Analytic Strategies and Cognitive Task Design to Study Neuropsychiatric Disorder

Longenecker等人2010年研究了精神分裂症患者和健康对照组工作清单回忆的潜在机制。我们小组使用了加州词汇学习测试(CVLT)，这是一种广泛使用的神经心理学词表学习任务，以检查回忆策略。CVLT列表由16个单词组成，包括4个语义类别--香料、工具、水果和衣服--每个类别包含4个单词。这些清单先是口头陈述，然后由受试者自由回忆。此前有报道称，回忆单词最多的受试者使用语义技术，而回忆能力较差的受试者通常使用连续技术。在这项研究中，我们开发了新的测量方法，独立于回忆性能来比较不同试验的策略，并确定诊断是否预测回忆策略。 最近的一项研究表明，精神分裂症患者在语言和视觉学习任务上表现不佳，但他们未受影响的兄弟姐妹只有在语言学习方面表现不佳，这表明语言处理可能代表了一种认知表型。使用CVLT检查言语记忆障碍的性质表明，该策略可以预测患者的表现。由于精神分裂症患者有广泛的缺陷，通过情景言语记忆任务来评估，我们使用语言障碍的受试者来检查在自由列表回忆过程中使用的策略的机制。使用从第一次和第五次试验中回忆的单词、两次试验之间的改善以及所有五次试验的总体回忆来衡量CVLT的表现。使用语义和序列聚类分数来分析所使用的助记策略。语义聚类被定义为对绝对相似的词的一个接一个的回忆，序列聚类被定义为按呈现顺序对词的顺序回忆。通过排列测试，计算和分析了新的召回率和转移概率。我们的受试者是根据CVLT在五次试验中的表现来选择的。从评估的450个受试者数据中，测试样本包括健康对照组(n=100)和精神分裂症患者(n=100)中表现最好的和最差的各50人。 结果：我们预测绩效水平将成为战略诊断的一个指标。同样，在以前报道的文献中，我们预计CVLT评分将预测成绩水平，而不考虑组别。一般来说，记忆最多的人使用语义聚类技术，而记忆较差的人使用连续技术。这一点在使用语义聚类技术的健康对照CVLT第五次试验中得到了证明，在该试验中，回忆的单词数量最多。患者的回忆模式和策略与控制组相似，表现水平相似，与诊断无关。表现不佳的人更有可能回忆起列表中的前两个和最后四个单词，而表现好的人在所有五个试验中都更多地参与了基于语义的转换，而表现不佳的人则没有。对我们的CVLT数据的检查表明，表现良好的患者和对照组都使用了连续策略和语义策略。这些策略似乎随着尝试学习而改变。由于CVLT标准评分方法没有捕捉到这些信息，我们开发了一种不同的方法来分析回忆，该方法考虑了与学习过程中表现变化相关的同时回忆策略。这一方法提供了一个框架，可以用来重新研究言语回忆不良的潜在认知机制，以及由大脑功能网络和遗传功能多态决定的神经生物学。使用这一工具，我们可以了解疾病和基因对言语回忆过程的调节过程中的个体差异。此外，我们开发了一种新的方法来计算单词列表自由回忆中的策略，该方法通过数据驱动的方法计算所有学习试验中回忆顺序之间的相似性。这种方法通过计算每对可能的单词的概率来提供更具体的信息，而不是计算每个试验的概率或所有五个试验的平均值。我们的方法评估单个单词的影响以及预先确定的类别和词序的交互作用。这种方法的另一个好处是，它可以被修改以适应额外的措施，例如通过回忆不相邻的单词或影响回忆顺序来激活语义。 结论：我们开发的新方法使用标准的CVLT分析公式来识别和解决问题。回忆和转移概率的测量表明，在自由回忆列表中，回忆的单词数量预测了所采用的策略，即语义策略导致更多的单词回忆。我们还发现，高表现的患者依赖于语义策略，这表明语义混乱可能不是精神分裂症患者的特征。总而言之，认知策略必须独立于整体表现来考虑，然后才能将糟糕的表现归因于学习过程的退化。我们的发现表明，偏离公认的、广泛使用的评分技术的重要性，特别是在精神分裂症患者的临床人群中，对他们来说，情景记忆缺陷是一个特征。此外，这种新的方法可以应用于其他词表范例和临床人群，并具有探索特定的词对词的关系及其如何影响单词回忆策略的敏感性。

Longenecker, et al 2010 investigated mechanisms underlying work list recall in patients with schizophrenia and healthy controls. Our group used the California Verbal Learning Test (CVLT), a widely used neuropsychological word list learning task, to examine recall strategies. The CVLT list consist of 16 words comprised of 4 semantic categories -spices, tools, fruits and clothes- each containing 4 words. The list were presented orally then freely recalled by the subject. It has been previously reported that subjects who recall the most words employ semantic techniques, as opposed to poor recall subjects who generally use serial techniques. In this study, we developed novel measures, independent of recall performance to compare strategies across trials and identify whether diagnosis predicts recall strategy. A recent study showed that patients with schizophrenia performed poorly on verbal and visual learning tasks, but their unaffected sibling showed poor performance only on measures of verbal learning, which suggest that verbal processing may represent a cognitive phenotype. Examining the nature of verbal memory impairment using the CVLT showed that strategy predicted performance of patients. Since patients with schizophrenia have a wide range of deficits, as assessed by episodic verbal memory tasks, we used subjects with verbal impairments to examine mechanisms of strategies used during free list recall. CVLT performance was measured using the words recalled from the first and fifth trials, the improvement between the two trials, and overall recall across all five trials. Mnemonic strategies employed were assayed using semantic and serial clustering scores. Semantic clustering was defined as recall of categorically similar words, one after another, and serial clustering was defined as sequential recall of words according to order of presentation. Novel measures of recall and transition probability were calculated and analyzed by permutation tests. Our subjects were selected based upon CVLT performance over five trials. From the 450 subject data evaluated, the test sample included 50 highest and 50 lowest performers among healthy controls (n=100) and patients with schizophrenia (n=100). Results: We predicted that performance level would be an indicator of strategy as diagnosis. Likewise, in previously reported literature, we expected CVLT scores would predict performance level regardless of group. Generally, people who recall the most words used the semantic clustering technique, and those who had poor recall employed the serial technique. This was shown in the healthy control CVLT fifth trial where the most number of words were recalled, used the semantic clustering technique. Recall patterns and strategies of patients resembled those from the control group with similar performance levels, regardless of diagnosis. Low performers were more likely to recall the first two and last four words from the list, while high performers increased involvement of semantic-based transition across all five trials, while low performers did not. Examination of our CVLT data showed that high performers, both patients and controls, employed both serial and semantic strategies. The strategies seemed to change with trial learning. Because this information was not captured on the CVLT standard scoring method, we developed a different approach to analyze recall which takes into account simultaneous recall strategies that relate the changes in performance during learning. This approach provides a framework with which to re-examine both the underlying cognitive mechanisms of poor verbal recall, but also the neurobiology as determined by both functional brain networks and genetic functional polymorphisms. Using this tool we may gain understanding of individual differences in modulation of the verbal recall process by both illness and genes. In addition, we developed a novel method to calculate strategy in free recall of word lists via a data-driven approach of computing similarities between recall orders across all learning trials. This approach offers more specific information by calculating probabilities for every possible pair of words rather than for each trial or an average value across all five trials. Our method evaluates the effects of individual words as well as the interaction of predetermined categories and word order. Another benefit of this approach is that it can be modified to adjust for additional measures such as semantic activation through recall of nonadjacent words or effects to recall order. Conclusion: The novel approach we developed identified and addresses issues using the standard CVLT analytic formulas. Recall and transition probability measures showed that in free recall of a list the number of words recalled is predictive of the strategy employed whereby semantic strategy resulted in a higher number of words recalled. We also showed that high performing patients depend on semantic strategies, which suggest that semantic disorganization may not be characteristic of patients with schizophrenia. In summary, cognitive strategy must be considered independently of overall performance before attributing poor performance to degraded learning processes. Our findings have shown the importance of departing from the accepted, widely used scoring techniques, particularly in clinical populations of patients with schizophrenia for whom episodic memory deficits are a characteristic feature. In addition, this novel approach can be applied to other word list paradigms and clinical populations and has the sensitivity to explore specific word-to-word relationships and how they impact verbal recall strategy.