Takahashi Ryohei, Kajikawa Yuya
Department of Innovation Science, School of Environment and Society, Tokyo Institute of Technology, 3-3-6 Shibaura, Minato-ku, Tokyo, Japan.
Department of Innovation Science, School of Environment and Society, Tokyo Institute of Technology, 3-3-6 Shibaura, Minato-ku, Tokyo, Japan.
Int J Med Inform. 2017 May;101:58-67. doi: 10.1016/j.ijmedinf.2017.02.004. Epub 2017 Feb 17.
Computer-aided diagnosis (CAD) has been a promising area of research over the last two decades. However, CAD is a very complicated subject because it involves a number of medicine and engineering-related fields. To develop a research overview of CAD, we conducted a literature survey with bibliometric analysis, which we report here. Our study determined that CAD research has been classified and categorized according to disease type and imaging modality. This classification began with the CAD of mammograms and eventually progressed to that of brain disease. Furthermore, based on our results, we discuss future directions and opportunities for CAD research. First, in contrast to the typical hypothetical approach, the data-driven approach has shown promise. Second, the normalization of the test datasets and an evaluation method is necessary when adopting an algorithm and a system. Third, we discuss opportunities for the co-evolution of CAD research and imaging instruments-for example, the CAD of bones and pancreatic cancer. Fourth, the potential of synergy with CAD and clinical decision support systems is also discussed.
在过去二十年中，计算机辅助诊断（CAD）一直是一个很有前景的研究领域。然而，CAD是一个非常复杂的课题，因为它涉及许多医学和工程相关领域。为了对CAD进行研究综述，我们进行了文献计量分析的文献调查，并在此报告结果。我们的研究确定，CAD研究已根据疾病类型和成像方式进行了分类。这种分类始于乳腺钼靶的CAD，最终发展到脑部疾病的CAD。此外，基于我们的研究结果，我们讨论了CAD研究的未来方向和机遇。首先，与典型的假设方法不同，数据驱动方法已显示出前景。其次，在采用算法和系统时，测试数据集的标准化和评估方法是必要的。第三，我们讨论了CAD研究与成像仪器共同发展的机遇，例如骨骼和胰腺癌的CAD。第四，还讨论了CAD与临床决策支持系统协同作用的潜力。
