Volume 5, Issue 6, 2023

Our Scientific Editor discusses the current use of artificial intelligence in writing academic papers and reports the updated guidelines for Brain Communications on the use of this tool in scientific writing.

This scientific commentary relates to ‘Clinical course of pathologically confirmed corticobasal degeneration and corticobasal syndrome’, by Aiba et al. (https://doi.org/10.1093/braincomms/fcad296).

This scientific commentary refers to ‘Functional hub disruption emphasizes consciousness recovery in severe traumatic brain injury’, by Oujamaa et al. (https://doi.org/10.1093/braincomms/fcad319).

Neuner et al. analysed intraoperative multichannel EEGs. With increasing exposure to anaesthesia, EEG microstate quantifiers and state space descriptors (parameters of global brain activity) showed different trajectories in non-suppression and suppression EEG and in patients with or without subsequent postoperative delirium. Their role in predicting postoperative delirium warrants further research.

Neurofilament light chain (NFL) is a valuable biomarker for neuroaxonal injury. Ongphichetmetha et al. assessed its diagnostic potential and created models to distinguish causes of altered consciousness. Cerebrospinal fluid NFL showed moderate accuracy, and the model using clinical data matched neurologists’ diagnoses. These findings underscore its use in consciousness diagnostics.

Llibre-Guerra et al. show that dominantly inherited and sporadic early-onset Alzheimer’s disease differ in baseline clinical presentation and cognitive profiles. However, longitudinal cognitive decline and biomarker profiles are similar across both groups. These findings suggest shared pathways between dominantly inherited and sporadic early-onset Alzheimer’s disease.

Nabizadeh et al. reported that the soluble Triggering Receptor Expressed on Myeloid Cell 2 in CSF is associated with slower tau aggregate accumulation. Also, it might be a protective factor against tau spreading through inter-connected regions in brain tissue.

Liu et al. report that individuals clinically diagnosed with dementia due to Alzheimer’s disease showed a positive association between heart rate variability and risk of developing agitation. Specifically, compared to non-agitated individuals, individuals with agitation showed reduced longitudinal decline in heart rate variability over 22–26 years of follow-up.

Using an artificial-intelligence-based volumetry in essential tremor and tremor-dominant Parkinson’s disease, Purrer et al. investigated structural brain features in different tremor types. They found similar volumetric patterns in both diseases but significant deviations from a normative database, particularly within the basal-ganglia-thalamocortical circuitry, suggesting that both diseases share structural changes, indicative of neurodegenerative mechanisms.

Brandel et al. administered a translated version of the Medical Research Council Prion Disease Scale by telephone and observed a similar decline in patients with sporadic Creutzfeldt–Jakob disease in France and the UK. This result validating the scale opens the possibility of using it as an evaluation criterion for international therapeutic trials.

Parent et al. have used multimodal MRI to measure tissue damage within white matter hyperintensities (radiological indicators of vascular dysfunction prevalent in aging and Alzheimer’s disease) and concluded that T2* is a microstructural measure sensitive to clinically-meaningful variations in white matter hyperintensity tissue damage.

Kikumoto et al. detected a characteristic radiological feature of pontine autosomal dominant angiopathy and leukoencephalopathy, which enables more accurate evaluations of patients with undetermined juvenile cerebral vascular disorder. This feature is multiple oval small infarctions in the pons resembling raisin bread, for which they coined the name ‘raisin bread sign’.

Hurley et al. report that exposure to bacterial membrane components (lipopolysaccharide and lipopeptides) and bacterial by-products (short-chain fatty acids) increases α-synuclein protein levels in mouse secretin tumour cell line 1 enteroendocrine cells. Enteroendocrine cells could therefore represent an initial peripheral site of Parkinson’s disease pathology under specific gut microbial conditions.

Using lesion network mapping, Zarifkar et al. report that eye-opening apraxia is uniquely associated with the dorsal anterior and posterior insula. The study suggests a complex interplay between salience detection and motor execution in the initiation of appropriate eyelid-opening responses, with neural substrates distinct from other common stroke-related syndromes.

Shamas et al. investigated inter-ictal EEG spikes in temporal lobe epilepsy. An increased number of brain regions with a high rate of spikes that lead spike propagation in and especially outside the seizure onset zone predicts seizure outcome indicates the spike network could inform the extent of the seizure network.

Papadelis et al. report abnormal electrophysiology (i.e. reduced evoked beta and gamma cortical activity) in individuals with succinic semialdehyde dehydrogenase deficiency, a disorder of γ-aminobutyric acid (GABA) degradation. These findings possibly represent elevated brain GABA’s negative effects on rhythmic synchronization and provide the grounds for the development of non-invasive biomarkers of GABA.

Owen et al. demonstrate that band power abnormalities relative to health derived from interictal magnetoencephalography data overlap with the interictal EEG electrode placement in good outcome patients only. Furthermore, a combination of both magnetoencephalography and intracranial EEG abnormalities is predictive of surgical outcome.

Nosková et al. report a girl with focal epilepsy and behavioural disorder carrying homozygous variant p.S905F in ADAM22 that is frequent in Roma population. The variant causes insufficient interaction of ADAM22 to a family of membrane-associated guanylate kinases and leads to milder phenotype compared to previously reported cases.

Kundu et al. introduces a novel approach for noninvasive imaging of human cortical lamina, combining diffusion–relaxation multidimensional MRI and unsupervised machine learning. It encodes microstructure and local chemical composition simultaneously, uncovering layer-specific multidimensional signatures that provide insights into intra-cortical organization. The approach offers potential for studying human neurodevelopment and related disorders.

Boito et al. investigated the integrity of brain white matter microstructure of patients previously hospitalized for COVID-19 and suffering from persisting symptoms at follow-up. Advanced diffusion MRI acquisition and analysis revealed widespread microstructural alterations of the white matter, compatible with vasogenic oedema, demyelination and axonal damage.

Held, Bauer et al. evaluated white matter microstructural topology and network connectivity in Rasmussen's encephalitis. Their results demonstrate a prominent involvement of interhemispheric fibres in disease progression and remodelling at both microstructural and network levels, leading to the conclusion that Rasmussen's encephalitis should be considered a whole-brain network disorder.

Boot et al. report a less-efficient functional brain network in young patients with post-stroke epilepsy compared with those without. These alterations, particularly involving contralesional brain regions, may contribute to cognitive impairment, suggesting a disrupted functional network as a potential pathophysiological mechanism for cognitive impairment in post-stroke epilepsy.

In this exploratory research, Ciliento et al., using high-density EEG overnight, detected EEG spikes in 46% of patients with amnestic mild cognitive impairment without any history of seizures versus 10% in controls. Complemented by MRI and neuropsychology, no significant morphological and cognitive differences emerged between patients with and without spikes.

Maher et al. report a promising pilot study of a deep learning algorithm that can stratify controls from patients with focal epilepsy and patient subgroups. Their interpretable model achieved an overall best accuracy of 72.92% when classifying patients versus controls and 67.86% when classifying patients according to seizure type.

Audrain et al. report the detection of seeding-competent transactive response DNA-binding protein of 43 kDa (TDP-43) in CSF of amyotrophic lateral sclerosis patients using seed amplification assay. These seeding species can be neutralized by an antibody targeting the C-terminal of TDP-43 providing evidence for an immunotherapy approach to target amyotrophic lateral sclerosis.

In this work, Miller et al. analyse cortical morphometry in both children and adults with persistent developmental stuttering and identify differences in cortical thickness in several key speech planning regions, particularly in children, with significant correlations between cortical thickness and stuttering severity in these brain regions.

Dessert et al. provide the first method for optimized stereo-EEG planning that considers spatial recording sensitivity based on patient-specific biophysical modelling. Optimized electrode configurations may improve both the safety and the efficacy of epileptogenic zone localization compared with manual planning.

In this retrospective analysis, Levy et al. show that localization of deep brain stimulation microelectrodes implanted in the subthalamic nucleus of patients with Parkinson’s disease is correlated with change in motor and non-motor symptoms. The analysis suggests an algorithm to optimize intraoperative decision-making of deep brain stimulation contact localization to improve symptoms.

Kremer et al. report that deep brain stimulation in the connectivity-derived 7-Tesla MRI subthalamic nucleus motor segment produces superior clinical outcomes, compared with stimulation in the non-motor-connected subthalamic nucleus. Multi- and single-unit activities, derived from microelectrode recordings, were not able to reliably distinguish the motor segment within the subthalamic nucleus.

Faigle et al. introduces the Brainbox, a waterproof and fully MRI-compatible container with an imprinted 3D coordinate system enabling whole-brain ex vivo imaging for efficient MRI-histopathology correlation. The Brainbox has been validated on eight human brains. Brainboxes are available upon request from our institution.

Ding et al. report that the amount and speed of producing discourse-relevant content in post-stroke aphasia were associated with damage to anterior dorsal white matter pathways. Therefore, damage to these pathways may be a useful biomarker for impaired production of relevant information and informs development of neurorehabilitation strategies.

There is concern that elite rugby participation adversely influences brain health. Parker et al. show that in-game head injury may lead to localized cortical swelling that correlates with a marker of astrocytic activation, whereas the same region exhibit reductions in cortical thickness that may occur in the longer term.

Heng and Rittman highlight a lack of ethnic diversity and disproportionately larger Caucasian populations in open dementia data sets, despite increasing recognition of the impact of ethnic differences in epidemiological measures and diagnostic biomarkers. It is hoped that this will help to raise awareness and broaden inclusion within the field.

In this study, Aumont et al. investigated the hippocampus, a brain region tied to memory. They found that visual memory was associated with the right hippocampus, while verbal memory was associated with both sides in older adults. Additionally, the hippocampus was smaller in the presence of local Alzheimer’s disease signature.

Joho et al. reported that long-term treatment with Bifidobacterium animalis subsp. lactis LKM512 and arginine improved mouse cognitive flexibility in middle-aged adulthood (over 13 months). It is possible that improvement of brain–gut axis can improve cognitive decline with aging.

Chen et al. generated a mouse model of the developmental and epileptic encephalopathy variant, SCN1B-c.265C>T, predicting SCN1B-p.R89C. Scn1b^C89/C89 mice more accurately model human variants with incomplete loss-of-function compared with Scn1b^−/− mice, with complete loss-of-function, adding to our translational toolbox to develop novel therapeutic strategies for the genetic epilepsies.

Poulos et al. report a retrospective cohort study of 40 patients with congenital myasthenic syndrome characterizing the natural history focused on respiratory outcomes. They provide key respiratory markers and useful insights that may help early diagnosis, surveillance and interventions in respiratory management of congenital myasthenic syndrome thereby improving outcomes.

Qiang et al. report that increased CSF 14-3-3β levels in patients with Alzheimer’s disease correlate with cognitive decline and biomarkers of the disease, suggesting CSF 14-3-3β as a potential biomarker and tool for monitoring disease progression.

Hoogenboom et al. report that repetitive subconcussive head impacts cause microstructural tissue changes in the developing rat brain, which are detectable with diffusion tensor imaging, and with suggestion of correlates in tissue pathology and behaviour. The model provides a starting point for further characterization of subconcussive head impacts in rodents.

Serrano et al. describe zebrafish models for investigating UBA5-related disease pathophysiology. uba5 mutants show locomotor impairment, cerebellar degeneration, delayed growth and reduced survival, reproducing phenotypes observed in affected individuals. Furthermore, involvement of the peripheral nervous system and widespread mitochondrial damage at early stages of the phenotype were identified.

Mole, Nelson and colleagues report on the largest analysis of the quality of phonemic fluency responses in focally lesioned patients. A novel transfer learning approach, capable of identifying left frontal patients based on quality of responses, is discussed. This approach predicted frontal dysfunction with far greater accuracy than other features.

Using MRI at 7T, Evangelisti et al. report that participants with early Huntington’s and Parkinson’s disease show opposite pattern of resting-state functional connectivity in the subthalamic nucleus and sensorimotor cortex. Despite going in opposite direction, functional connectivity values in each disorder showed pathological associations with motor and cognitive symptoms.

Boucher et al.’s findings suggest that examination of multiple diffusivity metrics may inform us about various physiological processes occurring within the lesional and perilesional white matter in the acute, subacute and chronic stages post-stroke, such as focal swelling, axonal damage or myelin loss.

Datta et al. report elevated ubiquitin C-terminal hydrolase-L1 (UCH-L1) and neurofilament-light chain (NF-L) in paediatric severe malaria. In cerebral malaria, NF-L is associated with mortality; UCH-L1 is associated with blood–brain barrier dysfunction and neurologic deficits over follow-up. Both markers are associated with worse cognition in cerebral malaria; UCH-L1 is associated with worse cognition in severe malarial anaemia.

Bauer et al. investigate subcortical grey matter in Rasmussen’s encephalitis longitudinally. They found ipsilesional (nucleus accumbens, caudate nucleus, putamen, thalamus) and contralesional (nucleus accumbens, caudate nucleus) atrophy. In type 1, but not 2, they observed contralesional volume increase (amygdala, hippocampus, pallidum, thalamus), which may be related to neuroplasticity or neuroinflammation.

Ishihara et al. report that early lesion features (necrotic core, cytotoxic and perilesional oedemas) following radiofrequency thalamotomy for movement disorders are best demarcated on T₂-weighted MRI. Lesion volumes were significantly reduced by 6 months after surgery. Overall, MRI demonstrates that radiofrequency thalamotomy produces focused and predictable lesion characteristics over time.

Aiba et al. elucidated that detailed clinical findings and course of 32 patients with corticobasal degeneration and 48 patients with corticobasal syndrome with background pathology. They uncovered that frozen gait at onset, dysarthria, personality changes and pyramidal signs may be useful clinical signs for predicting background pathologies in corticobasal syndrome.

See I. McGeachan and King (https://doi.org/10.1093/braincomms/fcad321) for a scientific commentary on this article.

Zare et al. report that cysteine–alanine–glutamine–lysine peptide efficiently binds to the demyelinated area in sections obtained from animal model of multiple sclerosis. When attached to the surface of a nanoparticle loaded with methylprednisolone, this peptide delivered them to the lesion site following systemic injections to mice with a focal demyelination in the brains.

Mustile et al. report that Parkinson’s disease patients showed impaired modulation of neural markers of proactive and reactive cognitive control while walking and stepping over obstacles. Using mobile EEG, the study revealed that Parkinson’s patients have difficulty both in the approach and once past the obstacle.

Faiman et al. explored whether resting-state EEG markers could differentiate between the diagnoses of epilepsy and psychogenic non-epileptic seizures. Two diagnostic accuracy studies (n = 148) tested several univariate markers using machine learning. Classification accuracy was low (45–60%), indicating challenges due to neurophysiological profiles overlap and stronger EEG dynamics in the data.

Tozlu et al. report that subcortical structural disconnectivity due to lesions identified with T2 fluid-attenuated inversion recovery occurs earlier across the spectrum of disability and cognitive impairment in people with multiple sclerosis. Paramagnetic rim lesion (PRL)–based structural disconnectivity in motor-related regions occurs earlier compared to non–PRL-based structural disconnectivity in the disability event sequence.

Du et al. identified cognitive trajectory profiles using person-level slopes and change point from a composite in late middle-aged adults. The groups were associated with differences in health/lifestyle and dementia-related (A/T/N/V) biomarkers. Identifying person-specific subclinical decline may provide an opportunity for early intervention.

Oujamaa et al. followed brain network changes of severe subacute traumatic brain-injured patients during clinical recovery. They used the hub disruption index to quantify the global brain network reorganization. The hub disruption index differentiates fully and minimally conscious patients, underlining the clinical relevance of this index to capture consciousness recovery.

See Edlow and Massimini (https://doi.org/10.1093/braincomms/fcad328) for a scientific commentary on this article.

Regio et al. report that long-term inactivation of wild-type huntingtin in striatal and cortical projecting neurons from adult mice is well tolerated.

Chari et al. studied navigation in a virtual visual environment in normal and vestibular-deficient men and women and found sex-dependent differences in the utilization of vestibular and visual information that have not been previously described.

Kricheldorff et al. compared performance in two adaptive control tasks in patients with Parkinson’s disease relative to healthy control participants. Successful reactive and proactive controls were accompanied by reduced conflict related theta power. Results demonstrate distinct impairments of proactive control in participants with Parkinson’s disease, tested on dopaminergic medications.

Kim et al. report on subthalamic nucleus deep brain stimulation–induced volumes of tissue activation associated with individual components of Parkinson’s disease bradykinesia, showing separate sites associated with the amplitude and frequency of hand movements. They also show that the stimulation of the amplitude ‘sweetspot’ correlates most strongly with changes in motor rating scales.

Sleep deprivation can impact plasma concentrations of neurodegeneration biomarkers. This human study offers in vivo evidence suggesting that meningeal lymphatic clearance capacity has a crucial role in sleep-induced changes in amyloid-β 40/amyloid-β 42 plasma concentrations, while glymphatic function may be particularly significant in altering the plasma concentration of phosphorylated tau peptide 181 during sleep deprivation.

Finding objective fluid-based biomarkers for amyotrophic lateral sclerosis that reflect the pathological processes occurring in a phenotypically heterogeneous disease is essential for assessing if treatments are working or not. This requires evaluating alternate non-invasive biofluids such as urine, with caveats such as standardizing collection, processing and validating candidates.

Hazan et al. summarize the ways in which online clinical tools could be used to support clinicians and patients, in a memory clinic setting, in the diagnosis of Alzheimer’s disease. This is increasingly important, specifically for clinics that implement plasma biomarkers. Further work is required to design and integrate related tools.