Periventricular Leukomalacia (PVL) is a white matter lesion surrounding the lateral ventricles of the brain occurring in the prenatal period, associated with a disorder of movement and posture, known as bilateral cerebral palsy. Children with PVL and bilateral cerebral palsy have spared verbal abilities, as measured by verbal Intelligence Quotient (verbal IQ) tests, while non-verbal intelligence and especially visuo-perceptual and visuo-spatial abilities are impaired. In addition some studies underline the impact of PVL also on executive function, especially in terms of working memory and in the ability to inhibit distraction.

Working Memory is the ability to retain and manipulate information for brief periods of time. It is important in several complex cognitive functions, such as academic learning and in planning and organizing daily life activities. School-based activities, indeed, such as math and reading depend on a student's ability to pay attention to several instructions or information and to hold and integrate them in their mind.

Recent behavioural and neurofunctional studies describes the effect of an evidence-based and computer-based training on working memory, the Cogmed Working MemoryTraining. Functional MRI show increase in parietal and prefrontal activity after this training, while the behavioural data demonstrate the generalization of this effect also on cognitive functions not directly trained, as attention, inhibition, learning and non-verbal reasoning. Cogmed Working MemoryTraining (RoboMemo®, CogMed-Cognitive Medical Systems, Stockholm, Sweden) is an online treatment comprising a number of visuo-spatial and verbal exercises that vary automatically depending on the individual child's performance in any given task. The training period is intensive and includes 25 home session for five weeks, 30-45 minutes each day. A Cogmed-trained coach monitors training progress and participants' commitment daily.

Only one ongoing study has used the CogMed training in children with cerebral palsy, but without investigating the correlation between behavioural findings with neurofunctional data.

The aim of this study is to analyze the effect of the working memory training with CogMed on trained and not directly trained cognitive abilities and on the changes in cortical electrophysiological reorganization during the sleep after training. The sleep analysis will be focused in particular on the slow waves activity [frequency range of 1-4.5 Hz] and on the sleep spindle [frequency range of 12-14Hz], which reflect the depth of sleep and are related to memory processes, learning and brain plasticity.

The results of this project will shed light on the mechanisms of neuroplasticity, by enhancing knowledge on the neuropsychological effects of a specific working memory training and on the neurophysiological underpinnings of these behavioural effects in a clinical population of children with congenital brain lesions, as PVL.

The procedures implemented to register patients data will include the following steps:

• The quality assurance plan will provide: - to verify that all clinical fields, necessary to include and select patients, are available in the registry (e.g. type of cerebral palsy; MRI documentation, age); - to monitor that patients' contact information are available as well as testing materials and software; - to set team meetings to uniform recruitment procedure, data collection and registration.
• Data checks will be conducted by two or more researches from the team;
• Source data verification: both paper and electronic case report forms will be used and mutually checked.
• Data dictionary: Pre-post measures will be chosen by the team and described in a report form together with references and normative data.
• Procedure Standardization to address registry operations: Researches will co-work in patient recruitment and will attribute an identification number to each patient enrolled according to the inclusion and exclusion criteria evaluated by the neuropsychiatric team and described below. All researchers will be trained to register data in an internal database and to note the presence of adverse events. The researcher providing the training will assure comprehension and subscription of the written consent. Consent form will be discussed by the team in order to provide explicit and clear instructions on the training protocol to parents and patients.
• Calculation of sample size: Sample size, to specify the number of participants necessary to demonstrate the training effect, will be calculated from the training effect size between the Start index and the Max index provided by CogMed training. Based on a previous pilot study in our institute (effect size =.54) and on findings by Lundqvist and collaborators (2010) (effect size=.95), the effect size estimates for this study will be .60 and the sample size 20 children.
• Design: A stepped randomized clinical trial design will be chose to split the participants in two groups, matching for age and sex, in order to control test-retest effects. Both groups will receive neuropsychological and electrophysiological assessments at the same initial time point (T1). One group (Cluster B) will be start Cogmed training at home immediately after T1, while the second group (Cluster A) will start the training in the following time point (T2). When the training will be completed, both groups will be retested with neuropsychological and electrophysiological assessments (T2). After this time point, the Cluster A will start Cogmed training and will be re-tested with the same neuropsychological and the electrophysiological assessments at the finally time points (T3).
• The following time points will be provided:

Initial time point (T1) will correspond to pre-training assessment for both Clusters; Second time point (T2) will be provided after 6 or 7 weeks from T1 and will coincide with post-training assessment for the Cluster B and an other pre-training assessment for the Cluster A; Finally time point (T3), after 6 or 7 weeks from T2, will constitute the post-training assessment for Cluster A.

• Intervention program: Cogmed Working MemoryTraining (RoboMemo®, CogMed-Cognitive Medical Systems, Stockholm, Sweden) is an online treatment comprising a number of visuo-spatial and verbal exercises that vary automatically depending on the individual child's performance in any given task. The training period is intensive and includes 25 home session for five weeks, 30-45 minutes each day. A Cogmed-trained coach monitors training progress and participants' commitment daily for all the training period, corresponding to time period T1-T2 for Cluster B and T2-T3 for Cluster A.
• Plan for Missing data: Unavailable or not interpretable (significantly out of range) pre-post training data (behavioural or neurophysiological) will be considered as missing. The patients who will not complete the training will be excluded from the analysis.
• Plan for Statistical analysis: Univariate and Multivariate statistical analysis will be performed to compare neuropsychological performance and electrophysiological patterns at the different time points . Correlational analysis will be also performed to study the association between neuropsychological and electrophysiological data.

Inclusion Criteria:

• PVL documented at brain MRI performed after age 2 years
• level I to III at the Manual Ability Classification System (MACS)
• IIntelligence Quotient above 80 in either verbal or non-verbal domains, as assessed by Wechsler preschool and primary scale of intelligence (WPPSI-III), Wechsler intelligence scale of children (WISC-III or WISC-IV) in the last year prior to recruitment

Exclusion Criteria:

• absence of a psychiatric disorder diagnosis or sensory deficits that preclude testing
• absence of drug-resistant epilepsy