Jobe Thomas H., Helgason Cathy M.
Department of Psychiatry, University of Illinois College of Medicine at Chicago, Chicago, USA
Neural Netw. 1998 Apr;11(3):549-555. doi: 10.1016/s0893-6080(97)00149-4.
Twentieth century medical science has embraced nineteenth century Boolean probability theory based upon two-valued Aristotelian logic. With the later addition of bit-based, von Neumann structured computational architectures, an epistemology based on randomness has led to a bivalent epidemiological methodology that dominates medical decision making. In contrast, fuzzy logic, based on twentieth century multi-valued logic, and computational structures that are content addressed and adaptively modified, has advanced a new scientific paradigm for the twenty-first century. Diseases such as stroke involve multiple concomitant causal factors that are difficult to represent using conventional statistical methods. We tested which paradigm best represented this complex multi-causal clinical phenomenon-stroke. We show that the fuzzy logic paradigm better represented clinical complexity in cerebrovascular disease than current probability theory based methodology. We believe this finding is generalizable to all of clinical science since multiple concomitant causal factors are involved in nearly all known pathological processes.
二十世纪医学科学采用了基于亚里士多德二值逻辑的十九世纪布尔概率论。随着后来基于比特的冯·诺依曼结构计算架构的加入，一种基于随机性的认识论导致了一种二值流行病学方法，这种方法主导着医疗决策。相比之下，基于二十世纪多值逻辑以及内容寻址和自适应修改的计算结构的模糊逻辑，为二十一世纪提出了一种新的科学范式。中风等疾病涉及多个伴随的因果因素，使用传统统计方法很难描述这些因素。我们测试了哪种范式最能代表这种复杂的多因果临床现象——中风。我们表明，与当前基于概率论的方法相比，模糊逻辑范式能更好地体现脑血管疾病的临床复杂性。我们相信这一发现可推广到所有临床科学领域，因为几乎所有已知的病理过程都涉及多个伴随的因果因素。
