Self-initiation of EEG-based communication in paralyzed patients.

OBJECTIVES: Severely paralyzed patients could learn to voluntarily generate slow cortical potential (SCP) shifts in their electroencephalogram and to use these signals to operate a communication device. To enhance the patients' autonomy, the present study describes the development of a permanently available communication system that can be turned on and off by locked-in patients without external assistance. A skill necessary for turning the system on is the ability to regulate one's slow potentials in the absence of continuous feedback. METHODS: A stepwise learning approach was employed to train two paralyzed patients to regulate their SCPs without continuous feedback. Elements of the original communication system were gradually removed and elements of the new stand-by mode were introduced. RESULTS: At the end of the learning procedure, both patients achieved correct response rates of above 84% in training sessions without continuous feedback. This skill enabled them to turn the communication device on and off without assistance from others. CONCLUSIONS: These findings suggest that severely paralyzed individuals can learn to operate an EEG-based communication device autonomously.

The thought translation device (TTD) for completely paralyzed patients.

The thought translation device trains locked-in patients to self-regulate slow cortical potentials (SCP's) of their electroencephalogram (EEG). After operant learning of SCP self-control, patients select letters, words or pictograms in a computerized language support program. Results of five respirated, locked-in-patients are described, demonstrating the usefulness of the thought translation device as an alternative communication channel in motivated totally paralyzed patients with amyotrophic lateral sclerosis.

A binary spelling interface with random errors.

An algorithm for design of a spelling interface based on a modified Huffman's algorithm is presented. This algorithm builds a full binary tree that allows to maximize an average probability to reach a leaf where a required character is located when a choice at each node is made with possible errors. A means to correct errors (a delete-function) and an optimization method to build this delete-function into the binary tree are also discussed. Such a spelling interface could be successfully applied to any menu-orientated alternative communication system when a user (typically, a patient with devastating neuromuscular handicap) is not able to express an intended single binary response, either through motor responses or by using of brain-computer interfaces, with an absolute reliability.

Geometric accuracy of field alignment in fractionated stereotactic conformal radiotherapy of brain tumors.

PURPOSE: To assess the accuracy of field alignment in patients undergoing three-dimensional (3D) conformal radiotherapy of brain tumors, and to evaluate the impact on the definition of planning target volume and control procedures. METHODS AND MATERIALS: Geometric accuracy was analyzed in 20 patients undergoing fractionated stereotactic conformal radiotherapy for brain tumors. Rigid head fixation was achieved by using cast material. Transfer of stereotactic coordinates was performed by an external positioning device. The accuracy during treatment planning was quantitatively assessed by using repeated computed tomography (CT) examinations in treatment position (reproducibility of isocenter). Linear discrepancies were measured between treatment plan and CT examination. In addition, for each patient, a series of 20 verifications were taken in orthogonal projections. Linear discrepancies were measured between first and all subsequent verifications (accuracy during treatment delivery). RESULTS: For the total group of patients, the distribution of deviations during treatment setup showed mean values between -0.3-1.2 mm, with standard deviations (SD) of 1.3-2.0 mm. During treatment delivery, the distribution of deviations revealed mean values between 0.7-0.8 mm, with SDs of 0.5-0.6 mm, respectively. For all patients, deviations for the transition to the treatment machine were similar to deviations during subsequent treatment delivery, with 95% of all absolute deviations between less than 2.8 and 4.6 mm. CONCLUSION: Random fluctuations of field displacements during treatment planning and delivery prevail. Therefore, our quantitative data should be considered when prescribing the safety margins of the planning target volume. Repeated CT examination are useful to detect operator errors and large random or systematic deviations before start of treatment. Control procedures during treatment delivery appear to be of limited importance. In addition, our findings should help to determine "cut-off points" for corrective actions in stereotactic conformal radiotherapy of brain tumors.

The thought translation device: a neurophysiological approach to communication in total motor paralysis.

A thought translation device (TTD) for brain-computer communication is described. Three patients diagnosed with amyotrophic lateral sclerosis (ALS), with total motor paralysis, were trained for several months. In order to enable such patients to communicate without any motor activity, a technique was developed where subjects learn to control their slow cortical potentials (SCP) in a 2-s rhythm, producing either cortical negativity or positivity according to the task requirement. SCP differences between a baseline interval and an active control interval are transformed into vertical or horizontal cursor movements on a computer screen. Learning SCP self regulation followed an operant-conditioning paradigm with individualized shaping procedures. After prolonged training over more than 100 sessions, all patients achieved self-control, leading to a 70-80% accuracy for two patients. The learned cortical skill enabled the patients to select letters or words in a language-supporting program (LSP) developed for inter-personal communication. The results demonstrate that the fast and stable SCP self-control can be achieved with operant training and without mediation of any muscle activity. The acquired skill allows communication even in total locked-in states.

Self-regulation of slow cortical potentials in completely paralyzed human patients.

The study was intended to answer the question whether self-regulation of brain activity can be operantly learnt when the brain is disconnected from motor periphery. Two neurological patients with nearly complete motor paralysis learned bi-directional control of their slow cortical potentials (SCP) at vertex. After 4-6 weeks training both patients could reliably differentiate between SCP shifts in a negative versus positive direction. With one patient, training has been continued for a subsequent 4 months, which resulted in precise self-control, i.e. the patient was able to produce positive SCP shifts on command with an accuracy of about 95%. This indicates that self-regulation of cortical excitability (as manifested in the SCP) does not require feedback loops from the periphery. Although we cannot rule out that healthy subjects may employ behavioral strategies such as muscle contractions or changes in breathing, obviously humans can also control their SCP without using these strategies.