This study investigates the memory capacities of Intensive Care Unit patients suffering from a Disorder of Consciousness (DoC) resulting from severe brain injury. Given the complexity of human memory, this study was designed to provide a nuanced profile of mnesic abilities by examining various types, stages, and temporal dimensions of memory. Specifically, the protocol targets progressively complex types of memory-perceptual, episodic, and autobiographical-alongside two key stages: encoding and retrieval. It also considers two temporal perspectives: (1) memory processes during the DoC (assessed indirectly through physiological markers) and (2) memory retrieval after the DoC, in patients who regain consciousness (assessed directly through subjective reports).
1. The investigator first aim to identify robust and specific physiological markers of memory processes occurring during the period of altered consciousness in the ICU. To this end, the investigator developed two auditory protocols using high-density electroencephalography (hdEEG) and electrocardiography (ECG) to assess the formation of new memories in non-communicating DoC patients (1a; anterograde memory). These protocols focus on two types of memory: perceptual memory, which enables the storage and interpretation of sensory information (e.g., recognition of the sounds of the artificial respirator or the voices of caregivers), and episodic memory, which allows individuals to recall specific events within their context (e.g., family visits, painful procedures) and serves as a foundation for autobiographical memory which supports the sense of identity. In practice, these protocols are conducted over two consecutive days in the ICU: Day 1 is dedicated to the putative encoding of new memories through exposure to specific auditory stimuli, while Day 2 assesses their long-term retrieval using 'old-new' paradigms, which contrast brain responses to stimuli presented on Day 1 ('old') with responses to novel stimuli introduced only on Day 2 ('new'). The investigator also developed a complementary protocol based on hdEEG, ECG and electrodermal and respiratory activity recordings, using auditory and olfactory stimuli, to investigate the retrieval of past autobiographical memories (1b; retrograde memory), that is memories from the patient's life prior to the onset of the disorder.
2. In parallel with this detection of implicit physiological correlates of memory, The investigator will also employ a direct approach in patients who have emerged from a DoC. Using semi-structured interviews, the investigator will gather explicit evidence of their subjective experiences and memories of the period of impaired consciousness. These may include internal conscious experiences (e.g., memories of dreams, hallucinations, emotions, or thoughts), external conscious experiences (e.g., family visits, painful procedures), as well as memories related to stimuli from experimental paradigms (e.g., auditory cues presented during physiological recordings) and memories of a musical memory probe played during the DoC. These interviews will be accompanied by neuropsychological tests assessing overall memory and cognitive functioning and the occurrence of PTSD.
The investigator will apply supervised machine learning techniques, including classification methods, to detect the presence or absence of memory processes at the individual patient level, based on predefined mnesic markers used as evaluation criteria. In parallel, The investigator will conduct second-order analyses, involving group-level comparisons between DoC patients and conscious ICU control patients, as well as between subgroups of DoC patients (e.g., UWS vs. MCS patients; focal vs. diffuse etiologies). These comparisons will be carried out using permutation statistics, statistical modeling, and/or parametric or non-parametric tests (e.g., Student's t-test, Wilcoxon-Mann-Whitney test).
