Background: The extent of psychiatric complications after head injury are not well recognized. We assessed the prevalence and treatment of psychiatric symptoms after traumatic intracranial hemorrhage (tICH) in a 10-years follow-up.

Methods: An observational, retrospective single-center cohort of consecutive tICH patients from HEAD Helsinki (Head trauma related health care Economics, Acute care and Development of long-term outcomes in Helsinki city region) study hospitalized at Helsinki University Hospital between 01 January and 31 December 2010. We reported primary outcomes as psychiatric symptoms and their subsequent treatment on secondary level psychiatry during the follow-up period 01 January 2010 and 31 December 2019. Logistic regression analysis was performed to study association between clinical aspects and primary outcomes.

Results: In our cohort of 385 patients (mean age 60.7 years, 66.5% male) with tICH, 66 (17.1%) had any psychiatric symptoms, and 48 (72.1%) of them experienced new symptoms (median time to onset 25 months), of which 26 (54.2%) were without any psychiatric history prior to tICH. A total of 35 patients (53.0%) received at least secondary psychiatric care, and 40 (60.6%) patients had initiated new psychotropics (median time to initiation 8 months). Compared to patients without psychiatric symptoms during the follow-up period, patients with psychiatric symptoms were younger (mean 49.1 years vs. 63.1 years, p < 0.001) and had less frequently larger (101-200 ml) intracranial hemorrhages (9.1% vs. 25.7%, p = 0.003). In multivariable analyses, younger age and epidural hemorrhage were independently associated with the development of psychiatric symptoms. Prior psychiatric medication and lower admission GCS score were associated with consequent psychiatric treatment.

Conclusions: Every sixth patient in our cohort had psychiatric symptoms after tICH. Over half of them received secondary level psychiatric care indicating the severity of psychiatric complications.

Psychological inflexibility is the central component in the Acceptance and Commitment Therapy (ACT) model of human suffering and mental health, with psychological flexibility representing positive health and wellbeing. Factors involved in psychological inflexibility/flexibility are proposed to relate to executive function including cognitive flexibility. Impairments in cognitive flexibility are commonly observed after traumatic brain injury (TBI) and the possible implications that this might have for psychological flexibility/inflexibility are unknown. This study aims to examine the relationship between these constructs in moderate to severe TBI (PTA > 24 hours). A total of 66 individuals, from the Liverpool Brain Injury Rehabilitation Unit, Sydney, Australia, were administered a battery of cognitive flexibility measures (Wisconsin Card Sort Test, Trail Making Test, Controlled Oral Word Associate Test, Stroop Test, The Alternate Uses Test ) in conjunction with their standard neuropsychological assessment. In addition to a general (Acceptance and Action Questionnaire-II; AAQ-II) and context specific (Acceptance and Action Questionnaire-Acquired Brain Injury; AAQ-ABI) measure of psychological inflexibility. Linear regression modelling was conducted to determine what measures of cognitive flexibility were predictive of psychological inflexibility. The context specific measure of psychological inflexibility model was significant (F(5,57) = 4.61, p < .001) and accounted for 29% of the variance in the AAQ-ABI. The Stroop colour-word interference score was the only significant measure of cognitive flexibility that predicted the AAQ-ABI (β =-.14, p< .001), a finding that remained after controlling for FSIQ and years of education. For the AAQ-II, the model was also significant (F(3, 63) = 3.10, p =.033) and accounted for 13.4% of the variance in the AAQ-II. Similarly, the Stroop colour-word interference score significantly predicted AAQ-II scores (β = -0.13, p =.024). This research provides support for the theory of executive function being an essential component of psychological flexibility and demonstrates empirical insight into these relationships. This may assist clinicians in understanding factors which may be perpetuating poor mental health in their clients with a TBI.

Objectives: Cognitive impairment is a common sequelae of moderate to severe traumatic brain injury (msTBI). The degree of persistence of cognitive impairment following msTBI, and the domains of cognitive function in which individuals continue to experience dysfunction, remain less clearly understood. A recent study utilizing the TRACK TBI dataset, a longitudinal, multisite, American study, identified that approximately 50% continue to experience cognitive deficits at 6 months following msTBI, and that there were diverse profiles of impairment, with the majority showing impairment to processing speed only. It is unclear how closely Canadian and American samples align. The present study was undertaken to examine the persistence and profiles of cognitive impairment in a large Canadian sample following msTBI.

Methods: The current study was a secondary analysis of n=197 patients with msTBI drawn from the Toronto Rehab TBI Recovery Study cohort, a multimodal longitudinal Canadian study comparable in assessment methods to TRACK TBI, including a cognitive assessment at approximately six months post-injury. A battery of standardized, performance-based neuropsychological tests assessed multiple cognitive domains, including memory, executive functioning (EF), and processing speed. Estimates of premorbid function were obtained using word-reading tests. Participants were assessed for impairment in each domain of cognitive functioning, operationalized as performance 1.5 standard deviations below the mean as per the TRACK TBI study. Following TRACK TBI, patients were classified into eight distinct impairment profiles: no cognitive impairment; memory impairment alone; processing speed impairment alone; EF impairment alone; memory and processing speed impairment; memory and EF impairment; processing speed and EF impairment; or impairment in all domains.

Results: Complete datasets were available for n=163 patients (mean age = 39.5, 75.5% percent male). In preliminary analyses, 30.1% of the sample demonstrated no evidence of cognitive impairment at approximately six months post-injury. Amongst the ~70% exhibiting persisting cognitive impairment, diverse cognitive profiles were observed, with impairment in memory and processing speed emerging as the most common profile (17.2%), followed by impairment in memory alone (16.0%).

Conclusions: With a prevalence of 70% impairment at roughly 6 months post-injury, 20% percent more individuals in this Canadian sample demonstrated persisting impairments on at least one cognitive outcome measure at 6 months post-injury as compared to the TRACK TBI study. The most prevalent profile of impairment was memory and processing speed, with a greater likelihood of memory impairment relative to isolated processing speed deficits. This stands in contrast to TRACK TBI, where patients showed isolated processing speed deficits as the most common profile. The paper will discuss possible explanations for the substantive differences between these Canadian and American cohorts.

Background: Neurophysiology is increasingly regaining centrality in diagnosing and managing Disorders of Consciousness (DOCs). Recently, the International Federation of Clinical Neurophysiology recommended a multimodal graded assessment approach for DOCs, integrating electroencephalography (EEG), Evoked Potentials (EPs), and Event-Related Potentials (ERPs). This study explores the potential of predicting diagnosis and etiology based on the persistence of afferent pathways identified through neurophysiological examinations.

Methods: We studied 146 DOC patients (52 females, mean age 50.3 ± 14.1 years). 82 were classified as Vegetative State/Unresponsive Wakefulness Syndrome (VS/UWS) and the remainder as Minimally Conscious State (MCS). Exclusion criteria included pre-existing severe psychiatric or neurological diseases. Etiologies were categorized as Anoxic Brain Injury (ABI), Traumatic Brain Injury (TBI), or OTHER causes. Multimodal EPs and auditory ERPs were used to assess auditory (AUD), somatosensory (SOM), and visual (VIS) responses. More specifically: auditory-Slow Vertex Potentials, Somatosensory EPs, and flash-Visual EPs. Endogenous cognitive responses (COG) were evaluated through MisMatch Negativity and P300. Preservation of each pathway was determined by the presence of a response from at least one side; COG was considered positive if either ERP component was present. Diagnostic group comparisons were made using Chi-Square and Wilcoxon tests (significance level: p < 0.01, Bonferroni corrected). We employed a Random Forest model with 5-fold cross-validation to predict diagnosis (VS/UWS vs. MCS) and etiology (ABI, TBI, OTHER), using AUD, SOM, VIS, and COG as variables. Accuracy and variable importance (via Gini's index) were computed, and splitting rules were extracted for interpretability.

Results: The diagnostic groups were homogeneous in sex and age but differed in etiology: ABI (51.2% VS/UWS, 17.18% MCS), TBI (26.8% VS/UWS, 31.25% MCS), and OTHER (22.0% VS/UWS, 50.8% MCS) (p = 3.46*10⁻⁴). Neurophysiological responses showed AUD in 45.1% VS/UWS, and 71.9% MCS (p = 0.024); VIS in 24.4% VS/UWS and 65.6% MCS (p = 5.7*10⁻³); and COG in 17.0% VS/UWS and 66.1% MCS (p = 1.48*10⁻⁵). Diagnosis was predicted with an accuracy of 67.78% ± 6.97% (chance level: 56.16%). VIS and AUD were the splitting variables for the diagnosis, the patient was classified as MCS in presence of both, while absence of either led to a VS/UWS classification. Etiology prediction had an accuracy of 52.51% ± 9.98% (chance level: 29.40%). For aetiology, the splitting variables were VIS and COG: their absence was associated with ABI; the only presence of VIS to TBI; presence of both, to OTHER.

Conclusion: Hence, the persistence of afferent pathways, and especially the visual one, holds significant informative power. Indeed, the VIS pathway conservation is a strong sign of MCS, while its loss is mainly associated to ABI, correlating with poorer outcomes. Our findings suggest that visual potentials can play a crucial role in guiding clinical decisions regarding DOC prognosis, diagnosis, and management.

Introduction: Severe acquired brain injuries can result in disorders of consciousness (DOC) like unresponsive wakefulness syndrome (UWS) and minimally conscious state (MCS), where patients show limited responses to stimuli. Bedside neurobehavioral assessments often lead to misdiagnosis, prompting the use of multimodal methods like electroencephalography (EEG) for better accuracy. Research in healthy individuals links EEG features with emotion recognition during audiovisual stimulation. However, studies on DOC patients are limited. Some evidence suggests that personalized stimuli, like a patient's name or familiar voices, can trigger identifiable responses. This study explores differences in EEG responses of healthy individuals, patients in MCS, and patients with UWS when exposed to emotionally-valenced videos featuring relatives or strangers.

Methods: A convenience sample of 17 healthy subjects, 7 patients in MCS, and 10 patients with UWS was recruited. The experimental design included one-minute baseline period, followed by 12 randomized 55-second video clips (six featuring relatives, six featuring strangers), each followed by a 50-second rest period. Videos, including audio, were presented using the HTC Vive Pro Eye 2 virtual reality headset. EEG was recorded using a 32-channel Brain Products LiveAmp amplifier. Standard preprocessing techniques were applied, and the signals were segmented into 5-second windows. Common spatial patterns (CSP) were extracted for different frequency bands and a Support Vector Machine (SVM) was trained using these data to classify the two types of stimuli for each subject. A correlation analysis was also conducted to explore the relationship between patient scores on the Coma Recovery Scale-Revised (CRS-R) and the accuracy of their trained classifiers.

Results: In healthy subjects and patients with DOC, the SVM models trained using the gamma and high-gamma frequency bands achieved the best performance. The average classification accuracy in distinguishing between videos of relatives and strangers aligned with the clinical condition, with healthy subjects achieving an accuracy of 0.79, patients in MCS scoring 0.62, and patients with UWS reaching 0.51. More importantly, the trained models successfully classified both stimuli for all healthy subjects, and for 4 out of the 17 patients with DOC (UWS=1, MCS=3). Interestingly, 3 of them (all in MCS) emerged from MCS 6 months after. There was a strong correlation (r=0.78) between the CRS-R scores of patients and the classification accuracy of their EEG signals.

Conclusion: These findings suggest that EEG-based machine learning models may be effective in detecting responses to emotional stimuli and could offer valuable insights into the emotional processing abilities, which can improve the assessment of the level of consciousness and even detect covert cognition, in patients with DOC.

Acknowledgments: This work was supported by Conselleria d’Innovació, Universitats, Ciència i Societat Digital of Generalitat Valenciana (CIDEXG/2022/15), Ministerio de Ciencia e Innovación (PID2022-141498OA-I00), Fundació la Marató de la TV3 (60/2023), and European Commission (HORIZON-MSCA-2022-PF-01-10110814).

Background: Music enhances arousal and attention when compared to white noise or disliked music or when compared to a control non-musical auditory stimulus, suggesting a potential impact of music therapy in patients with disorders of consciousness (DoC). Recently, binaural beats (perceptual phenomenon occurring when presented two tones with different frequency) also seems to influence arousal and attention. Such stimulation has nevertheless never been directly compared in DoCs.
Objective: In this multi-centric project, the short-term effects of both preferred music and binaural beats (with Beta waves, 20-30hz) on DOC patient’s responsiveness were compared.

Methods: A within subject design was used in one session and 2 conditions (music and beats) presented after a baseline in pseudo-randomized order between subjects. Stimuli were presented using earbuds at the patients’ bedside. Frequencies of behavioral responses were recorded at baseline and after each condition using selected items of the Coma Recovery Scale-Revised (CRS-R).

Participants: 77 patients (mean age: 51±20 years, 41 males) classified as being in a DoC (i.e., 53 minimally conscious state - MCS or 24 vegetative state - VS) based on the CRS-R were included in this study. Time since injury ranged from 6 months to 8 years. Etiologies included traumatic brain injuries (TBI; n = 27), stroke (n=17) and anoxia (n = 27).

Results: Using Wilcoxon tests, when each condition (music or beats) was compared to baseline, higher frequencies/responses were observed for arousal (w=-1.63; p=.05), visual tracking (w=-2.05; p=.02), movement to command (body-related; w=-1.95; p=.02), only for preferred music. No significant increase was observed for binaural beats. The effect for visual tracking was bigger for TBI/stroke than for anoxia (p=0.02). Finally, when comparing preferred music to baseline, we found a misdiagnosis of 21% (5/24) in VS and 29% (4/14) in MCS- (mostly, due to response to command), suggesting that using preferred music could help detecting signs of consciousness when diagnosing DoC.

Conclusion: According to our results, preferred music seems to promote arousal but also (and potentially, in turn) improves the detection of important signs of consciousness (such as willful body-related movements and visual tracking). Future studies should investigate the biomechanisms related to such improvements in patients with DoCs. Binaural beats (Beta waves) does not seem to lead to similar effects. However, using other frequencies (Theta or Alpha waves) might lead to different results and should be studied too.

The Brief Evaluation of Receptive Aphasia (BERA) is a novel tool designed to assess receptive phonological, semantic, and morphosyntactic abilities in patients with severe brain damage, based on visual fixation of a target image presented next to a specific distractor. This study aims to provide validation data regarding the administration of the BERA in post-comatose patients with severe brain injury. The BERA and Simplified Evaluation of CONsciousness Disorders (SECONDs) were administered to 30 French-speaking adult patients with preserved visual fixation or pursuit (21 patients in a minimally conscious state and 9 patients emerging from it). The BERA assessments were performed on two consecutive days, by two blinded raters, to determine the validity and reliability of the tool. Very strong and strong correlations were found between versions 1 and 2 (internal validity: ρ = 0.83) and between BERA scores and a language score extracted from the command-following and communication items of the SECONDs assessment (concurrent validity: ρ = 0.73). Excellent intra- and inter-rater reliability was also observed (ρ = 0.86 and ICC = 0.996). Results show that the BERA is a fast tool (µ = 13.9 min; σ = 4.7 min), which is appropriate to refine language profiles in minimally conscious and emerging patients. The BERA scores may indeed indicate selective receptive difficulties in phonological, semantic, and morphosyntactic domains, which could help guide speech-language therapy. These data finally suggest that the BERA may complement other scales such as the SECONDs for assessing and diagnosing post-comatose patients.