Tendolkar Indira, Arnold Jennifer, Petersson Karl Magnus, Weis Susanne, van Eijndhoven Philip, Buitelaar Jan, Fernández Guillén
Department of Psychiatry, University Medical Center Nijmegen, 6500-HB Nijmegen, The Netherlands.
Brain Res. 2007 Apr 20;1142:159-68. doi: 10.1016/j.brainres.2007.01.040. Epub 2007 Jan 20.
The medial temporal lobe (MTL) is crucial for declarative memory formation, but the function of its subcomponents in associative memory formation remains controversial. Most functional imaging studies on this topic are based on a stepwise approach comparing a condition with and one without associative encoding. Extending this approach we applied additionally a parametric analysis by varying the amount of associative memory formation. We found a hippocampal subsequent memory effect of almost similar magnitude regardless of the amount of associations formed. By contrast, subsequent memory effects in rhinal and parahippocampal cortices were parametrically and positively modulated by the amount of associations formed. Our results indicate that the parahippocampal region supports associative memory formation as tested here and the hippocampus adds a general mnemonic operation. This pattern of results might suggest a new interpretation. Instead of having either a fixed division of labor between the hippocampus (associative memory formation) and the rhinal cortex (non-associative memory formation) or a functionally unitary MTL system, in which all substructures are contributing to memory formation in a similar way, we propose that the location where associations are formed within the MTL depends on the kind of associations bound: If visual single-dimension associations, as used here, can already be integrated within the parahippocampal region, the hippocampus might add a general purpose mnemonic operation only. In contrast, if associations have to be formed across widely distributed neocortical representations, the hippocampus may provide a binding operation in order to establish a coherent memory.
内侧颞叶（MTL）对陈述性记忆的形成至关重要，但其各子成分在联想记忆形成中的功能仍存在争议。关于这一主题的大多数功能成像研究都基于一种逐步的方法，即比较有联想编码和无联想编码的两种情况。在此基础上，我们进一步采用参数分析方法，通过改变联想记忆形成的量来进行研究。我们发现，无论形成的联想数量多少，海马体的后续记忆效应大小几乎相似。相比之下，鼻周皮质和海马旁皮质的后续记忆效应则受到形成的联想数量的参数性正向调节。我们的结果表明，海马旁区域支持此处所测试的联想记忆形成，而海马体则增加了一种一般的记忆操作。这种结果模式可能暗示了一种新的解释。我们提出，内侧颞叶内联想形成的位置取决于所绑定的联想类型，而不是在海马体（联想记忆形成）和鼻周皮质（非联想记忆形成）之间存在固定的分工，也不是一个功能统一的内侧颞叶系统，即所有子结构都以类似的方式对记忆形成做出贡献：如果像这里使用的视觉单维联想已经可以在海马旁区域内整合，那么海马体可能仅增加一种通用的记忆操作。相反，如果联想必须在广泛分布的新皮质表征之间形成，那么海马体可能会提供一种绑定操作以建立连贯的记忆。
