Principal Investigator: Marzia De Lucia
What is the extent of preserved neural functions during coma ? What is their relation with patients’ level of consciousness ? The interest in these research questions is twofold; on one side understanding brain function in coma can help the development of quantitative markers for predicting a given patient’s prognosis as early as possible; on the other side, in basic neurosciences, comatose patients offers a virtual setting for the investigation of brain functions without consciousness.
These two main research interests develop within the more general framework of the study of the neural basis of human auditory perception and cognition in healthy and clinical populations. To answer these research questions we combine functional neuroimaging with methods of machine learning and computational modeling.
• Early prediction of comatose patients’ outcome : the evolution over days of auditory discrimination in comatose patients can predict their chance of surviving.
Decoding results and outcome prediction of post-anoxic comatose patients treated with therapeutic hypothermis. A. Average decoding performance (standard error indicated) for the control subjects (blue) and a group of comatose patients (red) recorded during hypothermia (TH) and normothermia (NT). Decoding was measured as the area under the curve (AUC) obtained when discriminating responses to standard versus deviant sounds in classic mismatch negativity paradigm. The mean decoding performance typically decreases from therapeutic hypothermia to normothermic conditions in non-survivors. (B) The difference in decoding performance from therapeutic hypothermia to normothermic conditions, evaluated independently for each patient, was used for predicting the patients’ outcome. All non-awakening patients showed a drop in their performance or no improvement (squares). An improvement on the average decoding performance from therapeutic hypothermia to normothermic conditions was only observed in survivors (rhombi), resulting in a positive predictive value of 100%.
• The role of subjects’ awareness in auditory categorization and association: reliable discrimination between complex sound categories can be achieved without awareness.
A. Average auditory evoked potentials of one comatose patient recorded during hypothermia and sedation and one control subject in response to human and animal vocalizations.
B. Results of the topographhic consistency test for the patient and control subject revealing long-lasting time-periods of evoked responses (i.e.,periods of 1-p > 0.999). Periods of differential activity in response to animal and human vocalizations estimated by a single-trial decoding analysis are highlighted in light blue.
These results show the first evidence in humans that reliable discrimination between complex sound categories can be achieved without awareness, and even in challenging clinical conditions such as those experienced by comatose patients during the acute phase.
• Multivariate analysis for decoding and prediction
Decoding algorithms are applied to neuroimaging datasets with the aim of extracting spatio-temporal patterns in the data that carry relevant information about differences in the response to experimental conditions. The application of these methods in fMRI and EEG studies has highlighted the role of widespread spatio-temporal patterns of activity encoding relevant sensory information, shifting the focus from the description of specialized brain regions to a network perspective. I am particularly interested in the development of such techniques and optimization of their applications to electrophysiological datasets for cognitive and clinical studies. One main challenge of building decoding algorithms for these types of applications is their neurophysiologic interpretability. These decoding methods also provide the basis for establishing a direct link between individual electrophysiological responses and behavior, by discovering which neurophysiologic signals best predict the model parameters estimated from subjects’ performance and by comparing neurophysiologically plausible models linking sensory processes to perception.
Tzovara A, Rossetti AO, Juan E, Suys T, Viceic D, Rusca M, Oddo M, De Lucia M (2016) Prediction of awakening from hypothermic post anoxic coma based on auditory discrimination. Annals Neurology 2016;79:748–757.
Tzovara A, Simonin A, Oddo M, Rossetti AO, De Lucia M (2015) Neural detection of complex sound sequences in the absence of consciousness. Brain, 138(Pt 5):1160-6.
Cossy N*, Tzovara A*, Simonin A, Rossetti AO, De Lucia M (2014) Robust discrimination between EEG responses to categories of environmental sounds in early coma. Frontiers in Psychology, 5, 155 *equal contribution
Tzovara A, Rossetti AO, Spierer L, Grivel J, Murray MM, Oddo M, De Lucia M (2013) Progression of auditory discrimination based on neural decoding predicts awakening from coma. Brain, 136(Pt 1):81-9.
Tzovara A., Murray M.M., Plomp G., HerzogM.H., Michel C.M. De Lucia M (2012) Decoding stimulus-related information from single-trial EEG responses based on voltage topographies. Pattern Recognition. 45( 6): 2109–2122.
Tzovara A, Murray MM, Bourdaud N, Chavarriaga R, del R. Millán J, De Lucia M (2012) The timing of exploratory decision-making revealed by single-trial topographic EEG analyses. Neuroimage. 60(4):1959-69.
De Lucia M, Tzovara A, Bernasconi F, Spierer L, Murray MM (2012) Auditory perceptual decision-making based on semantic categorization of environmental sounds. Neuroimage. 60(3):1704-15.
Bernasconi F*, De Lucia M*, Tzovara A*, Manual A, Murray MM, Spierer L. (2011) Noise in brain activity engenders perception and influences discrimination sensitivity. Journal of Neuroscience 31(49):17971-17981. *equal contribution
De Lucia M*, Constantinescu I*, Sterpenich V, Pourtois G, Seeck M, Schwartz S. (2011) Decoding sequence learning from single-trial intracranial EEG in humans. Plos One 6(12):e28630. *equal contribution
De Lucia M, Cocchi L., Martuzzi R., Meuli R.A., Clarke S., Murray M.M. (2010). Perceptual and semantic contributions to repetition priming of environmental sounds Cerebral Cortex 20(7): 1676-84.
Eureka Eurostars a principal investigator for three years -starting date july 2015- (E9361-ComAlert). "COnsciousness Monitoring, Assessment, prediction and Life activities with EEG in Real Time"; (510K chf)
Grant of the Service Projets et Organisation Stratégiques of the University Hospital of Lausanne for two years as principal investigator -starting date October 2012-. "Validation d'un test pour prédire les chances de réveil de patients dans le coma basé sur le eeg"; (213K chf)
Swiss National Foundation grant (CR32I3_143780) as co-applicant for three years -starting date March 2013-. “Early electrophysiological correlates of brain injury and outcome in comatose patients after cardiac arrest”. (281K chf)