoperability and latency in the practice of electronic health records, especially in primary care settings.
Currently, EHRs have become one of the most time-consuming tasks for healthcare staff in healthcare settings. AI technology promises to help healthcare providers collect, store, reformat, and track clinical data, as well as personalized plans and assessments.
AI in Diagnosis
Diagnostic errors are a serious threat to the quality and safety of healthcare. In the US, the estimated rate of outpatient diagnostic errors is 5.08%, equivalent to 12 million people per year. Approximately half of these errors are potentially harmful. AI technology has been used to improve diagnostic quality, particularly in radiology.
AI based on a dataset of 129,450 clinical images for diagnosing skin diseases has demonstrated that the system can classify skin cancers at a level comparable to dermatologists. An algorithm based on magnetic resonance imaging of heart movement allows for accurate prediction of outcomes in patients with pulmonary hypertension; an automated rhythm classification method in continuous electrocardiogram (ECG) analysis in critically ill patients. Another study examined promising results using AI in stroke imaging and suggested that AI technology could play a significant role in stroke patient management.
AI in medical robots
Applications of medical AI technology also include robots and assistive medical devices. For example, robotic telephones can facilitate communication between patients and healthcare professionals.
Assisted walking devices can help coordinate walking, standing, or sitting; and animal-like robots can communicate and identify patients. Robots can also be used in surgery as surgical assistants; the da Vinci surgical system is one of the most commonly used robotic surgical systems.
AI in Personalized Medicine
Personalized medicine is a new healthcare model where treatment and disease prevention are based on individual conditions, including genetic information, psychosocial characteristics, environment, and lifestyle. All this information generates a massive amount of data that can only be analyzed and integrated by AI technology.
AI in Healthcare Management
Integrating AI into healthcare management systems can help identify unnecessary diagnoses and treatments. Therefore, AI in healthcare is not only focused on classic patient-provider interactions but can also be used in managing healthcare systems for large-scale organizations. Systems can track healthcare spending, cost recovery, and response to treatment, thus improving population health and quality of care while reducing costs.
AI technology in healthcare has the potential to improve access to and quality of healthcare, and according to the authors, it is fully feasible to implement in developing countries. To achieve this goal, the authors propose developing a multi-tiered AI healthcare service network in developing countries, including: a primary AI healthcare application system (first level), a regional AI healthcare application system (middle level), and a national AI healthcare development center (highest level).
