Machine Learning for Differentiating Essential Tremor: A Scoping Review.

Tremor is a common yet diagnostically challenging neurological symptom spanning a range of disorders such as essential tremor (ET), Parkinson’s Disease (PD), dystonic tremor (DT), and functional tremor (FT) [ 3 , 4 ]. ET is the most prevalent, with an estimated global incidence of 13.3 per 1,000 individuals, rising to 57.9 per 1,000 in those over 65 years of age [ 1 , 5 ]. Despite its frequency, differentiating ET from other tremor types remains difficult due to overlapping clinical features, heterogeneous patient presentations, and variability in clinical interpretation [ 6 ]. Diagnosis typically relies on clinical examination with possible incorporation of a combination of structured tools, including tremor rating scales, Archimedes spiral analyses, and electromyography (EMG); however, these tools ultimately depend on pattern recognition [ 7 , 8 , 9 ]. As clinical data grows in volume and complexity, there is a growing need for tools, such as AI, that can support neurologists by streamlining this pattern-based decision-making. Importantly, ML-based approaches in this field are not intended to establish a biological ground truth for ET, which remains a clinical diagnosis, but rather to model and support clinician-based pattern recognition.

Machine learning (ML) offers a promising avenue for augmenting clinical diagnosis by supporting the pattern recognition that neurologists routinely apply in clinical practice [ 10 , 11 ]. As tremor assessment increasingly involves multimodal data such as accelerometry, EMG, or handwriting samples, ML models can assist in identifying subtle phenotypic differences that may complement traditional evaluation [ 11 , 12 , 13 ]. A range of approaches, including supervised and unsupervised learning as well as ensemble classifiers have been used to analyze these complex datasets and extract diagnostic features [ 11 , 12 , 13 ]. Beyond neurology, ML has already led to major breakthroughs in biomedical science, exemplified by AlphaFold’s success in protein structure prediction [ 14 ], and is actively being explored to assist in early disease detection, such as in breast cancer screening and diabetic retinopathy [ 15 , 16 ]. However, despite its potential, ML remains underutilized in tremor classification, particularly in differentiating ET from PD, DT, or FT, highlighting a gap between technical innovation and clinical implementation.

Recognizing the increasing need for objective methods to aid neurologists in their decision-making [ 17 , 18 ], this scoping review offers a timely first step toward addressing the ongoing diagnostic challenges posed by ET [ 3 , 11 ]. Given the substantial heterogeneity in study designs, patient populations, and methodologies, a scoping review approach was chosen to map the current literature and identify research gaps, rather than conduct a systematic review and meta-analysis. Due to the growing interest in leveraging AI for movement disorder diagnosis, this scoping review seeks to map and synthesize current research on the use of AI methods to distinguish ET from other tremor disorders. While previous reviews have explored machine learning approaches to tremor classification [ 18 ], important differences exist in scope and focus. In particular, the review by De et al. (2023) primarily addressed PD and emphasized accelerometer-based tremor assessment, with a focus on comparing machine learning methodologies [ 18 ]. In contrast, the present scoping review specifically targets ET and its differentiation from other tremor disorders. Furthermore, unlike prior work, we do not restrict inclusion based on data acquisition modality, thereby capturing studies using both wearable sensor-based and non-wearable clinical or instrumental data sources. This broader inclusion strategy allows for a more comprehensive mapping of AI/ML approaches applied to essential tremor across diverse datasets and methodological frameworks.

As a result, we examine the diversity of computational approaches employed, the types of clinical and sensor-based data analyzed, and the diagnostic performance achieved. In doing so, we aim to provide a comprehensive overview of the state of the field, identify methodological limitations, and outline opportunities for future development and validation of AI-assisted tremor classification tools. This synthesis highlights substantial heterogeneity across studies in three key domains: patient populations, tremor data, and AI techniques/data reporting, underscoring the need for a structured overview. By identifying methodological strengths, limitations, and evidence gaps, this review aims to inform future research priorities and facilitate clinical integration of AI through initiatives such as an expert–led consortium, ultimately improving differential diagnosis and access to high-quality care.

Methods

This scoping review was conducted in accordance with the 2020 PRISMA-ScR guidelines [ 2 ]. All studies identified for potential inclusion were indexed from PubMed, Cochrane Library, or Scopus. For each database, all articles published up to the date of data curation (April 2025) were included. The search strategy consisted of two parts: tremor and AI. Given the wide range of possible patient data types, the search was intentionally inclusive and did not restrict the type of data collected from the patient. Briefly, the types of tremor search consisted of medical subject headings (MeSH) of “Tremor” and “Essential Tremor.” The AI string contained terms related to machine learning and artificial intelligence including: “Machine Learning,” “Supervised Machine Learning,” “Unsupervised Learning,” and “Deep Learning.” These two strings were then combined using the AND boolean operator, which retrieves records containing both terms, according to the specific syntax requirements of each database. All search strategies used, along with subsequent article hits, can be found in the supplemental data S1.

Search strings yielded 548 studies, which were then uploaded to Covidence (Melbourne, Australia). Covidence is an online software that facilitates the screening of potential articles for scoping or systematic reviews; the software contains an artificial intelligence-based learning system that automatically removes duplicates and learns from the reviewers’ choices to remove irrelevant articles.

Eligible articles marked for inclusion in this scoping review included all published articles that used AI to distinguish ET from other tremors. Abstracts, commentaries, or posters were not included for analysis. Similarly, any article that only included animal studies, was not published (peer-reviewed), or was not written in English was excluded. Articles that used AI for different purposes were excluded. Additionally, studies that only compared ET to healthy controls or focused primarily on differentiating non-ET tremor types were excluded, as our scope was intentionally limited to studies distinguishing ET from at least one other tremor disorder. Review articles were not included in the final study count; however, the team reviewed them to identify any additional articles not detected by the search terms. No such additional articles were identified.

To assess article eligibility, two authors (DMF and KEH) initially screened each article based on the title and abstract. This screening used defined PICO criteria described briefly above and expanded in supplemental data S2. Both authors had to agree on article inclusion in order for it to proceed to a second review. If there was a conflict, a third, neutral author (SBR) made the final decision on inclusion. After the first review, included articles were reviewed for a second time, where the first two authors read the full text and decided on inclusion based on the PICO criteria. Again, the third author determined inclusion for any articles with disagreements between the original two authors.

If an article passed two rounds of screening, the authors then extracted data. Data extracted from each article included: study design, total number of participants (for each tremor type), the data recorded, the AI method implemented, and the resulting accuracy, sensitivity, and specificity.

The Joanna Briggs Institute (JBI) Critical Appraisal Tool was used to report bias for each paper [ 19 ]. The JBI Critical Appraisal Tool classifies papers based on their design (i.e., case report, clinical trial, cohort study, etc.), and reviewers score the risk of bias using a series of 8–11 questions [ 19 ]. Two reviewers (DMF and KEH) independently scored each paper. If there were conflicts, a third reviewer (SBR) broke the tie. The scores were then used to develop a bias assessment, which consisted of a scale with options of “yes, no, or unclear” for each specific question. The overall bias assessment for each study was an average of the questions that scored a “yes.”

Results

The process for manuscript selection is displayed in Figure 1 . A total of 721 articles were identified in the literature review. Covidence automatically removed 161 duplicates, and 12 additional duplicates were identified manually, leaving 548 articles for screening. Two reviewers independently assessed the deduplicated titles and abstracts from the literature search for relevance, which resulted in the exclusion of 451 additional articles. Subsequently, the 97 articles whose titles and abstracts met the initial inclusion criteria underwent a full-text review to evaluate study design quality and determine which met the final inclusion criteria, resulting in 46 included articles [ 11 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 , 64 ]. This process, including the rationale for article exclusion post-full-text review, is detailed in Figure 1 . The final set of studies included articles published between 1999 and 2025, reflecting a quarter of a century of using AI to differentiate ET from other types of tremors. Supplemental data S3 presents each article included in this scoping review, grouped by its study design. Table 1 contains the information below in one central format.

Study Types and Quality

No prospective studies met the inclusion criteria for this scoping review. All 46 articles best fit the classification of “Diagnostic test accuracy studies” from the JBI classification system. The average study quality was 83.91%. Complete bias assessment scores for each study can be found in the supplemental data S4.

Types of Tremor

A total of 6,051 patients were included across the 46 articles. The tremor types represented include ET, PD, physiological tremor, DT, cerebellar tremor, mixed PD and ET, essential tremor plus, cerebellar ataxia, and multiple sclerosis. The most common type of tremor differentiated from ET was PD, examined in 45 articles. This was followed by physiological tremor (2 articles), DT (2 articles), mixed PD and ET (2 articles), and cerebellar ataxia (2 articles). Supplemental data S5 summarizes the number of patients associated with each tremor type and the corresponding number of studies that distinguish them from ET.

Types of Data Acquired and AI Models

From each included article, both the type of data acquired from participants and the AI models employed were recorded. Common data types included accelerometer data (26 articles), EMG (9 articles), Archimedes spirals (6 articles), gyroscope data (5 articles), and video recordings (5 articles). A comprehensive list of all recorded patient data types is provided in supplemental data S6. The AI models used varied across studies, with the most frequently applied being support vector machines (18 articles), k-nearest neighbors (9 articles), and convolutional neural networks (8 articles) ( Figure 2 ). A comprehensive list of AI models used, along with their respective classifications, is presented in supplemental data S7.

Reported Accuracy, Sensitivity, and Specificity of Models

Of the 46 included articles, 44 reported model accuracy. Due to a large amount of heterogeneity in reporting, patient characteristics, and AI models used, the following data cannot be compared directly. Reported classification accuracies ranged from 60% to 100%; however, these values must be interpreted cautiously, as they reflect agreement with clinician-labeled diagnoses in the setting of known diagnostic uncertainty and heterogeneous study designs. Sensitivity and specificity were reported in 16 and 18 articles, respectively. Sensitivity ranged from 0.44 to 1.00, and specificity ranged from 0.64 to 1.00. Additionally, overall sensitivity and specificity were calculated from 15 articles that provided sufficient data to extrapolate true positives, false positives, true negatives, and false negatives. These values were pooled across studies, yielding an overall sensitivity of 0.83 and an overall specificity of 0.91. As stated, due to methodological variability across studies, including differences in data types and evaluation criteria, no definitive conclusions could be drawn about which model performed best or worse.

Discussion

This scoping review evaluated 46 studies that utilized ML models with electrophysiological signals, kinematic measurements, drawing tasks, functional assessments, clinical rating scales, and video recordings to distinguish ET from other tremor types across 6,501 patients. All studies met the JBI classification for diagnostic test accuracy studies, with an average quality score of 83.91%. The included studies employed a variety of ML approaches, with reported classification accuracies ranging from 60% to 100% and an overall average of 89%. However, there was substantial heterogeneity in patient demographic reporting, tremor data types, AI models used, and data reporting methods. This variability prevented direct comparison across studies and precluded systematic review or meta-analysis techniques. Importantly, given the clinical overlap between tremor syndromes, these findings highlight the potential of ML models as pattern recognition tools that can support and complement neurologists’ clinical expertise in the assessment of tremor disorders.

Examination of the patient populations and study designs revealed limitations in demographic reporting and study methodology. Across the 46 studies, reporting of basic patient demographic information such as race, gender, and age was inconsistent, limiting comparisons across cohorts. Furthermore, no study included case-matched controls, nor were any randomized controlled trials identified. These design elements are important for objectively evaluating AI performance in distinguishing ET from other tremor types. The emergence of the ET-plus classification introduces additional complexity for ML-based tremor studies. Because ET-plus is variably defined, inconsistently applied across centers, and often underreported, its presence likely contributes to diagnostic heterogeneity and label noise in existing datasets [ 3 , 6 ]. Addressing these limitations in future research would allow more robust comparisons across AI models and study settings.

Analysis of data acquisition methods across the 46 included studies revealed clear trends in the modalities used for ML model input. Accelerometer data was the most common modality, appearing in 26 studies, followed by EMG (9 articles), Archimedes spirals (6 articles), gyroscope data (5 articles), and video recordings (5 articles). From a practical standpoint, however, not all data acquisition methods translate easily into clinical practice. For example, while EMG provides rich electrophysiological detail and often incorporates accelerometer sensors, full EMG-based assessment workflows may be difficult to implement routinely in outpatient settings due to time requirements, technical expertise, reimbursement constraints, and patient tolerance [ 65 , 66 ]. In contrast, Archimedes’ spiral drawings, which were used in only six studies, are widely adopted in clinical assessment because of their simplicity, minimal equipment requirements, and ability to visually capture tremor amplitude and variability [ 67 , 68 , 69 ]. Although accelerometer and EMG-based approaches dominate the research space, they may be less feasible in typical outpatient neurology settings, highlighting a disconnect between research priorities and clinical practice. A potential middle ground may lie in smartphone-based assessments that integrate accelerometer and gyroscope data. Although only two studies utilized smartphones, they reported accuracies of 77.90% and 88.00%, comparable to more complex sensor configurations but with far greater theoretical ease of deployment [ 26 , 29 ].

In addition to the range of data types collected, the ML techniques used to distinguish ET from other tremor types also varied considerably. The most common approach was support vector machines (SVM), used in 18 studies, followed by k-nearest neighbors (KNN) in nine studies and convolutional neural networks (CNN) in eight studies. These are all supervised learning models that rely on labeled examples during training to classify new data. For instance, if the algorithm is trained using tremor patterns labeled as PD or ET, it learns decision boundaries that allow it to distinguish between these categories [ 18 ]. However, model training depends heavily on the availability and quality of labeled datasets, an area where the field still faces major limitations. Most studies relied on small, institution-specific datasets, underscoring the need for large, standardized, publicly available datasets. A consortium-led initiative would be well positioned to coordinate this effort and develop equitable AI models that could aid diagnosis or potentially streamline clinical visits for future patients [ 70 ].

While standardizing and simplifying data inputs is essential for clinical implementation, evaluating how different ML models perform is equally important, though not straightforward. Comparing diagnostic accuracy across studies proved challenging because of substantial methodological variability. Each study used its own combination of data types, preprocessing steps, model architectures, and evaluation criteria. In addition, many studies employed ensemble or hybrid approaches, making it difficult to isolate the contribution of individual ML methods. Reporting of evaluation metrics also varied; although most studies reported overall accuracy, fewer included sensitivity, specificity, or area under the curve, which are important for assessing clinical utility. Moreover, almost all models were trained and validated on internal datasets from single institutions, with little use of external validation cohorts. This limits understanding of model generalizability across diverse populations and clinical settings.

Among commonly used models, SVMs appeared in 18 studies, with reported accuracies ranging from 83% to 100%. KNN was used in nine studies, with accuracies ranging from 60% to 94.68%, while CNNs appeared in eight studies with accuracies between 70% and 95.40%. Although these ranges demonstrate generally strong performance, they also indicate that no single method consistently outperforms others across all scenarios. Differences in data types, preprocessing approaches, model architectures, and evaluation strategies preclude definitive conclusions regarding the best-performing AI approach. Furthermore, no studies directly compared model performance against clinicians, since clinician diagnosis typically served as the reference standard for labeling training and test data. Consequently, it remains unclear how these models perform relative to expert clinical judgment. For these reasons, a meta-analysis was not performed, and accuracy values should be interpreted within the context of each study’s design rather than through direct comparisons across models.

Future research should prioritize improved demographic reporting, larger datasets, and more consistent study methodologies. Coordinated multi-center efforts could establish shared protocols for collecting and reporting tremor data while enabling the development of large, publicly available datasets that allow fair comparison of AI models. These efforts should also prioritize tools that integrate into routine clinical practice. Nearly all included studies relied on device-based data such as accelerometers, EMG, and video recordings, whereas AI applications based solely on clinical history or physical examination data were notably absent. Given the practicality and lower resource demands of clinical data, exploring AI methods that incorporate or rely on clinical assessments represents an important future direction.

Although EMG provides detailed electrophysiological information, its routine clinical use can be limited by equipment requirements, time demands, and the need for specialized expertise [ 65 , 66 ]. In contrast, approaches such as Archimedes’ spirals, accelerometers, and gyroscopes are more accessible, particularly when integrated into smartphone-based platforms. One potential approach could involve a smartphone application developed collaboratively by movement disorder specialists and computer scientists. However, smartphone- and wearable-based assessments should currently be considered exploratory rather than validated clinical solutions. Although smartphones can capture accelerometer and gyroscope data, their size, weight, sensor placement, and the need for active holding may alter tremor biomechanics and signal characteristics, effects that wearable devices might mitigate. These factors have not been systematically studied and may introduce confounding effects. Consequently, both smartphone- and wearable-based approaches should currently be viewed as hypothesis-generating tools rather than ready-to-deploy clinical instruments.

Limitations

Several limitations should be considered when interpreting this review. First, only English-language articles were included, which may have excluded relevant non-English studies. Second, studies comparing ET only to healthy control groups or primarily focusing on differentiating another tremor type (e.g., PD) were excluded. Although these studies contribute to the broader literature on ML-based tremor classification, the scope of this review was intentionally limited to work differentiating ET from at least one other tremor type. Additionally, most included studies relied on clinically well-characterized and relatively unambiguous cases from single centers. As a result, current ML models have largely been trained and evaluated in idealized settings, and their performance in diagnostically challenging or longitudinally evolving tremor cases remains uncertain. Addressing this gap will be essential for meaningful clinical translation.

The level of methodological and reporting detail varied considerably across studies. Some articles focused primarily on ML algorithm development, whereas others emphasized data collection and patient characteristics. This inconsistency limited the ability to calculate overall sensitivity and specificity across all 46 studies, thereby affecting generalizability. Similarly, the types of patient data used for model training differed widely, making it difficult to determine which data modalities are most effective for ML-based tremor classification. These limitations further emphasize the need for standardized data collection protocols and reporting guidelines to improve model development and evaluation. No included studies employed case-matched controls or randomized controlled trial designs. Our exclusion of studies comparing ET only to healthy controls likely filtered out some RCTs, contributing to the absence of such designs. Future work incorporating case-matched cohorts and randomized methodologies would strengthen evidence regarding AI performance.

Finally, because clinician diagnosis served as the reference standard in nearly all studies, model performance is inherently constrained by the accuracy and consistency of clinical tremor classification itself. Some patients labeled with ET may not truly have the condition, while others diagnosed with alternative tremor types may actually have ET. This diagnostic uncertainty introduces label noise, meaning that ML model performance largely reflects agreement with clinician judgment rather than identification of a definitive biological substrate. In addition, diagnostic adjudication methods were inconsistently reported. Few studies described whether diagnoses were established by single clinicians, expert panels, or consensus processes, and descriptions of research diagnostic criteria were often limited. This lack of transparency reduces confidence in the ground truth labels used for model training and highlights a critical gap in the literature. Moreover, few studies reported detailed error analyses examining the nature of misclassifications, such as cases where ML models labeled tremor as ET while clinicians diagnosed another tremor type, or vice versa. Given the imperfect ground truth based on clinician diagnosis, further work examining model failure modes will be essential. Consequently, reported accuracies should be interpreted as measures of concordance with expert classification rather than objective diagnostic truth. This may also help explain why average diagnostic accuracy has remained relatively stable over the past two decades despite advances in AI technology.

Conclusion

In conclusion, this scoping review demonstrates that ML holds significant promise as a diagnostic aid for differentiating ET from other tremor types, a task that remains challenging due to overlapping clinical features and the absence of a definitive biomarker. Across 46 studies, a variety of ML models were trained using diverse data sources ranging from resource-intensive modalities such as EMG to more accessible inputs like smartphone-based accelerometry and gyroscope data. Although reported accuracies ranged from 60% to 100%, heterogeneity in data types, preprocessing methods, and model architectures limited direct comparison across studies. Nevertheless, these findings highlight that tremor classification is fundamentally a pattern recognition problem, an area in which ML is particularly well suited.

The clinical impact of these tools will depend on broader efforts to develop large, standardized, publicly available datasets, automate preprocessing pipelines, and prioritize clinically feasible data sources ( Figure 2 ). Rather than functioning as standalone solutions, ML systems should be designed to complement the clinical reasoning of movement disorder specialists and neurologists, particularly in diagnostically ambiguous cases. International societies could play a key role in coordinating these efforts by forming multidisciplinary task forces that include both movement disorder neurologists and computer scientists. When effectively integrated, ML models may enhance pattern recognition beyond what is achievable through clinical observation alone, providing neurologists with an additional layer of decision support to improve diagnostic confidence.

Financial Disclosures for the previous 12 months

KEH is supported by the National Science Foundation Research Traineeship (NRT) Program under Grant No. 2125872. AWA is a member of the faculty of the University of Colorado and is supported by university funds. She has served as a site investigator for studies sponsored by Michael J Fox Foundation for Parkinson’s Research, Parkinson Study Group, Aligning Science Across Parkinson’s (ASAP) Initiative, Koneksa Health, Biohaven Pharmaceuticals, Inc., Bial, and the NIH NINDS. She receives grant funding from NIH NICHD. She has received honoraria for consultancy to Aevum through Parkinson Study Group. She is a consultant for PhotoPharmics, Inc. The other authors declare that there are no additional disclosures to report.

Additional Files

The additional files for this article can be found as follows:

Systematic literature search conducted April 19, 2025.

PICO Criteria defined for article inclusion.

All studies included for article review with their study design.

Complete bias assessment scores for each study.

Patients from included studies divided into tremor types.

Full list of patient data acquired by each article.

Full list of AI models used by each article.

Ethics and Consent

The authors confirm that the approval of an institutional review board and informed patient consent was not required for this work. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Author Contributions

(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the first draft, B. Review and Critique

DMF: 1B, 1C, 2A, 2B, 3A

KEH: 1B, 1C, 2A, 2B, 3A

SBR: 1C, 3A

AM: 1C, 3A

JCM: 1C, 3B

AWA: 3B

AAM: 1A, 1B, 2C, 3B

David M. Fletcher and Kaitlyn E. Heintzelman contributed equally to this work.

References

1. Louis ED, McCreary M.. How Common is Essential Tremor? Update on the Worldwide Prevalence of Essential Tremor. Tremor Other Hyperkinet Mov (N Y). 2021;11(1):28. DOI: 10.5334/tohm.632

2. Tricco AC, Lillie E, Zarin W, et al. PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklist and Explanation. Ann Intern Med. 2018;169:467–473. DOI: 10.7326/M18-0850

3. Amlang CJ, Trujillo Diaz D, Louis ED.. Essential Tremor as a “Waste Basket” Diagnosis: Diagnosing Essential Tremor Remains a Challenge. Front Neurol. 2020;11:172. DOI: 10.3389/fneur.2020.00172

4. Lenka A, Jankovic J.. Tremor Syndromes: An Updated Review. Front Neurol. 2021;12:684835. DOI: 10.3389/fneur.2021.684835

5. Louis ED, Ottman R, Allen Hauser W.. How common is the most common adult movement disorder? Estimates of the prevalence of essential tremor throughout the world. Mov Disord. 1998;13(1):5–10. DOI: 10.1002/mds.870130105

6. Bhatia KP, Bain P, Bajaj N, et al. Consensus Statement on the classification of tremors. from the task force on tremor of the International Parkinson and Movement Disorder Society. Mov Disord. 2018;33(1):75–87. DOI: 10.1002/mds.27121

7. Ondo W, Shukla AW, Ondo C, Tremor Research Group OBOT.. TETRAS Spirals and Handwriting Samples: Determination of Optimal Scoring Examples. Tremor Other Hyperkinet Mov (N Y). 2021;11(1):50. DOI: 10.5334/tohm.665

8. Hess CW, Pullman SL.. Tremor: Clinical Phenomenology and Assessment Techniques. Tremor Other Hyperkinet Mov (N Y). 2012;2:02. DOI: 10.5334/tohm.115

9. Legrand AP, Rivals I, Richard A, et al. New insight in spiral drawing analysis methods – Application to action tremor quantification. Clin Neurophysiol. 2017;128(10):1823–1834. DOI: 10.1016/j.clinph.2017.07.002

10. Ahsan MM, Luna SA, Siddique Z.. Machine-Learning-Based Disease Diagnosis: A Comprehensive Review. Healthcare (Basel). 2022;10(3):541. DOI: 10.3390/healthcare10030541

11. Xing X, Luo N, Li S, Zhou L, Song C, Liu J.. Identification and Classification of Parkinsonian and Essential Tremors for Diagnosis Using Machine Learning Algorithms. Front Neurosci. 2022;16:701632. DOI: 10.3389/fnins.2022.701632

12. Peng Y, Han S, Wu D, et al. A deep learning approach to remotely assessing essential tremor with handwritten images. Sci Rep. 2025;15(1):10783. DOI: 10.1038/s41598-025-94729-0

13. Kline A, Wang H, Li Y, et al. Multimodal machine learning in precision health: A scoping review. NPJ Digit Med. 2022;5(1):171. DOI: 10.1038/s41746-022-00712-8

14. Jumper J, Evans R, Pritzel A, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583–589. DOI: 10.1038/s41586-021-03819-2

15. McKinney SM, Sieniek M, Godbole V, et al. International evaluation of an AI system for breast cancer screening. Nature. 2020;577(7788):89–94. DOI: 10.1038/s41586-019-1799-6

16. Gulshan V, Peng L, Coram M, et al. Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs. JAMA. 2016;316(22):2402–2410. DOI: 10.1001/jama.2016.17216

17. Vizcarra JA, Yarlagadda S, Xie K, Ellis CA, Spindler M, Hammer LH.. Artificial Intelligence in the Diagnosis and Quantitative Phenotyping of Hyperkinetic Movement Disorders: A Systematic Review. J Clin Med. 2024;13(23):7009. DOI: 10.3390/jcm13237009

18. De A, Bhatia KP, Volkmann J, Peach R, Schreglmann SR.. Machine Learning in Tremor Analysis: Critique and Directions. Mov Disord. 2023;38(5):717–731. DOI: 10.1002/mds.29376

19. Aromataris E, Lockwood C, Porritt K, Pilla B, Jordan Z, editors. JBI Manual for Evidence Synthesis [Internet]. JBI; 2024. [cited 2025 Jun 9]. Available from: https://jbi-global-wiki.refined.site/space/MANUAL .

20. Ai L, Wang J, Huang L, Wang X.. Hand Tremor Classification Using Bispectrum Analysis of Acceleration Signals and Back-Propagation Neural Network. In: Liu D, Fei S, Hou Z, Zhang H, Sun C. editors. Advances in Neural Networks – ISNN 2007. 2007; p 1202–1210. DOI: 10.1007/978-3-540-72393-6_142

21. Ai L, Wang J, Wang X.. Multi-features fusion diagnosis of tremor based on artificial neural network and D–S evidence theory. Signal Process. 2008;88(12):2927–2935. DOI: 10.1016/j.sigpro.2008.06.018

22. Ai L, Wang J, Yao R.. Classification of parkinsonian and essential tremor using empirical mode decomposition and support vector machine. Digit Signal Process. 2011;21(4):543–550. DOI: 10.1016/j.dsp.2011.01.010

23. Anandapadmanabhan R, Vishnoi A, Raman G, et al. Deep Learning–Based Artificial Intelligence Algorithm to Classify Tremors from Hand-Drawn Spirals. Mov Disord. 2025;40(6):1172–1181. DOI: 10.1002/mds.30176

24. Aracri F, Quattrone A, Bianco MG, et al. Multimodal imaging and electrophysiological study in the differential diagnosis of rest tremor. Front Neurol. 2024;15:1399124. DOI: 10.3389/fneur.2024.1399124

25. Aubin PM, Serackis A, Griskevicius J.. Support Vector Machine Classification of Parkinson’s Disease, Essential Tremor and Healthy Control Subjects Based on Upper Extremity Motion. In: 2012 International Conference on Biomedical Engineering and Biotechnology. Macau, China; May 2012. pp. 900–904. DOI: 10.1109/iCBEB.2012.387

26. Balachandar A, Algarni M, Oliveira L, et al. Are smartphones and machine learning enough to diagnose tremor? J Neurol. 2022;269(11):6104–6115. DOI: 10.1007/s00415-022-11293-7

27. Chandra Reddy G, Kumar AT, Rebello A, Brindha A, Pa S, Sunitha KA.. Enhanced Computer Vision Technique for Differentiating Tremor Types. In: 2024 5th International Conference on Biomedical Engineering (IBIOMED). Bali, Indonesia; 2024. p 169–174. DOI: 10.1109/iBioMed62485.2024.10875825

28. Darnall ND, Donovan CK, Aktar S, et al. Application of machine learning and numerical analysis to classify tremor in patients affected with essential tremor or Parkinson’s disease. Gerontechnology. 2012;10(4):208–219. DOI: 10.4017/gt.2012.10.4.002.00

29. Duque JDL, Egea AJS, Reeb T, Rojas HAG, Gonzalez-Vargas AM.. Angular Velocity Analysis Boosted by Machine Learning for Helping in the Differential Diagnosis of Parkinson’s Disease and Essential Tremor. IEEE Access. 2020;8:88866–88875. DOI: 10.1109/ACCESS.2020.2993647

30. Engin M, Demirag S, Engin E, et al. The classification of human tremor signals using artificial neural network. Expert Syst Appl. 2007;33(3):754–761. DOI: 10.1016/j.eswa.2006.06.014

31. Ferreira GAS, Teixeira JLS, Rosso ALZ, De Sá AMFLM.. On the classification of tremor signals into dyskinesia, Parkinsonian tremor, and Essential tremor by using machine learning techniques. Biomed Signal Process Control. 2022;73:103430. DOI: 10.1016/j.bspc.2021.103430

32. Ghassemi NH, Marxreiter F, Pasluosta CF, et al. Combined accelerometer and EMG analysis to differentiate essential tremor from Parkinson’s disease. Annu Int Conf IEEE Eng Med Biol Soc. 2016;2016:672–675. DOI: 10.1109/EMBC.2016.7590791

33. González R, Barrientos A, Del Cerro J, Coca B.. DIMETER: A Haptic Master Device for Tremor Diagnosis in Neurodegenerative Diseases. Sensors (Basel). 2014;14(3):4536–4559. DOI: 10.3390/s140304536

34. Groznik V, Guid M, Sadikov A, et al. Elicitation of neurological knowledge with argument-based machine learning. Artif Intell Med. 2013;57(2):133–144. DOI: 10.1016/j.artmed.2012.08.003

35. Hossen A. Discrimination between parkinsonian tremor and essential tremor using artificial neural network with hybrid features. Technol Health Care. 2022;30(3):691–702. DOI: 10.3233/THC-213324

36. Hossen A, Muthuraman M, Raethjen J, Deuschl G, Heute U.. A Neural Network Approach to Distinguish Parkinsonian Tremor from Advanced Essential Tremor. In: Deep K, Nagar A, Pant M, Bansal JC, editors. Proceedings of the International Conference on Soft Computing for Problem Solving (SocProS 2011); 2012. pp. 1051–1060. DOI: 10.1007/978-81-322-0487-9_96

37. Ishii N, Mochizuki Y, Shiomi K, Nakazato M, Mochizuki H.. Spiral drawing: Quantitative analysis and artificial-intelligence-based diagnosis using a smartphone. J Neurol Sci. 2020;411:116723. DOI: 10.1016/j.jns.2020.116723

38. Jakubowski J, Kwiatos K, Chwaleba A, Osowski S.. Higher order statistics and neural network for tremor recognition. IEEE Trans Biomed Eng. 2002;49(2):152–159. DOI: 10.1109/10.979354

39. Kovalenko E, Talitckii A, Anikina A, et al. Distinguishing Between Parkinson’s Disease and Essential Tremor Through Video Analytics Using Machine Learning: A Pilot Study. IEEE Sens J. 2021;21(10):11916–11925. DOI: 10.1109/JSEN.2020.3035240

40. Lee RE, Chan PY.. Explainable artificial intelligence for searching frequency characteristics in Parkinson’s disease tremor. Sci Rep. 2023;13(1):18622. DOI: 10.1038/s41598-023-45802-z

41. Li J, Zhu H, Li J, et al. A Wearable Multi-Segment Upper Limb Tremor Assessment System for Differential Diagnosis of Parkinson’s Disease Versus Essential Tremor. IEEE Trans Neural Syst Rehabil Eng. 2023;31:3397–3406. DOI: 10.1109/TNSRE.2023.3306203

42. Lin S, Gao C, Li H, et al. Wearable sensor-based gait analysis to discriminate early Parkinson’s disease from essential tremor. J Neurol. 2023;270(4):2283–2301. DOI: 10.1007/s00415-023-11577-6

43. Locatelli P, Alimonti D, Traversi G, Re V.. Classification of Essential Tremor and Parkinson’s Tremor Based on a Low-Power Wearable Device. Electronics. 2020;9(10):1695. DOI: 10.3390/electronics9101695

44. Moon S, Song HJ, Sharma VD, et al. Classification of Parkinson’s disease and essential tremor based on balance and gait characteristics from wearable motion sensors via machine learning techniques: a data-driven approach. J NeuroEngineering Rehabil. 2020;17(1):125. DOI: 10.1186/s12984-020-00756-5

45. Nanayakkara GRR, Chan PY.. Subclinical tremor differentiation using long short-term memory networks. Phys Eng Sci Med. 2025;48(2):591–602. DOI: 10.1007/s13246-025-01526-0

46. Nanda SK, Lin WY, Lee MY, Chen RS.. A quantitative classification of essential and Parkinson’s tremor using wavelet transform and artificial neural network on sEMG and accelerometer signals. In: 2015 IEEE 12th International Conference on Networking, Sensing and Control. Taipei, Taiwan, 2015; p 399–404. DOI: 10.1109/ICNSC.2015.7116070

47. Oktay AB, Kocer A.. Differential diagnosis of Parkinson and essential tremor with convolutional LSTM networks. Biomed Signal Process Control. 2020;56:101683. DOI: 10.1016/j.bspc.2019.101683

48. Piepjohn P, Bald C, Kuhlenbäumer G, Becktepe JS, Deuschl G, Schmidt G.. Real-time classification of movement patterns of tremor patients. Biomed Tech (Berl). 2022;67(2):119–130. DOI: 10.1515/bmt-2021-0140

49. Ranjan R, Palaniswami M, Bhushan B.. A Machine Learning Approach for Classification of Tremor–A Neurological Movement Disorder. In: Bi Y, Bhatia R, Kapoor S, editors. Intelligent Systems and Applications (IntelliSys 2019); 2019. pp. 1289–1307. DOI: 10.1007/978-3-030-29513-4_95

50. Saad M, Hefner S, Donovan S, et al. Development of a Tremor Detection Algorithm for Use in an Academic Movement Disorders Center. Sensors. 2024;24(15):4960. DOI: 10.3390/s24154960

51. Sanderson JB, Yu JH, Liu DD, et al. Multi-Dimensional, Short-Timescale Quantification of Parkinson’s Disease and Essential Tremor Motor Dysfunction. Front Neurol. 2020;11:886. DOI: 10.3389/fneur.2020.00886

52. Seedat N, Aharonson V, Schlesinger I.. Automated machine vision enabled detection of movement disorders from hand drawn spirals. In: 2020 IEEE International Conference on Healthcare Informatics (ICHI). Oldenburg, Germany; 2020. pp. 1–5. DOI: 10.1109/ICHI48887.2020.9374333

53. Shahtalebi S, Atashzar SF, Patel RV, Jog MS, Mohammadi A.. A deep explainable artificial intelligent framework for neurological disorders discrimination. Sci Rep. 2021;11(1):9630. DOI: 10.1038/s41598-021-88919-9

54. Shahtalebi S, Atashzar SF, Samotus O, Patel RV, Jog MS, Mohammadi A.. PHTNet: Characterization and Deep Mining of Involuntary Pathological Hand Tremor using Recurrent Neural Network Models. Sci Rep. 2020;10(1):2195. DOI: 10.1038/s41598-020-58912-9

55. Skaramagkas V, Andrikopoulos G, Kefalopoulou Z, Polychronopoulos P.. Towards Differential Diagnosis of Essential and Parkinson’s Tremor via Machine Learning. In: 2020 28th Mediterranean Conference on Control and Automation (MED). Saint-Raphaël, France, 2020; pp. 782–787. DOI: 10.1109/MED48518.2020.9182922

56. Skaramagkas V, Andrikopoulos G, Kefalopoulou Z, Polychronopoulos P.. A Study on the Essential and Parkinson’s Arm Tremor Classification. Signals. 2021;2(2):201–224. DOI: 10.3390/signals2020016

57. Spyers-Ashby JM, Stokes MJ, Bain PG, Roberts SJ.. Classification of normal and pathological tremors using a multidimensional electromagnetic system. Med Eng Phys. 1999;21(10):713–723. DOI: 10.1016/S1350-4533(00)00004-7

58. Surangsrirat D, Thanawattano C, Pongthornseri R, Dumnin S, Anan C, Bhidayasiri R.. Support vector machine classification of Parkinson’s disease and essential tremor subjects based on temporal fluctuation. Annu Int Conf IEEE Eng Med Biol Soc. 2016;2016:6389–6392. DOI: 10.1109/EMBC.2016.7592190

59. Tang L, Hu Q, Wang X, et al. A multimodal fusion network based on a cross-attention mechanism for the classification of Parkinsonian tremor and essential tremor. Sci Rep. 2024;14(1):28050. DOI: 10.1038/s41598-024-79111-w

60. Tavakkoli P, Kobravi HR, Shoeibi A, Zakhirehdari N.. Discrimination of Parkinsonian Tremor from Essential Tremor Using the Nonlinear Analysis of Wrist Muscle EMG Signals. In: Goh J, editor. The 15th International Conference on Biomedical Engineering; 2014. pp. 667–670. DOI: 10.1007/978-3-319-02913-9_170

61. Teo YX, Lee RE, Nurzaman SG, Tan CP, Chan PY.. Action tremor features discovery for essential tremor and Parkinson’s disease with explainable multilayer BiLSTM. Comput Biol Med. 2024;180:108957. DOI: 10.1016/j.compbiomed.2024.108957

62. Vescio B, De Maria M, Crasà M, et al. Development of a New Wearable Device for the Characterization of Hand Tremor. Bioengineering. 2023;10(9):1025. DOI: 10.3390/bioengineering10091025

63. Weede P, Smietana PD, Kuhlenbäumer G, Deuschl G, Schmidt G.. Two-Stage Convolutional Neural Network for Classification of Movement Patterns in Tremor Patients. Information. 2024;15(4):231. DOI: 10.3390/info15040231

64. Yang TL, Lin CH, Chen WL, Lin HY, Su CS, Liang CK.. Hash Transformation and Machine Learning-Based Decision-Making Classifier Improved the Accuracy Rate of Automated Parkinson’s Disease Screening. IEEE Trans Neural Syst Rehabil Eng. 2020;28(1):72–82. DOI: 10.1109/TNSRE.2019.2950143

65. Chen JJ, Swope DM.. Essential Tremor: Diagnosis and Treatment. Pharmacotherapy. 2003;23(9):1105–1122. DOI: 10.1592/phco.23.10.1105.32750

66. Jerath NU, Strader SB, Reddy CG, Swenson A, Kimura J, Aul E.. Factors influencing aversion to specific electrodiagnostic studies. Brain Behav. 2014;4(5):698–702. DOI: 10.1002/brb3.240

67. Pullman SL. Spiral Analysis: A New Technique for Measuring Tremor With a Digitizing Tablet. Mov Disord. 2008;13(S3):85–89. DOI: 10.1002/mds.870131315

68. Solé-Casals J, Anchustegui-Echearte I, Marti-Puig P, et al. Discrete Cosine Transform for the Analysis of Essential Tremor. Front Physiol. 2019;9:1947. DOI: 10.3389/fphys.2018.01947

69. Otero JFA, López-de-Ipina K, Caballer OS, et al. EMD-based data augmentation method applied to handwriting data for the diagnosis of Essential Tremor using LSTM networks. Sci Rep. 2022;12(1):12819. DOI: 10.1038/s41598-022-16741-y

70. International Parkinson and Movement Disorder Society [Internet] . Tremor Study Group. [cited 2025 Jul 1]. Available from: https://www.movementdisorders.org/MDS/About/Study-Groups/Tremor-Study-Group.htm .

Associated Data

Supplementary Materials

Systematic literature search conducted April 19, 2025.

PICO Criteria defined for article inclusion.

All studies included for article review with their study design.

Complete bias assessment scores for each study.

Patients from included studies divided into tremor types.

Full list of patient data acquired by each article.

Full list of AI models used by each article.