Influence of next-generation artificial intelligence on headache research, diagnosis and treatment: the junior editorial board members' vision - part 1.

Artificial intelligence (AI) is revolutionizing the field of biomedical research and treatment, leveraging machine learning (ML) and advanced algorithms to analyze extensive health and medical data more efficiently. In headache disorders, particularly migraine, AI has shown promising potential in various applications, such as understanding disease mechanisms and predicting patient responses to therapies. Implementing next-generation AI in headache research and treatment could transform the field by providing precision treatments and augmenting clinical practice, thereby improving patient and public health outcomes and reducing clinician workload. AI-powered tools, such as large language models, could facilitate automated clinical notes and faster identification of effective drug combinations in headache patients, reducing cognitive burdens and physician burnout. AI diagnostic models also could enhance diagnostic accuracy for non-headache specialists, making headache management more accessible in general medical practice. Furthermore, virtual health assistants, digital applications, and wearable devices are pivotal in migraine management, enabling symptom tracking, trigger identification, and preventive measures. AI tools also could offer stress management and pain relief solutions to headache patients through digital applications. However, considerations such as technology literacy, compatibility, privacy, and regulatory standards must be adequately addressed. Overall, AI-driven advancements in headache management hold significant potential for enhancing patient care, clinical practice and research, which should encourage the headache community to adopt AI innovations.

The impact of primary headaches on disability outcomes: a literature review and meta-analysis to inform future iterations of the Global Burden of Disease study.

BACKGROUND: The burden and disability associated with headaches are conceptualized and measured differently at patients' and populations' levels. At the patients' level, through patient-reported outcome measures (PROMs); at population level, through disability weights (DW) and years lived with a disability (YLDs) developed by the Global Burden of Disease Study (GBD). DW are 0-1 coefficients that address health loss and have been defined through lay descriptions. With this literature review, we aimed to provide a comprehensive analysis of disability in headache disorders, and to present a coefficient referring to patients' disability which might inform future GBD definitions of DW for headache disorders. METHODS: We searched SCOPUS and PubMed for papers published between 2015 and 2023 addressing disability in headache disorders. The selected manuscript included a reference to headache frequency and at least one PROM. A meta-analytic approach was carried out to address relevant differences for the most commonly used PROMs (by headache type, tertiles of medication intake, tertiles of females' percentage in the sample, and age). We developed a 0-1 coefficient based on the MIDAS, on the HIT-6, and on MIDAS + HIT-6 which was intended to promote future DW iterations by the GBD consortium. RESULTS: A total of 366 studies, 596 sub-samples, and more than 133,000 single patients were available, mostly referred to cases with migraine. Almost all PROMs showed the ability to differentiate disability severity across conditions and tertiles of medication intake. The indexes we developed can be used to inform future iterations of DW, in particular considering their ability to differentiate across age and tertiles of medication intake. CONCLUSIONS: Our review provides reference values for the most commonly used PROMS and a data-driven coefficient whose main added value is its ability to differentiate across tertiles of age and medication intake which underlie on one side the increased burden due to aging (it is likely connected to the increased impact of common comorbidities), and by the other side the increased burden due to medication consumption, which can be considered as a proxy for headache severity. Both elements should be considered when describing disability of headache disorders at population levels.

Machine learning approach for Migraine Aura Complexity Score prediction based on magnetic resonance imaging data.

BACKGROUND: Previous studies have developed the Migraine Aura Complexity Score (MACS) system. MACS shows great potential in studying the complexity of migraine with aura (MwA) pathophysiology especially when implemented in neuroimaging studies. The use of sophisticated machine learning (ML) algorithms, together with deep profiling of MwA, could bring new knowledge in this field. We aimed to test several ML algorithms to study the potential of structural cortical features for predicting the MACS and therefore gain a better insight into MwA pathophysiology. METHODS: The data set used in this research consists of 340 MRI features collected from 40 MwA patients. Average MACS score was obtained for each subject. Feature selection for ML models was performed using several approaches, including a correlation test and a wrapper feature selection methodology. Regression was performed with the Support Vector Machine (SVM), Linear Regression, and Radial Basis Function network. RESULTS: SVM achieved a 0.89 coefficient of determination score with a wrapper feature selection. The results suggest a set of cortical features, located mostly in the parietal and temporal lobes, that show changes in MwA patients depending on aura complexity. CONCLUSIONS: The SVM algorithm demonstrated the best potential in average MACS prediction when using a wrapper feature selection methodology. The proposed method achieved promising results in determining MwA complexity, which can provide a basis for future MwA studies and the development of MwA diagnosis and treatment.

P3 latency as a biomarker for the complexity of migraine with aura: Event-related potential study.

BACKGROUND: This study aimed to compare the P3 component between patients who have migraines with aura and healthy subjects, and to compare different subtypes of migraine with aura relative to the complexity of migraine aura. METHODS: Average Migraine Aura Complexity Score was calculated for each MwA patient. Visual oddball paradigm was used to elicit the P3 component. P3 amplitudes and latencies elicited from frequent and rare stimuli, as well as from difference wave, were compared with healthy subjects. Subsequently, subtypes of migraine with aura were compared and Average Migraine Aura Complexity Score was used to explore the connection between features of the P3 and complexity of migraine with aura. RESULTS: 37 patients who have migraine with aura (16 with simple aura and 21 with complex aura) patients and 28 healthy subjects were studied. Patients who have migraine with aura had significantly prolonged latencies compared to healthy subjects (411 ± 39 ms vs 372 ± 34 ms, p < 0.001) relative to a rare condition. Patients who have complex aura significantly differs from patients who have simple aura (427 ± 34 ms vs 389 ± 35 ms, p = 0.004) and healthy subjects (372 ± 34 ms, p < 0.001) relative to P3 latency in a rare condition and the patients who have complex aura significantly differs from healthy subjects (442 ± 37 ms vs 394 ± 33 ms, p < 0.001) relative to P3 latency in difference wave. P3 latency from rare condition positively correlated with the Average Migraine Aura Complexity Score (p < 0.001). CONCLUSIONS: Visual oddball paradigm, particularly rare stimuli, could serve as a potential new tool for deep profiling of different clinical complexities among patients who have migraine with aura. Also, the present pattern of P3 components provided new evidence for the cognitive dysfunctions in patients who have migraine with aura.

Characteristics of N400 component elicited in patients who have migraine with aura.

BACKGROUND: This study aimed to examine the N400 effect and event-related potentials (ERPs) elicited from congruent and incongruent stimuli in patients who have migraines with aura (MwA). METHODS: A total of 33 MwA patients and 20 healthy controls (HCs) were studied. They were balanced in age (35.12 ± 8.94 vs 34.70 ± 9.59 years, p = 0.872) and sex (69.7 vs 75.0% females, p = 0.761). ERPs were measured in response to both stimuli, where pictures were preceded with an object name that either matched or mismatched with the object. Averaged amplitudes, peaks, peak latencies, difference waves and topography were compared between MwA and HCs. RESULTS: MwA patients had significantly lower averaged amplitudes at the Fz and F4 sites during incongruent stimuli, as well as reduced peaks at the C3 and Pz sites. Topography showed a more widespread N400 effect over scalp relative to HCs. The difference ERP waveforms did not differ in the N400 effect between groups, but the P600 effect was significantly stronger in the HCs group relative to the MwA group at the Pz (6.52 ± 2.57 vs. 3.50 ± 3.15, p = 0.001) and P4 (5.86 ± 2.79 vs. 3.95 ± 3.64, p = 0.040) sites. CONCLUSIONS: Picture-word matching tasks could serve as a potential new method for the investigation of semantic processing in MwA patients.

Transcranial Doppler Evaluation of the Cerebral Vasculature in Women Patients who Have Migraine with Aura.

BACKGROUND: Previous studies suggest that increased cerebrovascular reactivity might be a feature of patients who have migraine with aura (MwA). The correlation between the clinical presentation of migraine with aura and transcranial Doppler parameters remains unclear. OBJECTIVE: The main aim of this study was to explore cerebral blood flow, vascular resistance, and cerebrovascular reactivity in women MwA. Also, the relationships between hemodynamic conditions and aura characteristics are examined. DESIGN: Cross-sectional study. SETTING: Headache Center, Neurology Clinic, Clinical Center of Serbia. SUBJECTS: Fifty-four women MwA and 49 healthy controls (HCs). METHODS: Transcranial Doppler sonography examination was used to determine blood flow mean velocity (MV) and pulsatility index (PI), as well as breath-holding index (BHI), in 15 arterial segments comprising the circle of Willis. RESULTS: A total of 54 women MwA and 49 HCs were studied. The PIs of all segments of the left and right middle cerebral arteries and the left and right anterior cerebral arteries were significantly higher in MwA with regards to HCs. Also, both the left and right BHIs were significantly higher in MwA than HCs. In addition, MVs of the right vertebral artery and the first segment of the basilar artery were significantly lower in MwA than HCs. Longer duration of migraine aura showed a weak negative correlation with the PI of the left posterior cerebral artery. CONCLUSIONS: Our findings suggest increased vessel pulsatility, abnormal cerebrovascular reactivity, and decreased cerebral blood flow velocity in several arterial segments of the Willis circle in women MwA.

Dysphasia and Other Higher Cortical Dysfunctions During the Migraine Aura-a Systematic Review of Literature.

PURPOSE OF THE REVIEW: Although visual and somatosensory disturbances are the most common migraine aura (MA) symptoms, patients can also experience other symptoms during their MA. The aim of this review is to provide an overview of studies that report symptoms of dysphasia and other higher cortical dysfunctions (HCDs) during MA, as well as to determine the frequency of HCDs. RECENT FINDINGS: Five studies met the inclusion criteria, corresponding to 697 patients overall. The most frequently reported HCDs were those of the language group (range 10-53%). The occurrence of visual HCDs was noted in 12-40 patients, somatosensory HCDs in 12-20%, and memory disturbances in 10-22% of the patients during MAs. MA is associated with a wide range of neurological symptoms, including symptoms of HCD. A better strategy for investigation of the HCD symptoms is needed to correctly stratify patients thus allowing meaningful studies of aura pathophysiology.

Proposal for a Migraine Aura Complexity Score.

OBJECTIVES: Currently, there is no scoring system for assessing the complexity of migraine aura. Our goal was to develop a Migraine Aura Complexity Score that synthesizes the quantity and quality of aura symptoms and to test its applicability in neuroimaging studies. METHODS: Patients with migraine aura were interviewed in order to obtain characteristics of migraine aura. Explorative and confirmatory analyses were used to develop the Migraine Aura Complexity Score. Median values were derived from 10 consecutive migraine auras in each patient. The Migraine Aura Complexity Score was correlated with an average cortical thickness of different brain areas in studied patients. The Surface-based Morphometric Analysis approach was used to estimate cortical thickness. RESULTS: This study included 23 (16 females and seven males) migraineurs with aura. Confirmatory factor analysis suggested the second-order model with three-factor measurement for grading migraine aura. The first factor is linked to higher cortical dysfunction during migraine aura, while the second is associated with the degree of involvement of primary visual and somatosensory cortices; the third linked symptoms of somatosensory aura and hand and head involvement. Positive correlation of Migraine Aura Complexity Score and averaged cortical thickness were found in the left and right hemispheres overall (r = 0.568, p = 0.007; r = 0.617, p = 0.003) and in some of their regions. CONCLUSIONS: This study demonstrates that the Migraine Aura Complexity Score could be a valuable tool for assessing migraine aura. The score could be used in neuroradiological studies in order to achieve a stratification of patients with migraine aura.

Features of migraine aura in teenagers.

BACKGROUND: Complex migraine aura in teenagers can be complicated to diagnose. The aim of this study was to present detailed features of migraine aura in teenage migraineurs. METHODS: This cross-sectional study was conducted in the period from 2008 till 2013. A total number of 40 teenage migraineurs (20 females and 20 males) met criteria for this study. The patients were interviewed using a specially designed questionnaire for collecting data about migraine aura features. Main outcome measures were frequency of visual, somatosensory and higher cortical dysfunction (HCD) symptoms in teenage migraineurs population during the aura, and also within each individual. RESULTS: Visual aura was reported in every attack, followed by somatosensory (60%) and dysphasic (36.4%) aura. Scintillating scotoma and blurry vision were mostly reported and predominant visual symptoms. The most common somatosensory symptom was numbness in hand. HCD were reported by 22 (55%) patients. Slowed speech was mostly reported symptom of HCD, followed by dyslexia, déjà vu phenomenon, color dysgnosia, and dyspraxia. In patients with HCD, aura frequency per year (6.18 ± 3.17 vs. 3.33 ± 2.03, p = 0.003) and prevalence of somatosensory symptoms (77.3% vs. 38.9%, p = 0.014) were significantly higher than in patients without HCD. CONCLUSIONS: Aura symptoms vary to a great extent in complexity in teenage migraineurs. Consequently, results obtained in this study provide useful information for clinicians when faced with unusual migraine aura.

Processing of visual hapaxes in picture naming task: An event-related potential study.

Object recognition and visual categorization are typically swift and seemingly effortless tasks that involve numerous underlying processes. In our investigation, we utilized a picture naming task to explore the processing of rarely encountered objects (visual hapaxes) in comparison to common objects. Our aim was to determine the stage at which these rare objects are classified as unnamable. Contrary to our expectations and in contrast to some prior research on event-related potentials (ERPs) with novel and atypical objects, no differences between conditions were observed in the late time windows corresponding to the P300 or N400 components. However, distinctive patterns between hapaxes and common objects surfaced in three early time windows, corresponding to the posterior N1 and P2 waves, as well as a widespread N2 wave. According to the ERP data, the differentiation between hapaxes and common objects occurs within the first 380 ms of the processing line, involving only limited and indirect top-down influence.

Underestimated phenomena: higher cortical dysfunctions during migraine aura.

INTRODUCTION: Aura occurs in 20-30% of patients with migraine. Some descriptions of aura go far beyond the most frequent visual and sensory symptoms, suggesting the involvement of different cortical areas. The aim of this prospective study was to evaluate the frequency and types of disorders of higher cortical functions (HCF) that occur during visual and/or sensory aura. METHODS: We interviewed 60 patients with visual and/or sensory aura about HCF disorders of praxia, gnosia, memory, and speech, during aura. Patients were divided into two groups, with and without HCF disorders, and were compared in terms of demographic data and aura characteristics. RESULTS: From all 60 patients, 65% reported at least one HCF disorder during aura. The patients with HCF disorders had longer-lasting auras (28.51 ± 16.39 vs. 19.76 ± 11.23, P = 0.016). The most common HCF disorders were motor dysphasia (82.05%) and dysnomia (30.74%). Motor dysphasia was more often reported by patients with visual as well as sensory aura ( P = 0.002). The number of HCF disorders correlated with the aura duration ( P = 0.003). CONCLUSION: According to our results, HCF disorders during aura occur more often than previously thought. The aura duration has some influence on the HCF disorders.

Role of sensory processing sensitivity and high sensation seeking in migraine with typical aura.

The psychological differences between migraine with aura (MwA) patients and healthy controls (HCs) have not been sufficiently investigated in the current scientific literature. Taking this into account, the present study aimed to examine differences between MwA patients and HCs in sensory processing sensitivity factors, high sensation seeking factors, depression, and anxiety. Mentioned variables were also used to determine their predictive role in predicting the group membership (MwA patients vs HCs). The Highly Sensitive Person Scale, the revised High Sensation Seeking Test, and the Hospital Anxiety and Depression Scale were administered to a sample of seventy-one respondents (39 MwA patients and 32 HCs). MwA patients had a significantly higher score for the low sensory threshold (sensory processing sensitivity factor) in comparison with HCs (4.36 ± 1.4 vs 3.45 ± 1.1, p = 0.003). There was no significant difference in other subscales of sensory processing sensitivity, as well as regarding the high sensation seeking, anxiety, and depression scores, between those two groups. The logistic regression model correctly classified 79.5% MwA patients and 66.7% HCs. The low sensory threshold was a statistically significant predictor for MwA patients (p = 0.001). Our results indicate a certain similarity in the brain sensitivities of MwA patients and people with sensory processing sensitivity trait. Moreover, this shows that the constructs of sensitivity in migraine patients and highly sensitive people overlap to an extent, suggesting the similarity between the conceptualization of sensitivity in the psychological literature and the conceptualization of sensitivity in the medical literature.

Migraine with aura detection and subtype classification using machine learning algorithms and morphometric magnetic resonance imaging data.

Introduction: Migraine with aura (MwA) is a neurological condition manifested in moderate to severe headaches associated with transient visual and somatosensory symptoms, as well as higher cortical dysfunctions. Considering that about 5% of the world's population suffers from this condition and manifestation could be abundant and characterized by various symptoms, it is of great importance to focus on finding new and advanced techniques for the detection of different phenotypes, which in turn, can allow better diagnosis, classification, and biomarker validation, resulting in tailored treatments of MwA patients. Methods: This research aimed to test different machine learning techniques to distinguish healthy people from those suffering from MwA, as well as people with simple MwA and those experiencing complex MwA. Magnetic resonance imaging (MRI) post-processed data (cortical thickness, cortical surface area, cortical volume, cortical mean Gaussian curvature, and cortical folding index) was collected from 78 subjects [46 MwA patients (22 simple MwA and 24 complex MwA) and 32 healthy controls] with 340 different features used for the algorithm training. Results: The results show that an algorithm based on post-processed MRI data yields a high classification accuracy (97%) of MwA patients and precise distinction between simple MwA and complex MwA with an accuracy of 98%. Additionally, the sets of features relevant to the classification were identified. The feature importance ranking indicates the thickness of the left temporal pole, right lingual gyrus, and left pars opercularis as the most prominent markers for MwA classification, while the thickness of left pericalcarine gyrus and left pars opercularis are proposed as the two most important features for the simple and complex MwA classification. Discussion: This method shows significant potential in the validation of MwA diagnosis and subtype classification, which can tackle and challenge the current treatments of MwA.

Analysis of neuropsychological and neuroradiological features for diagnosis of Alzheimer's disease and mild cognitive impairment.

BACKGROUND: Age-related neurodegenerative diseases are constantly increasing with prediction that in 2050 over 60 % of population will suffer from some level of cognitive impairment. A cure for the Alzheimer's disease (AD) does not exist, so early diagnosis is of a great importance. Machine learning techniques can help in early diagnosis with deep medical data processing, disease understanding, intervention analysis and knowledge discovery for achieving better medical decision making. METHODS: In this paper, we analyze the dataset consisting of 90 individuals and 482 input features. We investigate the achieved AD prediction performances using seven classifiers and five feature selection algorithms. We pay special focus on analyzing performance by utilizing only a subset of best ranked attributes to establish the minimum amount of input features that ensure acceptable performance. We also investigate the significance of neuropsychological (NP) and neuroradiological (NR) attributes for the AD diagnosis. RESULTS: The accuracy for the whole set of attributes ranged between 66.22 % and 81.00 %, and the weighted average AUROC was between 76.3 % and 95.0 %. The best results were achieved by the naive Bayes classifier and the Relief feature selection algorithm. Additionally, Support Vector Machines classifier shows the most stable results since it depends the least on the feature selection algorithm which is used. As the main result of this paper, we compare the performance of models trained with automatically selected features to models trained with hand-selected features performed by medical experts (NP and NR features). CONCLUSIONS: The results reveal that unlike the NR attributes, the NP attributes achieve a good performance that is comparable to the full set of attributes, which suggests that they possess a high predictive power for AD diagnosis.

Volume alterations of brainstem subregions in migraine with aura.

BACKGROUND: The brainstem plays a significant role in migraine pathogenesis, but a relationship between volume alterations of brainstem subregions and migraine aura characteristics has not been sufficiently investigated. The aim of this study is to compare the volume of the brainstem, and its subregions, between patients with a migraine with aura (MwA) and healthy controls (HC), and also to correlate characteristics of MwA and the volume of the brainstem subregions. METHODS: Forty-two MwA and 42 HCs, balanced by sex and age, were selected for this study. Total brainstem volume changes as well as volume changes in the pons, medulla, midbrain and the superior cerebellar peduncles were investigated in MwA relative to HCs. In addition, the relationships between brainstem subregions and aura characteristics (aura duration, the frequency of the aura, occurrence of somatosensory and dysphasic aura, duration of a headache, intensity of headache pain and disease duration) were explored in MwA. RESULTS: MwA patients had a larger brainstem volume relative to HCs (25,941.35 ±â€¯2559.2 mm3 vs. 25,179.32 ±â€¯2019.1 mm3; p = .008), as well as the midbrain and pons (6155.98 ±â€¯565.7 mm3 vs. 5964.22 ±â€¯457.0 mm3, p = .002; 15,105.13 ±â€¯1765.5 mm3 vs. 14,539.89 ±â€¯1408.4 mm3, p = .007, respectively). Total brainstem volume, as well as volumes of brainstem subregions, were not significantly correlated to the MwA characteristics. CONCLUSION: The results of this study reveal that a migraine with aura is associated with a larger volume of the brainstem with a particular involvement of the midbrain and pons.

Subcortical Volume Changes in Migraine with Aura.

BACKGROUND AND PURPOSE: Various features of the cerebral cortex and white matter have been extensively investigated in migraine with aura (MwA), but the morphological characteristics of subcortical structures have been largely neglected. The aim of this study was to identify possible differences in subcortical structures between MwA patients and healthy subjects (HS), and also to determine the correlations between the characteristics of migraine aura and the volumes of subcortical structures. METHODS: Thirty-two MwA patients and 32 HS matched by sex and age were analyzed in this study. Regional subcortical brain volumes were automatically calculated using the FSL/FMRIB Image Registration and Segmentation Tool software (https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Glossary). A general linear model analysis was used to investigate differences in the volume of subcortical structures between the MwA patients and HS. A partial correlation test was used to assess correlations between the volume of subcortical structures and characteristics of MwA. RESULTS: The volumes of the right globus pallidus, left globus pallidus, and left putamen were significantly smaller in MwA patients than in HS (mean±SD): 1,427±135 mm³ vs. 1,557±136 mm³ (p<0.001), 1,436±126 mm³ vs. 1,550±139 mm³ (p=0.001), and 4,235±437 mm³ vs. 4,522±412 mm³ (p=0.006), respectively. There were no significant relationships between subcortical structures and clinical parameters. CONCLUSIONS: These findings suggest that both the globus pallidi and left putamen play significant roles in the pathophysiology of the MwA. Future studies should determine the cause-and-effect relationships, since these could not be discriminated in this study due to its cross-sectional design.

Application of the Migraine Aura Complexity Score (MACS): Clinical and Neuroimaging Study.

Background: Manifestations of typical migraine aura can be numerous. Investigation of its pathophysiological mechanisms can be challenging if a stratification of phenotypes is not performed. In this context, the Migraine Aura Complexity Score (MACS), recently developed, may help. Here we aimed to categorize migraine patients into homogenous groups using MACS and to compare those groups with respect to patients' characteristics and neuroimaging findings. Methods: Participants who have a migraine with aura (MwA) were interviewed after each attack in order to obtain the characteristics of migraine aura. Thereafter, we scored the complexity of their auras by MACS. The MACS was used to categorize patients into three groups: MwA-S (with simple aura), MwA-MC (with moderately complex aura), and MwA-C (with complex aura). The patient characteristics and estimated cortical thickness of regions of interest, which are potentially linked to the symptoms that develop during the aura, were used to compare these groups. Results: In total, 338 MwA attacks were recorded in analyzed groups. Scotoma was the most frequently reported symptom in the groups, followed by somatosensory aura in the MwA-C group and zig-zag lines in the MwA-MC and MwA-S groups. Patients in the MwA-C and MwA-MC groups had a thicker cortex in the left primary visual cortex with respect to MwA-S group. In addition, patients in the MwA-C group had a thicker cortex in several visual and somatosensory cortical regions relative to the MwA-S group. Conclusions: Our results show that the newly developed MACS can be used for the stratification of MwA patients, herewith allowing the better investigation of changes in migraineurs' brains.