Applications of the Natural Language Processing Tool ChatGPT in Clinical Practice: Comparative Study and Augmented Systematic Review.

BACKGROUND: This research integrates a comparative analysis of the performance of human researchers and OpenAI's ChatGPT in systematic review tasks and describes an assessment of the application of natural language processing (NLP) models in clinical practice through a review of 5 studies. OBJECTIVE: This study aimed to evaluate the reliability between ChatGPT and human researchers in extracting key information from clinical articles, and to investigate the practical use of NLP in clinical settings as evidenced by selected studies. METHODS: The study design comprised a systematic review of clinical articles executed independently by human researchers and ChatGPT. The level of agreement between and within raters for parameter extraction was assessed using the Fleiss and Cohen κ statistics. RESULTS: The comparative analysis revealed a high degree of concordance between ChatGPT and human researchers for most parameters, with less agreement for study design, clinical task, and clinical implementation. The review identified 5 significant studies that demonstrated the diverse applications of NLP in clinical settings. These studies' findings highlight the potential of NLP to improve clinical efficiency and patient outcomes in various contexts, from enhancing allergy detection and classification to improving quality metrics in psychotherapy treatments for veterans with posttraumatic stress disorder. CONCLUSIONS: Our findings underscore the potential of NLP models, including ChatGPT, in performing systematic reviews and other clinical tasks. Despite certain limitations, NLP models present a promising avenue for enhancing health care efficiency and accuracy. Future studies must focus on broadening the range of clinical applications and exploring the ethical considerations of implementing NLP applications in health care settings.

Cortical reactivity to transcranial magnetic stimulation predicts risk of post-stroke delirium.

OBJECTIVE: Post-stroke delirium (PSD) is a frequent and with regard to outcome unfavorable complication in acute stroke. The neurobiological mechanisms predisposing to PSD remain poorly understood, and biomarkers predicting its risk have not been established. We tested the hypothesis that hypoexcitable or disconnected brain networks predispose to PSD by measuring brain reactivity to transcranial magnetic stimulation with electroencephalography (TMS-EEG). METHODS: We conducted a cross-sectional study in 33 acute stroke patients within 48 hours of stroke onset. Brain reactivity to single-pulse TMS of dorsolateral prefrontal cortex, primary motor cortex and superior parietal lobule of the right hemisphere was quantified by response intensity, effective connectivity, perturbational complexity index (PCIST), and natural frequency of the TMS-EEG response. PSD development was clinically tracked every 8 hours before and for 7 days following TMS-EEG. RESULTS: Fourteen patients developed PSD while 19 patients did not. The PSD group showed lower excitability, effective connectivity, PCIST and natural frequency compared to the non-PSD group. The maximum PCIST over all three TMS sites demonstrated largest classification accuracy with a ROC-AUC of 0.943. This effect was independent of lesion size, affected hemisphere and stroke severity. Maximum PCIST and maximum natural frequency correlated inversely with delirium duration. CONCLUSIONS: Brain reactivity to TMS-EEG can unravel brain network states of reduced excitability, effective connectivity, perturbational complexity and natural frequency that identify acute stroke patients at high risk for development of delirium. SIGNIFICANCE: Findings provide novel insight into the pathophysiology of pre-delirium brain states and may promote effective delirium prevention strategies in those patients at high risk.

Corticospinal excitability is highest at the early rising phase of sensorimotor µ-rhythm.

Alpha oscillations are thought to reflect alternating cortical states of excitation and inhibition. Studies of perceptual thresholds and evoked potentials have shown the scalp EEG negative phase of the oscillation to correspond to a short-lasting low-threshold and high-excitability state of underlying visual, somatosensory, and primary motor cortex. The negative peak of the oscillation is assumed to correspond to the state of highest excitability based on biophysical considerations and considerable effort has been made to improve the extraction of a predictive signal by individually optimizing EEG montages. Here, we investigate whether it is the negative peak of sensorimotor µ-rhythm that corresponds to the highest corticospinal excitability, and whether this is consistent between individuals. In 52 adult participants, a standard 5-channel surface Laplacian EEG montage was used to extract sensorimotor µ-rhythm during transcranial magnetic stimulation (TMS) of primary motor cortex. Post-hoc trials were sorted from 800 TMS-evoked motor potentials (MEPs) according to the pre-stimulus EEG (estimated instantaneous phase) and MEP amplitude (as an index of corticospinal excitability). Different preprocessing transformations designed to improve the accuracy by which µ-alpha phase predicts excitability were also tested. By fitting a sinusoid to the MEP amplitudes, sorted according to pre-stimulus EEG-phase, we found that excitability was highest during the early rising phase, at a significant delay with respect to the negative peak by on average 45° or 10 ms. The individual phase of highest excitability was consistent across study participants and unaffected by two different EEG-cleaning methods that utilize 64 channels to improve signal quality by compensating for individual noise level and channel covariance. Personalized transformations of the montage did not yield better prediction of excitability from µ-alpha phase. The relationship between instantaneous phase of a brain oscillation and fluctuating cortical excitability appears to be more complex than previously hypothesized. In TMS of motor cortex, a standard surface Laplacian 5-channel EEG montage is effective in extracting a predictive signal and the phase corresponding to the highest excitability appears to be consistent between individuals. This is an encouraging result with respect to the clinical potential of therapeutic personalized brain interventions in the motor system. However, it remains to be investigated, whether similar results can be obtained for other brain areas and brain oscillations targeted with EEG and TMS.

Decision support by machine learning systems for acute management of severely injured patients: A systematic review.

Introduction: Treating severely injured patients requires numerous critical decisions within short intervals in a highly complex situation. The coordination of a trauma team in this setting has been shown to be associated with multiple procedural errors, even of experienced care teams. Machine learning (ML) is an approach that estimates outcomes based on past experiences and data patterns using a computer-generated algorithm. This systematic review aimed to summarize the existing literature on the value of ML for the initial management of severely injured patients. Methods: We conducted a systematic review of the literature with the goal of finding all articles describing the use of ML systems in the context of acute management of severely injured patients. MESH search of Pubmed/Medline and Web of Science was conducted. Studies including fewer than 10 patients were excluded. Studies were divided into the following main prediction groups: (1) injury pattern, (2) hemorrhage/need for transfusion, (3) emergency intervention, (4) ICU/length of hospital stay, and (5) mortality. Results: Thirty-six articles met the inclusion criteria; among these were two prospective and thirty-four retrospective case series. Publication dates ranged from 2000 to 2020 and included 32 different first authors. A total of 18,586,929 patients were included in the prediction models. Mortality was the most represented main prediction group (n = 19). ML models used were artificial neural network ( n = 15), singular vector machine (n = 3), Bayesian network (n = 7), random forest (n = 6), natural language processing (n = 2), stacked ensemble classifier [SuperLearner (SL), n = 3], k-nearest neighbor (n = 1), belief system (n = 1), and sequential minimal optimization (n = 2) models. Thirty articles assessed results as positive, five showed moderate results, and one article described negative results to their implementation of the respective prediction model. Conclusions: While the majority of articles show a generally positive result with high accuracy and precision, there are several requirements that need to be met to make the implementation of such models in daily clinical work possible. Furthermore, experience in dealing with on-site implementation and more clinical trials are necessary before the implementation of ML techniques in clinical care can become a reality.

Convolutional Neural Networks in Spinal Magnetic Resonance Imaging: A Systematic Review.

OBJECTIVE: Convolutional neural networks (CNNs) are being increasingly used in the medical field, especially for image recognition in high-resolution, large-volume data sets. The study represents the current state of research on the application of CNNs in image segmentation and pathology detection in spine magnetic resonance imaging. METHODS: For this systematic literature review, the authors performed a systematic initial search of the PubMed/Medline and Web of Science (Core collection) databases for eligible investigations. The authors limited the search to observational studies. Outcome parameters were analyzed according to the inclusion criteria and assigned to 3 groups: 1) segmentation of anatomical structures, 2) segmentation and evaluation of pathologic structures, and 3) specific implementation of CNNs. RESULTS: Twenty-four retrospectively designed articles met the inclusion criteria. Publication dates ranged from 2017 to 2021. In total, 14,065 patients with 113,110 analyzed images were included. Most authors trained their network with a training-to-testing ratio of 80/20, while all but 2 articles used 5- to 10-fold cross-validation. Nine articles compared their performance results with other neural networks and algorithms, and all 24 articles described outcomes as positive. CONCLUSIONS: State-of-the-art CNNs can detect and segment-specific anatomical landmarks and pathologies across a wide range, comparable to the skills of radiologists and experienced clinicians. With rapidly evolving network architectures and growing medical image databases, the future is likely to show growth in the development and refinement of these capable networks. However, the aid of automated segmentation and classification by neural networks cannot and should not be expected to replace clinical experts.

Personalized neurorehabilitative precision medicine: from data to therapies (MWKNeuroReha) - a multi-centre prospective observational clinical trial to predict long-term outcome of patients with acute motor stroke.

BACKGROUND: Stroke is one of the most frequent diseases, and half of the stroke survivors are left with permanent impairment. Prediction of individual outcome is still difficult. Many but not all patients with stroke improve by approximately 1.7 times the initial impairment, that has been termed proportional recovery rule. The present study aims at identifying factors predicting motor outcome after stroke more accurately than before, and observe associations of rehabilitation treatment with outcome. METHODS: The study is designed as a multi-centre prospective clinical observational trial. An extensive primary data set of clinical, neuroimaging, electrophysiological, and laboratory data will be collected within 96 h of stroke onset from patients with relevant upper extremity deficit, as indexed by a Fugl-Meyer-Upper Extremity (FM-UE) score ≤ 50. At least 200 patients will be recruited. Clinical scores will include the FM-UE score (range 0-66, unimpaired function is indicated by a score of 66), Action Research Arm Test, modified Rankin Scale, Barthel Index and Stroke-Specific Quality of Life Scale. Follow-up clinical scores and applied types and amount of rehabilitation treatment will be documented in the rehabilitation hospitals. Final follow-up clinical scoring will be performed 90 days after the stroke event. The primary endpoint is the change in FM-UE defined as 90 days FM-UE minus initial FM-UE, divided by initial FM-UE impairment. Changes in the other clinical scores serve as secondary endpoints. Machine learning methods will be employed to analyze the data and predict primary and secondary endpoints based on the primary data set and the different rehabilitation treatments. DISCUSSION: If successful, outcome and relation to rehabilitation treatment in patients with acute motor stroke will be predictable more reliably than currently possible, leading to personalized neurorehabilitation. An important regulatory aspect of this trial is the first-time implementation of systematic patient data transfer between emergency and rehabilitation hospitals, which are divided institutions in Germany. TRIAL REGISTRATION: This study was registered at ClinicalTrials.gov ( NCT04688970 ) on 30 December 2020.

Analysis of the paraspinal muscle morphology of the lumbar spine using a convolutional neural network (CNN).

PURPOSE: This single-center study aimed to develop a convolutional neural network to segment multiple consecutive axial magnetic resonance imaging (MRI) slices of the lumbar spinal muscles of patients with lower back pain and automatically classify fatty muscle degeneration. METHODS: We developed a fully connected deep convolutional neural network (CNN) with a pre-trained U-Net model trained on a dataset of 3,650 axial T2-weighted MRI images from 100 patients with lower back pain. We included all qualities of MRI; the exclusion criteria were fractures, tumors, infection, or spine implants. The training was performed using k-fold cross-validation (k = 10), and performance was evaluated using the dice similarity coefficient (DSC) and cross-sectional area error (CSA error). For clinical correlation, we used a simplified Goutallier classification (SGC) system with three classes. RESULTS: The mean DSC was high for overall muscle (0.91) and muscle tissue segmentation (0.83) but showed deficiencies in fatty tissue segmentation (0.51). The CSA error was small for the overall muscle area of 8.42%, and fatty tissue segmentation showed a high mean CSA error of 40.74%. The SGC classification was correctly predicted in 75% of the patients. CONCLUSION: Our fully connected CNN segmented overall muscle and muscle tissue with high precision and recall, as well as good DSC values. The mean predicted SGC values of all available patient axial slices showed promising results. With an overall Error of 25%, further development is needed for clinical implementation. Larger datasets and training of other model architectures are required to segment fatty tissue more accurately.

Induction of LTD-like corticospinal plasticity by low-frequency rTMS depends on pre-stimulus phase of sensorimotor µ-rhythm.

BACKGROUND: Neural oscillations reflect rapidly changing brain excitability states. We have demonstrated previously with EEG-triggered transcranial magnetic stimulation (TMS) of human motor cortex that the positive vs. negative peak of the sensorimotor µ-oscillation reflect corticospinal low-vs. high-excitability states. In vitro experiments showed that induction of long-term depression (LTD) by low-frequency stimulation depends on the postsynaptic excitability state. OBJECTIVE/HYPOTHESIS: We tested the hypothesis that induction of LTD-like corticospinal plasticity in humans by 1 Hz repetitive TMS (rTMS) is enhanced when rTMS is synchronized with the low-excitability state, but decreased or even shifted towards long-term (LTP)-like plasticity when synchronized with the high-excitability state. METHODS: We applied real-time EEG-triggered 1-Hz-rTMS (900 pulses) to the hand area of motor cortex in healthy subjects. In a randomized double-blind three-condition crossover design, pulses were synchronized to either the positive or negative peak of the sensorimotor µ-oscillation, or were applied at random phase (control). The amplitude of motor evoked potentials was recorded as an index of corticospinal excitability before and after 1-Hz-rTMS. RESULTS: 1-Hz-rTMS at random phase resulted in a trend towards LTD-like corticospinal plasticity. RTMS in the positive peak condition (i.e., the low-excitability state) induced significant LTD-like plasticity. RTMS in the negative peak condition (i.e., the high-excitability state) showed a trend towards LTP-like plasticity, which was significantly different from the other two conditions. CONCLUSION: The level of corticospinal depolarization reflected by phase of the µ-oscillation determines the degree of corticospinal plasticity induced by low-frequency rTMS, a finding that may guide future personalized therapeutic stimulation.

The shaky ground truth of real-time phase estimation.

Instantaneous phase of brain oscillations in electroencephalography (EEG) is a measure of brain state that is relevant to neuronal processing and modulates evoked responses. However, determining phase at the time of a stimulus with standard signal processing methods is not possible due to the stimulus artifact masking the future part of the signal. Here, we quantify the degree to which signal-to-noise ratio and instantaneous amplitude of the signal affect the variance of phase estimation error and the precision with which "ground truth" phase is even defined, using both the variance of equivalent estimators and realistic simulated EEG data with known synthetic phase. Necessary experimental conditions are specified in which pre-stimulus phase estimation is meaningfully possible based on instantaneous amplitude and signal-to-noise ratio of the oscillation of interest. An open source toolbox is made available for causal (using pre-stimulus signal only) phase estimation along with a EEG dataset consisting of recordings from 140 participants and a best practices workflow for algorithm optimization and benchmarking. As an illustration, post-hoc sorting of open-loop transcranial magnetic stimulation (TMS) trials according to pre-stimulus sensorimotor µ-rhythm phase is performed to demonstrate modulation of corticospinal excitability, as indexed by the amplitude of motor evoked potentials.

Brain State-dependent Brain Stimulation with Real-time Electroencephalography-Triggered Transcranial Magnetic Stimulation.

The effect of a stimulus to the brain depends not only on the parameters of the stimulus but also on the dynamics of brain activity at the time of the stimulation. The combination of electroencephalography (EEG) and transcranial magnetic stimulation (TMS) in a real-time brain state-dependent stimulation system allows the study of relations of dynamics of brain activity, cortical excitability, and plasticity induction. Here, we demonstrate a newly developed method to synchronize the timing of brain stimulation with the phase of ongoing EEG oscillations using a real-time data analysis system. This real-time EEG-triggered TMS of the human motor cortex, when TMS is synchronized with the surface EEG negative peak of the sensorimotor µ-alpha (8-14 Hz) rhythm, has shown differential corticospinal excitability and plasticity effects. The utilization of this method suggests that real-time information about the instantaneous brain state can be used for efficacious plasticity induction. Additionally, this approach enables personalized EEG-synchronized brain stimulation which may lead to the development of more effective therapeutic brain stimulation protocols.

Effects of antiepileptic drugs on cortical excitability in humans: A TMS-EMG and TMS-EEG study.

Brain responses to transcranial magnetic stimulation (TMS) recorded by electroencephalography (EEG) are emergent noninvasive markers of neuronal excitability and effective connectivity in humans. However, the underlying physiology of these TMS-evoked EEG potentials (TEPs) is still heavily underexplored, impeding a broad application of TEPs to study pathology in neuropsychiatric disorders. Here we tested the effects of a single oral dose of three antiepileptic drugs with specific modes of action (carbamazepine, a voltage-gated sodium channel (VGSC) blocker; brivaracetam, a ligand to the presynaptic vesicle protein VSA2; tiagabine, a gamma-aminobutyric acid (GABA) reuptake inhibitor) on TEP amplitudes in 15 healthy adults in a double-blinded randomized placebo-controlled crossover design. We found that carbamazepine decreased the P25 and P180 TEP components, and brivaracetam the N100 amplitude in the nonstimulated hemisphere, while tiagabine had no effect. Findings corroborate the view that the P25 represents axonal excitability of the corticospinal system, the N100 in the nonstimulated hemisphere propagated activity suppressed by inhibition of presynaptic neurotransmitter release, and the P180 late activity particularly sensitive to VGSC blockade. Pharmaco-physiological characterization of TEPs will facilitate utilization of TMS-EEG in neuropsychiatric disorders with altered excitability and/or network connectivity.

Munc13-3 Is Required for the Developmental Localization of Ca2+ Channels to Active Zones and the Nanopositioning of Cav2.1 Near Release Sensors.

Spatial relationships between Cav channels and release sensors at active zones (AZs) are a major determinant of synaptic fidelity. They are regulated developmentally, but the underlying molecular mechanisms are largely unclear. Here, we show that Munc13-3 regulates the density of Cav2.1 and Cav2.2 channels, alters the localization of Cav2.1, and is required for the development of tight, nanodomain coupling at parallel-fiber AZs. We combined EGTA application and Ca2+-channel pharmacology in electrophysiological and two-photon Ca2+ imaging experiments with quantitative freeze-fracture immunoelectron microscopy and mathematical modeling. We found that a normally occurring developmental shift from release being dominated by Ca2+ influx through Cav2.1 and Cav2.2 channels with domain overlap and loose coupling (microdomains) to a nanodomain Cav2.1 to sensor coupling is impaired in Munc13-3-deficient synapses. Thus, at AZs lacking Munc13-3, release remained triggered by Cav2.1 and Cav2.2 microdomains, suggesting a critical role of Munc13-3 in the formation of release sites with calcium channel nanodomains.

Is there still a role for vancomycin in skin and soft-tissue infections?

PURPOSE OF REVIEW: Skin and soft-tissue infections (SSIs) are among the commonest infections encountered in clinical practice. Spread of methicillin-resistant Staphylococcus aureus SSIs continues to increase in both health care and community settings and presents a challenge for the best treatment choice. Vancomycin has been the mainstay of SSIs treatment, but recently its use has been questioned because of concerns about its efficacy, tolerability, and unfavorable pharmacokinetic/pharmacodynamic profile. The purpose of this review is to establish the current role for vancomycin in light of the literature published from January 2007 to September 2017 on comparison with both old and new alternatives. RECENT FINDINGS: Meta-analyses show better clinical and microbiological outcomes for drugs approved for the treatment of SSI, including those sustained by methicillin-resistant S. aureus, in the last 10 years than for vancomycin. The newer glycopeptides and linezolid decrease the total treatment costs compared with vancomycin, by reducing the length of stay or avoiding the hospitalization. SUMMARY: Vancomycin is noninferior in efficacy and safety to all comparator drugs, including the newest on the market. However, the SSI treatment evidence base presents several shortcomings limiting the clinical applicability of the results. High-level clinical trials should be performed to obtain results that can be generalized and applied effectively in clinical practice.

Effect of antibiotic stewardship on the incidence of infection and colonisation with antibiotic-resistant bacteria and Clostridium difficile infection: a systematic review and meta-analysis.

BACKGROUND: Antibiotic stewardship programmes have been shown to reduce antibiotic use and hospital costs. We aimed to evaluate evidence of the effect of antibiotic stewardship on the incidence of infections and colonisation with antibiotic-resistant bacteria. METHODS: For this systematic review and meta-analysis, we searched PubMed, the Cochrane Database of Systematic Reviews, the Cochrane Central Register of Controlled Trials, and Web of Science for studies published from Jan 1, 1960, to May 31, 2016, that analysed the effect of antibiotic stewardship programmes on the incidence of infection and colonisation with antibiotic-resistant bacteria and Clostridium difficile infections in hospital inpatients. Two authors independently assessed the eligibility of trials and extracted data. Studies involving long-term care facilities were excluded. The main outcomes were incidence ratios (IRs) of target infections and colonisation per 1000 patient-days before and after implementation of antibiotic stewardship. Meta-analyses were done with random-effect models and heterogeneity was calculated with the I2 method. FINDINGS: We included 32 studies in the meta-analysis, comprising 9 056 241 patient-days and 159 estimates of IRs. Antibiotic stewardship programmes reduced the incidence of infections and colonisation with multidrug-resistant Gram-negative bacteria (51% reduction; IR 0·49, 95% CI 0·35-0·68; p<0·0001), extended-spectrum ß-lactamase-producing Gram-negative bacteria (48%; 0·52, 0·27-0·98; p=0·0428), and meticillin-resistant Staphylococcus aureus (37%; 0·63, 0·45-0·88; p=0·0065), as well as the incidence of C difficile infections (32%; 0·68, 0·53-0·88; p=0·0029). Antibiotic stewardship programmes were more effective when implemented with infection control measures (IR 0·69, 0·54-0·88; p=0·0030), especially hand-hygiene interventions (0·34, 0·21-0·54; p<0·0001), than when implemented alone. Antibiotic stewardship did not affect the IRs of vancomycin-resistant enterococci and quinolone-resistant and aminoglycoside-resistant Gram-negative bacteria. Significant heterogeneity between studies was detected, which was partly explained by the type of interventions and co-resistance patterns of the target bacteria. INTERPRETATION: Antibiotic stewardship programmes significantly reduce the incidence of infections and colonisation with antibiotic-resistant bacteria and C difficile infections in hospital inpatients. These results provide stakeholders and policy makers with evidence for implementation of antibiotic stewardship interventions to reduce the burden of infections from antibiotic-resistant bacteria. FUNDING: German Center for Infection Research.

Developmental tightening of cerebellar cortical synaptic influx-release coupling.

Tight coupling between Ca(2+) channels and the sensor for vesicular transmitter release at the presynaptic active zone (AZ) is crucial for high-fidelity synaptic transmission. It has been hypothesized that a switch from a loosely coupled to a tightly coupled transmission mode is a common step in the maturation of CNS synapses. However, this hypothesis has never been tested at cortical synapses. We addressed this hypothesis at a representative small cortical synapse: the synapse connecting mouse cerebellar cortical parallel fibers to Purkinje neurons. We found that the slow Ca(2+) chelator EGTA affected release significantly stronger at immature than at mature synapses, while the fast chelator BAPTA was similarly effective in both groups. Analysis of paired-pulse ratios and quantification of release probability (pr) with multiple-probability fluctuation analysis revealed increased facilitation at immature synapses accompanied by reduced pr. Cav2.1 Ca(2+) channel immunoreactivity, assessed by quantitative high-resolution immuno-electron microscopy, was scattered over immature boutons but confined to putative AZs at mature boutons. Presynaptic Ca(2+) signals were quantified with two-photon microscopy and found to be similar between maturation stages. Models adjusted to fit EGTA dose-response curves as well as differential effects of the Ca(2+) channel blocker Cd(2+) indicate looser and less homogenous coupling at immature terminals compared with mature ones. These results demonstrate functionally relevant developmental tightening of influx-release coupling at a single AZ cortical synapse and corroborate developmental tightening of coupling as a prevalent phenomenon in the mammalian brain.