Though artificial intelligence (AI) has been the basis for many Hollywood films and sci-fi novels, it will take years or decades to advance AI up to the extent of imagination shown in them. AI has transformed various sectors such as education, business, manufacturing, law, and others. It has made leaps into medicine industry as well. Robotic science has made leaps and assisted doctors in performing surgeries. There is a possibility that AI will transform the medicine sector; however, it is too challenging and too far from now.
Jia Li, the head of research and development at Google Cloud, outlined few details on new research conducted on application of AI in radiology imaging at EmTech Digital, a conference organized by MIT Technology Review in San Francisco. The identification of disease in real clinical setting is possible with machine learning. The identification requires a rationale for diagnosis. Some of the medical experts opined that the entire areas of some of the medical works such as analysis of radiological images can be automated with AI.
The most popular machine learning technique, deep learning was used to determine abnormalities in chest X-rays. As Li and colleagues had a very less set of data to work on, they used another dataset to accelerate the learning process. They made sure their analysis showed the area of an image that was important to diagnose abnormality.
Researchers have been working with medical experts from Stanford. They highlighted that AI can automate only small part of work of radiologists. Medical experts need to understand the history of a patient, communicate a diagnosis, and identify the correct treatment. AI can play a critical role in this. Li outlined that doctors will not be replaced by AI in near future. She said, “We can assist [doctors] to make better judgments and make the process more efficient.”
The research in this field is critical as it addresses major challenges in application of AI in real-life medical situations. Most of the demonstrations in AI for medical sector have not considered the broader context. They involve large and perfectly annotated data sets. Google cloud offers a major platform for AI in medicine industry. Many experts believe delivering AI through cloud will be a major and lucrative trend in coming years.
Till now, the role of AI has been of assistant. It can assist doctors and healthcare professionals to gain insights on past cases of patients, reduce the dependence on them, and eliminate human errors. Doctors have nearly 80 patients per week. However, they may not provide the equal attention to every patient. The fatigue makes them prone to mistakes. However, AI can reduce their errors and improve quality of patients’ lives.
AI can be used to conduct clinical research trials. Application of advanced predictive analysis to determine candidates for trials can help medical professionals find out correct participants. According to McKinsey, there are many machine learning applications to help increase efficiency of clinical trial, including finding out best sample sizes for better efficiency and usage of electronic medical records to eliminate data errors. Though there are many applications of AI in medical sector, it is too far from replacing doctors. AI technology can assist them; but cannot replace.