Dumkrieger, Gina

Aim Post-traumatic headache (PTH) attributed to mild traumatic brain injury (mTBI) frequently co-occurs with other physical, cognitive, and emotional symptoms. This study assessed associations between PTH improvement and these symptoms in individuals with acute headaches (APTH) and compared symptoms between those who had recently developed persistent PTH (PPTH) to those who had long-standing PPTH. Methods This study included 105 individuals with APTH (mean age = 43.4, SD = 15.8; 70 females, 35 males) and 46 individuals with PPTH (mean age = 38.2, SD = 10.9; 16 females, 30 males) with an average duration of 11.3 years from two prospective longitudinal studies. For those who had APTH, PTH improvement was assessed at three months post-enrollment using an electronic headache diary. Participants completed questionnaires evaluating physical (including headache symptoms, photosensitivity, hyperacusis, insomnia, autonomic symptoms, and disability), emotional (including depression, anxiety, pain catastrophizing, and post-traumatic stress disorder screening), and cognitive symptoms (including memory and attention). The APTH cohort completed questionnaires at baseline, follow-up #1 (three to four months from baseline), and follow-up #2 (six to seven months from baseline). Mixed Effects Models were used to analyze the temporal trend of symptoms among the APTH improved and non-improved groups. Results Among the 105 individuals with APTH, 60 experienced PTH improvement and 45 did not. At baseline, compared to those who had PTH improvement, those who did not have PTH improvement had higher SCAT total scores and physical subscores, greater levels of hyperacusis, pain catastrophizing, anxiety and depression. At follow-up #1 (three to four months), these individuals continued to show more pronounced physical and emotional symptoms and worse cognitive function and insomnia. Most of these group differences persisted at follow-up #2 (six to seven months). Most symptoms showed progressive improvement over three months for individuals who had PTH improvement but not for those without PTH improvement. Furthermore, individuals with long-standing PPTH showed more severe insomnia, pain catastrophizing, and PTSD compared to those who had recently developed PPTH. Conclusion Individuals with APTH who did not experience PTH improvement at three months had more severe physical, cognitive, and emotional symptoms at baseline compared to those who had PTH improvement. While most symptoms normalized in individuals with PTH improvement, symptoms overall persisted at three and at six months for those with PTH persistence. Individuals who had long-standing PPTH had more severe insomnia and more negative cognitive/emotional responses to pain compared to those who had recently developed PPTH.

Background This narrative review establishes the current state of the art of machine learning approaches for prediction of migraine attacks. Related concepts are highlighted including the identification of triggers or premonitory symptoms and methods for evaluating prediction models. Existing efforts at machine learning prediction of individual migraine headaches and attacks are reviewed in detail. Challenges in this task are discussed. Results A variety of input data and modeling approaches have been used. It is consistently found that individualized models provide better results compared to a generalized model and achievable performance varies considerably between individuals. Patient needs should be assessed to discover what a valuable prediction looks like. The field should develop common standards for evaluating migraine prediction algorithms. Conclusions/Interpretations While the problem is far from solved there is great potential and reason to believe that feasible solutions that improve the quality of life of those with migraine are within our grasp.

Abstract Objectives/Background Post‐traumatic headache (PTH) is a common symptom following mild traumatic brain injury (mTBI). Currently, there is no identified way to accurately predict if, when, and at what pace a person will have PTH improvement. In our prior studies, we focused on predicting headache improvement at 3 months post‐mTBI. However, that approach may overlook individual differences in how headaches evolve over time. This study aimed to identify individual subgroups based on their headache trajectories and to develop machine‐learning (ML) models for early prediction of headache evolution. Methods Participants with acute PTH completed a daily electronic headache diary over 3 months, recording their headache‐related symptoms. Tensor decomposition was utilized to extract latent factors underlying the time‐varying symptoms. We applied clustering techniques on the latent factors to identify patient subgroups with varying headache improvement trajectory. Next, we developed an ML method to classify each individual into a headache trajectory subgroup as early as possible within the 3‐month interval. Results Seventy‐three individuals with acute PTH (mean age = 44.8 years, SD = 14.0; 50 females/23 males) were enrolled between 0 and 59 days post‐mTBI. Data from 54 individuals were used as the training cohort for model training, and 19 individuals were used as the test cohort for model evaluation. Tensor decomposition extracted two latent factors: one factor representing the overall state of PTH severity and disability and the other representing the improvement state of these symptoms over 3 months. Clustering identified four patient subgroups with distinct headache evolution trajectories: (1) severe symptoms without improvement, (2) severe symptoms with mild improvement, (3) milder symptoms with substantial improvement, and (4) mildest symptoms with mild improvement. The proposed ML model achieved 0.80 cross‐validation accuracy in classifying individuals with PTH into subgroups for the training cohort and 0.84 accuracy for the test cohort. Notably, the model required only the first 2 weeks of headache data to accurately identify the subgroup with the mildest headaches, 3 additional weeks to identify the subgroup with the most severe headaches and no improvement in 3 months, and 2 additional weeks to distinguish the remaining subgroups. Conclusion This study identified subgroups of individuals with acute PTH with distinct headache improvement trajectories. The proposed ML method accurately classified individuals into these subgroups using the minimally necessary early headache data for each person, including detecting the subgroup with the mildest headaches at 2 weeks. This approach could offer an estimated forecast of headache burden over time and could assist clinicians with determining treatment needs and eligibility for PTH clinical trials.

Background Headache is a common symptom following mild traumatic brain injury (mTBI). Post-traumatic headache (PTH), a secondary headache disorder that develops after mTBI, often persists for months or years. To identify potential recovery mechanisms and prognostic biomarkers, the present study investigated whether longitudinal changes in pain-induced brain activation differ between healthy controls (HC) and PTH participants showing headache improvement and those without improvement. Methods Thirty-three participants who met International Classification of Headache Disorders, 3rd edition, criteria for acute PTH within 59 days post-mTBI and 33 HC participants were included with no significant differences in demographics. All participants underwent functional magnetic resonance imaging scans at baseline, four weeks, and 16 weeks post-enrollment using a thermal stimulation paradigm with noxious and non-painful heat stimuli. ‘Painful vs. Non-Painful Heat’ contrasts were generated using SPM12. PTH improvement was assessed at three months post-enrollment via electronic headache diaries. Two-sample t -tests compared the brain activation between HC and PTH at baseline. Linear mixed-effects models examined longitudinal changes for HC, PTH improvement and non-improvement groups across visits. Generalized linear models compared these groups within visits. Results Baseline analysis revealed several regions with significantly higher activation in acute PTH compared to HC, including bilateral postcentral gyrus, right superior temporal gyrus, right middle temporal gyrus, left inferior parietal gyrus, right superior parietal gyrus, left ventral striatum, left olfactory cortex, left gyrus rectus, and left middle occipital gyrus. Over time, the PTH improvement group demonstrated progressive normalization across all identified brain regions, whereas the non-improvement group showed only partial normalization in left ventral striatum, left olfactory cortex, and left gyrus rectus. Sustained elevated activation in specific regions distinguished PTH participants without headache improvement from those with headache improvement, suggesting potential biomarkers for persistent PTH. Conclusions Our findings demonstrate significantly altered pain-induced brain activations in participants with acute PTH compared to HC. Longitudinal analysis revealed distinct recovery trajectories: progressive normalization in the improvement group versus persistent alterations in the non-improvement group. These neuroimaging patterns may serve as biomarkers for identifying individuals at risk for persistent PTH, with implications for early intervention and personalized treatment approaches.

Abstract Objective The goal is to provide an overview of artificial intelligence (AI) and machine learning (ML) methodology and appraisal tailored to clinicians and researchers in the headache field to facilitate interdisciplinary communications and research. Background The application of AI to the study of headache and other healthcare challenges is growing rapidly. It is critical that these findings be accurately interpreted by headache specialists, but this can be difficult for non‐AI specialists. Methods This paper is a narrative review of the fundamentals required to understand ML/AI headache research. Using guidance from key leaders in the field of headache medicine and AI, important references were reviewed and cited to provide a comprehensive overview of the terminology, methodology, applications, pitfalls, and bias of AI. Results We review how AI models are created, common model types, methods for evaluation, and examples of their application to headache medicine. We also highlight potential pitfalls relevant when consuming AI research, and discuss ethical issues of bias, privacy and abuse generated by AI. Additionally, we highlight recent related research from across headache‐related applications. Conclusion Many promising current and future applications of ML and AI exist in the field of headache medicine. Understanding the fundamentals of AI will allow readers to understand and critically appraise AI‐related research findings in their proper context. This paper will increase the reader's comfort in consuming AI/ML‐based research and will prepare them to think critically about related research developments.

Abstract Objective To develop machine learning models using patient and migraine features that can predict treatment responses to commonly used migraine preventive medications. Background Currently, there is no accurate way to predict response to migraine preventive medications, and the standard trial‐and‐error approach is inefficient. Methods In this cohort study, we analyzed data from the Mayo Clinic Headache database prospectively collected from 2001 to December 2023. Adult patients with migraine completed questionnaires during their initial headache consultation to record detailed clinical features and then at each follow‐up to track preventive medication changes and monthly headache days. We included patients treated with at least one of the following migraine preventive medications: topiramate, beta‐blockers (propranolol, metoprolol, atenolol, nadolol, timolol), tricyclic antidepressants (amitriptyline, nortriptyline), verapamil, gabapentin, onabotulinumtoxinA, and calcitonin gene‐related peptide (CGRP) monoclonal antibodies (mAbs) (erenumab, fremanezumab, galcanezumab, eptinezumab). We pre‐trained a deep neural network, “TabNet,” using 145 variables, then employed TabNet‐embedded data to construct prediction models for each medication to predict binary outcomes (responder vs. non‐responder). A treatment responder was defined as having at least a 30% reduction in monthly headache days from baseline. All model performances were evaluated, and metrics were reported in the held‐out test set (train 85%, test 15%). SHapley Additive exPlanations (SHAP) were conducted to determine variable importance. Results Our final analysis included 4260 patients. The responder rate for each medication ranged from 28.7% to 34.9%, and the mean time to treatment outcome for each medication ranged from 151.3 to 209.5 days. The CGRP mAb prediction model achieved a high area under the receiver operating characteristics curve (AUC) of 0.825 (95% confidence interval [CI] 0.726, 0.920) and an accuracy of 0.80 (95% CI 0.70, 0.88). The AUCs of prediction models for beta‐blockers, tricyclic antidepressants, topiramate, verapamil, gabapentin, and onabotulinumtoxinA were: 0.664 (95% CI 0.579, 0.745), 0.611 (95% CI 0.562, 0.682), 0.605 (95% CI 0.520, 0.688), 0.673 (95% CI 0.569, 0.724), 0.628 (0.533, 0.661), and 0.581 (95% CI 0.550, 0.632), respectively. Baseline monthly headache days, age, body mass index (BMI), duration of migraine attacks, responses to previous medication trials, cranial autonomic symptoms, family history of headache, and migraine attack triggers were among the most important variables across all models. A variable could have different contributions; for example, lower BMI predicts responsiveness to CGRP mAbs and beta‐blockers, while higher BMI predicts responsiveness to onabotulinumtoxinA, topiramate, and gabapentin. Conclusion We developed an accurate prediction model for CGRP mAbs treatment response, leveraging detailed migraine features gathered from a headache questionnaire before starting treatment. Employing the same methods, the model performances for other medications were less impressive, though similar to the machine learning models reported in the literature for other diseases. This may be due to CGRP mAbs being migraine‐specific. Incorporating medical comorbidities, genomic, and imaging factors might enhance the model performance. We demonstrated that migraine characteristics are important in predicting treatment responses and identified the most crucial predictors for each of the seven types of preventive medications. Our results suggest that precision migraine treatment is feasible.

Abstract Objective To explore hidden Markov models (HMMs) as an approach for defining clinically meaningful headache‐frequency‐based groups in migraine. Background Monthly headache frequency in patients with migraine is known to vary over time. This variation has not been completely characterized and is not well accounted for in the classification of individuals as having chronic or episodic migraine, a diagnosis with potentially significant impacts on the individual. This study investigated variation in reported headache frequency in a migraine population and proposed a model for classifying individuals by frequency while accounting for natural variation. Methods The American Registry for Migraine Research (ARMR) was a longitudinal multisite study of United States adults with migraine. Study participants completed quarterly questionnaires and daily headache diaries. A series of HMMs were fit to monthly headache frequency data calculated from the diary data of ARMR. Results Changes in monthly headache frequency tended to be small, with 47% of transitions resulting in a change of 0 or 1 day. A substantial portion (24%) of months reflected daily headache with individuals ever reporting daily headache likely to consistently report daily headache. An HMM with four states with mean monthly headache frequency emissions of 3.52 (95% Prediction Interval [PI] 0–8), 10.10 (95% PI 4–17), 20.29 (95% PI 12–28), and constant 28 days/month had the best fit of the models tested. Of sequential month‐to‐month headache frequency transitions, 12% were across the 15‐headache days chronic migraine cutoff. Under the HMM, 38.7% of those transitions involved a change in the HMM state, and the remaining 61.3% of the time, a change in chronic migraine classification was not accompanied by a change in the HMM state. Conclusion A divide between the second and third states of this model aligns most strongly with the current episodic/chronic distinction, although there is a meaningful overlap between the states that supports the need for flexibility. An HMM has appealing properties for classifying individuals according to their headache frequency while accounting for natural variation in frequency. This empirically derived model may provide an informative classification approach that is more stable than the use of a single cutoff value.

Objective While a substantial body of research describes the disabling impacts of migraine attacks, less research has described the impacts of migraine on physical functioning between migraine attacks. The objective of this study is to describe physical impairment during and between migraine attacks as a dimension of burden experienced by people living with chronic migraine. Methods The physical impairment domain of the Migraine Physical Function Impact Diary was recorded in headache diaries from the Medication Overuse Treatment Strategy trial. Days with moderate to severe headache were used to approximate migraine attacks. Factor analysis and regression analysis were used to describe associations between migraine and physical impairment. Results 77,662 headache diary entries from 720 participants were analyzed, including 25,414 days with moderate to severe headache, 19,149 days with mild headache, and 33,099 days with no headache. Mean physical impairment score was 41.5 (SD = 26.1) on days with moderate to severe headache, 12.8 (SD = 15.0) on days with mild headache, and 5.2 (SD = 13.1) on days with no headache. Physical impairment on days with mild headache and days with no headache was significantly associated with days since last moderate to severe headache, physical impairment with last moderate to severe headache, mild headache (compared to no headache), depression, hypersensitivities and cranial autonomic symptoms. Conclusions Physical impairment occurs on migraine and non-migraine days. Study participants with frequent headaches, symptoms of depression, hypersensitivities and cranial autonomic symptoms experience physical impairment at a higher rate on days with no headache and days with mild headache. Clinical Trial Registration: ClinicalTrials.gov (NCT02764320)

Abstract Objective To develop a natural language processing (NLP) algorithm that can accurately extract headache frequency from free‐text clinical notes. Background Headache frequency, defined as the number of days with any headache in a month (or 4 weeks), remains a key parameter in the evaluation of treatment response to migraine preventive medications. However, due to the variations and inconsistencies in documentation by clinicians, significant challenges exist to accurately extract headache frequency from the electronic health record (EHR) by traditional NLP algorithms. Methods This was a retrospective cross‐sectional study with patients identified from two tertiary headache referral centers, Mayo Clinic Arizona and Mayo Clinic Rochester. All neurology consultation notes written by 15 specialized clinicians (11 headache specialists and 4 nurse practitioners) between 2012 and 2022 were extracted and 1915 notes were used for model fine‐tuning (90%) and testing (10%). We employed four different NLP frameworks: (1) ClinicalBERT (Bidirectional Encoder Representations from Transformers) regression model, (2) Generative Pre‐Trained Transformer‐2 (GPT‐2) Question Answering (QA) model zero‐shot, (3) GPT‐2 QA model few‐shot training fine‐tuned on clinical notes, and (4) GPT‐2 generative model few‐shot training fine‐tuned on clinical notes to generate the answer by considering the context of included text. Results The mean (standard deviation) headache frequency of our training and testing datasets were 13.4 (10.9) and 14.4 (11.2), respectively. The GPT‐2 generative model was the best‐performing model with an accuracy of 0.92 (0.91, 0.93, 95% confidence interval [CI]) and R 2 score of 0.89 (0.87, 0.90, 95% CI), and all GPT‐2–based models outperformed the ClinicalBERT model in terms of exact matching accuracy. Although the ClinicalBERT regression model had the lowest accuracy of 0.27 (0.26, 0.28), it demonstrated a high R 2 score of 0.88 (0.85, 0.89), suggesting the ClinicalBERT model can reasonably predict the headache frequency within a range of ≤ ± 3 days, and the R 2 score was higher than the GPT‐2 QA zero‐shot model or GPT‐2 QA model few‐shot training fine‐tuned model. Conclusion We developed a robust information extraction model based on a state‐of‐the‐art large language model, a GPT‐2 generative model that can extract headache frequency from EHR free‐text clinical notes with high accuracy and R 2 score. It overcame several challenges related to different ways clinicians document headache frequency that were not easily achieved by traditional NLP models. We also showed that GPT‐2–based frameworks outperformed ClinicalBERT in terms of accuracy in extracting headache frequency from clinical notes. To facilitate research in the field, we released the GPT‐2 generative model and inference code with open‐source license of community use in GitHub. Additional fine‐tuning of the algorithm might be required when applied to different health‐care systems for various clinical use cases.

Background Our prior work demonstrated that questionnaires assessing psychosocial symptoms have utility for predicting improvement in patients with acute post-traumatic headache following mild traumatic brain injury. In this cohort study, we aimed to determine whether prediction accuracy can be refined by adding structural magnetic resonance imaging (MRI) brain measures to the model. Methods Adults with acute post-traumatic headache (enrolled 0–59 days post-mild traumatic brain injury) underwent T1-weighted brain MRI and completed three questionnaires (Sports Concussion Assessment Tool, Pain Catastrophizing Scale, and the Trait Anxiety Inventory Scale). Individuals with post-traumatic headache completed an electronic headache diary allowing for determination of headache improvement at three- and at six-month follow-up. Questionnaire and MRI measures were used to train prediction models of headache improvement and headache trajectory. Results Forty-three patients with post-traumatic headache (mean age = 43.0, SD = 12.4; 27 females/16 males) and 61 healthy controls were enrolled (mean age = 39.1, SD = 12.8; 39 females/22 males). The best model achieved cross-validation Area Under the Curve of 0.801 and 0.805 for predicting headache improvement at three and at six months. The top contributing MRI features for the prediction included curvature and thickness of superior, middle, and inferior temporal, fusiform, inferior parietal, and lateral occipital regions. Patients with post-traumatic headache who did not improve by three months had less thickness and higher curvature measures and notably greater baseline differences in brain structure vs. healthy controls (thickness: p < 0.001, curvature: p = 0.012) than those who had headache improvement. Conclusions A model including clinical questionnaire data and measures of brain structure accurately predicted headache improvement in patients with post-traumatic headache and achieved improvement compared to a model developed using questionnaire data alone.