Integrating Artificial Intelligence into the Practice of Transfusion Medicine

摘要

OVERVIEW OF ARTIFICIAL INTELLIGENCE Artificial intelligence (AI) refers to computer systems capable of performing complex tasks that historically only a human could do, such as reasoning, making decisions, or solving-problems. John McCarthy first described the term as early as 1956, as the science and engineering of making intelligent machines. AI and machine learning (AIML), an offshoot of computer science, are rapidly evolving in various fields of healthcare. Today, AIML is used to describe the use of computers and technology to simulate intelligent behavior and critical thinking akin to a human being.[1] AI is an umbrella term that encompasses a wide variety of technologies, including machine learning, deep learning, and natural language processing. Machine learning may use a flowchart-based algorithmic approach or a database approach. Google Translate uses deep learning algorithms to translate text from one language to another. ChatGPT uses large language models to generate text in response to questions or comments.[2] AI has the potential to be applied in almost every field of medicine, be it in radiology for interpreting X-rays or computed tomography scans or in robotic surgery for performing complex surgeries guided by a surgeon. AI is also used to automate appointments, online check-in of patients at outpatient departments, digitization of medical records, and so on.[3] A computer may use a flowchart-based approach as with patient history taking, to narrow down the diagnosis or possibilities using a large amount of data, fed into machine-based cloud networks (neural networks). Artificial neural networks are organized in several layers to imitate how the human brain interprets and draws conclusions from information. AI may also use a database approach as with radiology films which are based on deep learning or pattern recognition. AI applications have transformed health care the world over, particularly in the field of medical imaging and diagnostics.[3,4] Unknown to us, AI has been in use in our field of immunnohematology and transfusion medicine. This editorial is aimed to explore areas where AIML can be employed in transfusion medicine. ARTIFICIAL INTELLIGENCE IN TRANSFUSION MEDICINE –SUBSPECIALTIES Artificial intelligence in donor registration and counseling The donor selection deferral criteria are so vast and ever-changing that it is nearly impossible to remember all of them. Human intelligence is limited and often needs assistance from books, documents, or Google searches to accept or defer a donor. This is where AI is helpful. The amount of information that an AI tool can remember is many times more than the human mind. Once the selection deferral criteria are fed in or access to appropriate documents specified, AI-operated systems can answer any donor query in a standardized manner that too with scientific reference. Donor counseling involves complex topics such as exploring donor’s sexual health, his habits such as intravenous drug abuse, and answering doubts or queries that the donor may have. However, donor queries are fairly repetitive such as – “I have diabetes or high blood pressure can I donate?” and “I am on medication for hypothyroidism can I donate?” These largely repetitive queries can probably be better handled by the AI system once programmed. For instance, “Is the process of blood donation painful?” can be answered differently by different blood centers, by different staff in the same blood center, or by the same staff on different occasions. This is because human intelligence is dependent on our memory, mood, energy levels, and time at our disposal. All these can be overcome and our replies can be standardized using an AI tool. While answers to all questions may not be available on Day1, as the tool learns more and more or gets access to good reliable documents, it can possibly give out standardized replies with no dependence on memory or mood. Such tools can be linked with PubMed to source authe