In: Computer Science
Jiang, F., Jiang, Y., Zhi, H., Dong, Y., Li, H., Ma, S., ... & Wang, Y. (2017). Artificial intelligence in healthcare: past, present and future. Stroke and vascular neurology, 2(4), 230-243.
1.The artificial neural network (ANN)-a machine learning technique inspired by the human neuronal synapse system-was introduced in the 1950s. However, the ANN was previously limited in its ability to solve actual problems, due to the vanishing gradient and overfitting problems with training of deep architecture, lack of computing power, and primarily the absence of sufficient data to train the computer system. Interest in this concept has lately resurfaced, due to the availability of big data, enhanced computing power with the current graphics processing units, and novel algorithms to train the deep neural network. Recent studies on this technology suggest its potentially to perform better than humans in some visual and auditory recognition tasks, which may portend its applications in medicine and healthcare, especially in medical imaging, in the foreseeable future. This review article offers perspectives on the history, development, and applications of deep learning technology, particularly regarding its applications in medical imaging.
2.This review firstly offers a basic overview of deep learning particularly for image data analysis to give knowledge to nuclear medicine physicians and researchers. Because of the unique characteristics and distinctive aims of various types of molecular imaging, deep learning applications can be different from other fields. In this context, the review deals with current perspectives of deep learning in molecular imaging particularly in terms of the development of biomarkers. Finally, future challenges of deep learning application for molecular imaging and future roles of experts in molecular imaging are discussed.
3.Deep learning is not only a new wave of research, development and application in the field of medical imaging (and other imaging fields such as homeland security screening) but also a paradigm shift. This could be the magic wand to achieving optimal results cost-effectively, especially from huge and compromised data, as well as for problems that are nonlinear, nonconvex, and overly complex. However, my perspective of deep imaging could be overly optimistic, and must be balanced by controversies, potential difficulties and justified concerns. It has taken decades for the neural network approach to outperform the human in some recognition tasks, and hence the general success of deep learning for image reconstruction must rely on some new twists that take time to develop and realize. The big data, learning architecture, performance evaluation, and potential translation may all demand significant efforts.
4.Artificial intelligence (AI) aims to mimic human cognitive functions. It is bringing a paradigm shift to healthcare, powered by increasing availability of healthcare data and rapid progress of analytics techniques. We survey the current status of AI applications in healthcare and discuss its future. AI can be applied to various types of healthcare data (structured and unstructured). Popular AI techniques include machine learning methods for structured data, such as the classical support vector machine and neural network, and the modern deep learning, as well as natural language processing for unstructured data. Major disease areas that use AI tools include cancer, neurology and cardiology. We then review in more details the AI applications in stroke, in the three major areas of early detection and diagnosis, treatment, as well as outcome prediction and prognosis evaluation. It has been concluded with discussion about pioneer AI systems, such as IBM Watson, and hurdles for real-life deployment of AI.
5.
Artificial intelligence (AI) is already widely employed in various medical roles, and ongoing technological advances are encouraging more widespread use of AI in imaging. This is partly driven by the recognition of the significant frequency and clinical impact of human errors in radiology reporting, and the promise that AI can help improve the reliability as well the efficiency of imaging interpretation. AI in imaging was first envisioned in the 1960s, but initial attempts were limited by the technology of the day. It was the introduction of artificial neural networks and AI based computer aided detection (CAD) software in the 1980s that marked the advent of widespread integration of AI within radiology reporting. CAD is now routinely used in mammography, with consistent evidence of equivalent or improved lesion detection, with small increases in recall rates. Significant false positive rates remain a limitation for CAD, although these have markedly improved in the last decade. Other challenges include the difficulty clinicians encounter in trying to understand the reasoning of an AI system, which may limit their confidence in its advice, and a question mark hangs over who should be liable if CAD makes an error. The future integration of CAD with PACS promises the development of more comprehensively intelligent systems that can identify multiple, challenging diagnoses, and a move towards more individualised patient outcome predictions based upon AI analysis.