Amnesia involves profound memory loss, which is usually caused by physical brain injury, diseases that attack the brain, or the effects of drugs that alter brain function. Amnesia can also have psychological causes, such as traumatic and emotional events.

There are many kinds of amnesia that create different symptoms, sometimes symptoms lasting temporary (e.g. transient global amnesia) otherwise they would last permanently. In most cases, amnesia occurs mainly because of damage to the brain that result in hindered brain function, especially in terms of functionality related to long and short-term memory. There are several types of amnesia but mainly if individuals with amnesia exhibit a severe impairment of new learning in other word inability to formation of new enduring declarative memories, it is anterograde amnesia. As opposite form of amnesia, on the other hand, if individuals with amnesia cannot remember previously familiar information or the episodic memory preceding brain damage it is retrograde amnesia.
Amnesia can affect anyone, occur at any age. Moreover due to its variety of causes, the processing of the amnesia is also variable. Although individuals with amnesia can exhibit preserved intelligence, reasoning abilities, attention and remote memory, it is clear that the memory loss is the main issue of amnesia. People with amnesia cannot keep their job due to their memory loss and impaired memory capacity after the onset of brain damage, family of amnesic people also suffer from the dependence of their life activates which is especially related with memory function. The initial questions of this study are whether amnesia is healed? How is amnesia treated? If there is efficient rehabilitation strategy for amnesic patients’ memory loss so that they can keep their job or activity for living by themselves, it would be a great hope for all amnesic individuals as well as their family. Through literature review, particularly by case study review, this study explores some rehabilitation interventions for individuals with amnesia.

Methods
Case study I
Although learning new information is impaired in severe anterograde amnesia, it seems that most amnesic individuals with damage to the medial temporal lobe (the hippocampal region, entorhinal, perirhinal, and parahippocampal cortices) have some capacity to learn new information. Bayley and Squire (2002) tested new semantic learning in an amnesic patient, E.P., who has extensive damage limited to the medial temporal lobe. At the time of study, E.P. was 79-years old who developed profound anterograde and retrograde amnesia after brain damage. He scored 98 on the Wechsler Adult Intelligence Scale-III (WAIS-III) as full scare IQ, also scored 94, 57, 82, 61, and 56 on the attention-concentration, verbal memory, visual memory, general memory, and delayed memory subtest on the Wechsler Memory Scale-Revised (WMS-R). E.P. was presented in 24 study sessions and tested on his recognition and recall performance for factual information (novel tree-word sentences) in three tests (cued-recall test, two-alternative test, forced-choice test). In order to compare E.P.’s performances with controls, E.P. and controls followed a ‘study only’ or ‘errorless learning’ procedure. Although E.P. did much poorer performances when compared with controls, he demonstrated considerable learning of new information. E.P improved gradually and unmistakably across the sessions. However, results presented that E.P.’s learning was depended on non-declarative memory, was relatively inflexible, and was not accompanied by conscious knowledge. Authors suggested that factual information can be acquired not only by declarative memory but also by non-declarative memory.

Case study II
Although the study of Baley et al. (2002) suggested that ‘errorless learning’ techniques helped a patient with amnesia to acquire at least some form of factual information (three-word sentences) by non-declarative memory, the learning was rigidly organized and unavailable as conscious knowledge. Patient E.P. was able to learn many three-word facts when a related cue was presented as same as the one actually used in training but he could not respond correctly when a cue was substituted to different one. In E.P.’s case, even small changes in questions that used to probe the memory dropped performances significantly. Stark and colleagues (2005) explained this problem as due to artificially eliminated variability during errorless learning procedures. With a modification of the paradigm of Baley et al.’ study, Stark and his colleagues reported a case study of patient T.E., whose amnesia was caused by damaged to the hippocampus and extrahippocampal areas in the medical temporal lobe, to explore whether the addition of variance into the training within the errorless learning would affect the ability to generalize performance to new versions of the trained sentences. T.E. was 68 years old at the time of study who was severe amnesic in everyday life. He did not remember the day or time, did not recall some family members’ name, wholly dependent on his wife in daily life, unable to maintain a job either.

On the Wechsler Memory Scale-III (WMS-III), he scored 88, 58, 56, and 49 on the working memory, auditory delayed memory, visual delayed memory, and general memory subtests. During eight test sessions, T.E. was trained on new semantic sentences that was an extension of the same set of three-word sentences used by Bayley et al. (2002) both with and without variance and was tested on repeated and related novel sentences. T.E. was finely able to learned new semantic information in both conditions but in generalization of this ability to the not studied items, T.E. performed significantly better following variance training than no variance training that was only used in Bayley et al. study. Results of this study suggested that the variance condition in the training set not only helped T.E. could recall and recognition new semantic information but also could generalized this information in a flexible way.