Human patient models of memory Flashcards
HM in detail
Henry Molaison, underwent resection of MTL in 1953 for epilepsy. Studied originally by Scoville and Milner, 1957. Brain was fixed in gelatine and sliced over 53 hours using a microtome. They aimed for anterior 2-2.5 cm of hippocampus and subicular cortex, but they also got central perirhinal, some of entorhinal, most of amygdala except CeN. Seizure frequency was reduced, but severe anterograde episodic and semantic loss, retrograde episodic loss. Intact implicit memory (mirror writing), IQ and STM - although perhaps the stimuli used were too familiar.
K.F.
parieto-occipital damage, intact LTM but impaired STM. This contradicts the Atkinson and Shiffrin multi-store model, there must be some way for info to go straight into LTM. Non-declarative memory normal.
Impaired more when presented with verbal stimuli than visual (i.e. selective impairment of phonological loop?)
Parkinson’s patients
deterioration of the nigrostriatal pathway. Intact declarative, impaired non-declarative. Can test using e.g. a weather-predicting game, with four cards. Controls and MTL amnesics will improve at predicting over several trials, PD patients will not. But amnesics will not remember declarative stuff e.g. what the screen looked like, whereas PD patients will.
Vargha-Khadem’s children
1997 - Beth, Jon and Kate. Perinatal hypoxia caused early hippocampal damage. All three had intact semantic memory, and performed well at school, but impaired episodic memory. Jon performed normally on recognition tests (normally designed to test episodic memory), but showed the early electrical activity associated with ‘familiarity’ and not that associated with ‘recall’, showing that recognition tests can miss episodic altogether. Non-declarative memory normal.
R.B.
Cardiac arrest following open heart surgery led to localised and selective anoxia and death of all CA1 hippocampal cells, while CA3 and DG were untouched. He showed a marked anterograde amnesia, worse for episodic than semantic (though not as bad as HM), and only a little retrograde amnesia. No intellectual impairment, non-declarative memory normal. Monkeys with global hypoxia show CA1 damage, and are impaired at DNMTS task.
N.A.
Stabbed through the right nostril with a miniature fencing foil in 1960. Damage to mediodorsal nucleus of the thalamus, mammillary bodies, mammillothalamic tract. Anterograde amnesia, but no damage to higher cognitive functions Non-declarative memory normal.
Korsakoff’s
Damage to medial thalamus, mammillary bodies, fornix. Retrograde and anterograde amnesia. Some reduction in cognitive function (resembling frontal patients), perhaps associated with frontal lobe atrophy sometimes seen. Johnson et al 1985 - Korsakoff’s patients can acquire affective reactions (presented with pairs of photos of men + biographical info depicting one as the ‘good guy’ and the other as the ‘bad guy’. When tested, patients couldn’t remember any of the info, but said they preferred the good guy.) Also remember Jimmy G. Non-declarative memory normal.
Semantic dementia
Warrington (1975) was the first to describe selective impairment of semantic memory, but Hodges et al 1992 were the ones to propose calling it ‘semantic dementia’
- damage to perirhinal cortex, lateral temporal lobe and temporal pole. Early on, there’s intact episodic but impaired semantic LTM. Recognition memory is normal, until the images to be recognised are perceptually different. This led to the multiple input theory, and fucked with Tulving’s SPI model. BUT perhaps recognition memory is not a true test of episodic memory - see Jon.
Delay-Brion/Papez circuit
1937 - Papez suggested a circuit involving mammillary bodies, hippocampus, anterior thalamic nuclei and cingulate cortex was behind emotion.
1969 - Delay and Brion suggested that damage to a circuit involving mammillary bodies, hippocampus, anterior thalamic nuclei would be enough to precipitate anterograde amnesia.
