Sage Journals: Discover world-class research

Abstract

Background

Mental health issues like insomnia, anxiety, and depression have increased significantly. Artificial intelligence (AI) has shown promise in diagnosing and providing personalized treatment.

Objective

This study aims to systematically review the application of AI in addressing insomnia, anxiety, and depression, identifying key research hotspots, and forecasting future trends through bibliometric analysis.

Methods

We analyzed a total of 875 articles from the Web of Science Core Collection (2000–2024) using bibliometric tools such as VOSviewer and CiteSpace. These tools were used to map research trends, highlight international collaboration, and examine the contributions of leading countries, institutions, and authors in the field.

Results

The United States and China lead the field in terms of research output and collaborations. Key research areas include “neural networks,” “machine learning,” “deep learning,” and “human-robot interaction,” particularly in relation to personalized treatment approaches. However, challenges around data privacy, ethical concerns, and the interpretability of AI models need to be addressed.

Conclusions

This study highlights the growing role of AI in mental health research and identifies future priorities, such as improving data quality, addressing ethical challenges, and integrating AI more seamlessly into clinical practice. These advancements will be crucial in addressing the global mental health crisis.

Keywords

Artificial intelligence insomnia anxiety depression mental health bibliometrics

Introduction

As global mental health issues worsen, the demand for research and treatment of conditions such as insomnia, anxiety, and depression has dramatically increased. These mental health challenges not only significantly reduce patients' quality of life but also place a considerable burden on society and the economy.^1,2 Insomnia, anxiety, and depression are intricately interconnected, with evidence supporting this relationship on both psychological and physiological levels. Insomnia is recognized as a significant risk factor for both anxiety and depression, and it is also a common symptom of these conditions.³ Research suggests that insomnia affects the hypothalamic-pituitary-adrenal axis, leading to abnormal secretion of stress hormones like cortisol, which can exacerbate or trigger symptoms of anxiety and depression.^4,5 Moreover, insomnia is linked to neurotransmitter imbalances, such as reduced serotonin and dopamine levels, which are strongly associated with the onset of depression.^6,7 Likewise, anxiety and depression can disrupt sleep patterns, worsening insomnia and creating a vicious cycle.^8,9 This interplay between insomnia, anxiety, and depression complicates treatment, emphasizing the need for multidimensional diagnostics and personalized therapeutic strategies.

Artificial intelligence (AI), with its capabilities in data processing, pattern recognition, and predictive analytics, has emerged as a critical tool in mental health interventions.^10–12 Research highlights its vast potential in personalized treatment, real-time monitoring, and risk prediction. For example, AI utilizes multisource data to develop innovative tools for the accurate identification, dynamic prediction, and optimization of mental health interventions, particularly demonstrating significant value in assessing and supporting university students' mental health.¹³ Machine learning algorithms can analyze physiological signals and behavioral patterns to effectively detect mental health risks, such as insomnia.¹⁴ Additionally, the use of social media data offers fresh perspectives for population-level mental health interventions.¹⁵ Bone et al.'s dynamic prediction model, which integrates symptom data to forecast treatment outcomes, provides essential support for refining personalized intervention strategies.¹⁶ By analyzing large-scale, complex clinical datasets, AI can uncover patterns that traditional methods often miss, offering robust scientific support for the early diagnosis and precise intervention of mental health disorders.

However, the application of AI in mental health still faces numerous challenges. First, the performance of AI models relies heavily on high-quality, large-scale annotated datasets, which are difficult to acquire in mental health due to strict privacy and ethical concerns.^17,18 Additionally, the “black box” nature of many AI models makes their decision-making processes hard to interpret, leading to skepticism from both clinicians and patients and limiting AI's clinical utility.^19,20 To address these issues, researchers are exploring explainable AI technologies that aim to enhance the transparency and trustworthiness of these models. Furthermore, there is a growing emphasis on multidisciplinary collaboration to ensure the safe and effective integration of AI in mental health care.

This study applies bibliometric analysis methods^21,22 to systematically review and analyze research on AI's application in insomnia, anxiety, and depression. By analyzing relevant articles published over the past 24 years from the Web of Science database, this study aims to uncover key research trends, development trajectories, and challenges in this area. Specifically, it examines research institutions, countries, authors, keywords, and co-cited references to identify the core research themes and current frontiers in the field. The findings provide valuable insights for advancing AI's application in mental health and outline future directions for research and clinical practice.

Methods

Data sources and search strategy

We conducted a comprehensive search for relevant articles in the Web of Science Core Collection (WoSCC) database. To ensure the accuracy and reliability of the results, two independent researchers were responsible for screening and reviewing the literature. Any disagreements between the two researchers were resolved through discussion, or if necessary, by consulting a third researcher to reach a consensus. The criteria for article retrieval were as follows: (1) Search strategy: The search terms were determined by the TS (“subject”, including title, abstract, author keywords, and keywords Plus). Specific content is TL = (“insomnia” OR “agrypnia” OR “sleeplessness” OR “aypnia” OR “hyposomnia” OR “anxiety” OR “anxiety disorder” OR “anxiety neurosis” OR “depression” OR “depressive disorder”) AND TL = ("deep learning” OR “machine learning” OR “neural network” OR “support vector machine” OR “decision tree” OR “k-nearest neighbor” OR “data mining” OR “feature extraction” OR “graph mining” OR “reinforcement learning” OR “supervised learning” OR “unsupervised clustering” OR “multimodal learning” OR “large language model” OR “knowledge graph” OR “graphical model” OR “Bayesian network” OR “robotics” OR “image segmentation” OR “expert system” OR “computational intelligence” OR “artificial intelligence”); (2) The publication types were limited to “article” or “review”; (3) The publication dates ranged from 2000 to 2024; (4) The publication language was restricted to English; (5) The collected information included publication details, authors, countries, institutions, journals, keywords, and other relevant data. The data collection and retrieval strategy are shown in Figure 1.

Research ethics

The data for this study were obtained from the publicly available WoSCC database. All data were acquired through legitimate channels and in accordance with the database's terms of use. As the data are anonymized and publicly accessible, they do not involve personal privacy and do not require ethical approval or informed consent.

Statistical analysis

This study employed bibliometric methods and integrated three tools—VOSviewer (v1.6.20), CiteSpace (v6.3.R1), and Scimago Graphica (v1.0.42)—to systematically analyze and visualize collaboration networks, research hotspots, and knowledge evolution in the field.²³ VOSviewer was used to reveal collaboration patterns and key nodes among countries, institutions, and authors.²⁴ CiteSpace focused on keyword co-occurrence analysis, citation bursts, and timeline visualizations to track research hotspots and frontier dynamics. Scimago Graphica enhanced the visual presentation by optimizing chart designs and improving interpretability.²⁵

Annual growth and academic collaboration

The growth trends of the research field were analyzed based on annual publication output. Collaboration networks, constructed using VOSviewer, provided insights into the intensity of international collaborations, the central roles of key institutions, and the academic contributions of influential authors.

Keyword analysis

CiteSpace was employed to conduct keyword co-occurrence and citation burst analyses, identifying emerging research directions within the field. Through timeline mapping, it uncovered the evolutionary patterns and shifting trajectories of research hotspots, providing a comprehensive view of the phased development of key themes.

Reference analysis

CiteSpace was employed to perform citation burst detection and timeline view analysis, identifying highly cited references and key academic breakthroughs while tracing the evolutionary trajectory of the knowledge structure.

Results

Trends in annual publications

Based on the search strategy, we retrieved a total of 875 articles from the WoS database, covering a span of 24 years. The trend of annual publications is shown in Figure 2. During the first decade (2000–2010), the number of publications per year in this field remained below 25. In the subsequent eight years (2011–2018), the annual publication count exhibited a fluctuating but overall increasing trend.

Figure 1.

The data collection and retrieval strategy. The flowchart illustrates the search process, exclusion criteria, and subsequent bibliometric analysis conducted on the selected publications.

Figure 2.

Global publication trends in artificial intelligence (AI)-based research on insomnia, anxiety, and depression. The blue line represents the actual number of publications, while the red line denotes the exponential fitting trend, illustrating the growth trajectory.

Distribution of countries and institutions

A total of 74 countries and 1643 institutions have published works on this topic (Table 1). An analysis of collaboration networks among countries indicates that international cooperation in this field is relatively active, with the United States and China occupying leading positions in global partnerships and maintaining close and stable collaborations with other countries (Figure 3).

Figure 3.

Distribution and co-authorship network of countries/regions. The size of each node represents the research output and impact of a country or region, while the lines indicate collaborative relationships. The thickness of the lines reflects the strength of these collaborations, and different colors distinguish various collaboration clusters.

Table 1.

Top 10 countries by total number of publications.

Rank	Country	Total publications	Total link strength	Total citations
1	China	274	55	3320
2	United States	214	132	5004
3	United Kingdom	71	73	1779
4	India	66	38	977
5	South Korea	65	16	638
6	Canada	59	52	1065
7	Australia	54	53	1209
8	Germany	46	50	1260
9	Kingdom of Saudi Arabia	35	24	418
10	Italy	34	32	818

Note: China ranks first with 274 publications, followed by the United States (214) and the United Kingdom (71). The United States leads in total citations (5004) and total link strength (132), highlighting its academic influence in this field.

To explore the core institutions and their collaborative relationships in this field, we established a co-authorship network of institutions. As shown in Table 2, the University of Toronto has published the highest number of papers (n = 22), followed by the University of Pittsburgh (n = 17). These institutions have received 622 and 552 citations, respectively, underscoring their strong influence in this research area. Using VOSviewer, we constructed a co-authorship network among institutions (Figure 4). The analysis indicates that the University of Toronto, the Chinese Academy of Sciences, and McMaster University are central to the collaboration network, actively collaborating with other institutions.

Figure 4.

The co-authorship network of institutions. Node size represents the number of publications, line thickness indicates the strength of collaboration, and colors differentiate the research groups.

Table 2.

Top 10 institutions by number of publications.

Rank	Institution	Country	Total publications	Total citations	Total link strength
1	University of Toronto	Canada	22	622	21
2	University of Pittsburgh	United States	17	552	8
3	King's College London	United Kingdom	14	486	4
4	Chinese Academy of Sciences	China	14	398	10
5	Harvard Medical School	United States	14	364	5
6	Southeast University	China	14	136	3
7	Lanzhou University	China	13	247	7
8	University of Melbourne	Australia	12	513	3
9	McMaster University	Canada	12	314	10
10	Capital Medical University	China	12	270	6

Analysis of authors and co-authorship networks

A total of 4156 authors have contributed to research in this field. Citation count is a key metric for assessing academic influence, and Table 3 lists the top ten authors ranked by citation count. Uddhav Rajendra Acharya holds the highest number of citations (413), followed by Reza Rostami (374) and Paolo Brambilla (330). Among these, Paolo Brambilla has the highest number of published papers (n = 7). The co-authorship network among authors was analyzed (Figure 5(a)), revealing that both Paolo Brambilla and Reza Rostami have a substantial number of collaborators and maintain close cooperation with other authors(Figure 5(b) and (c)). However, overall, the scale of collaboration among authors is relatively small, indicating that the overall connectivity among researchers in this field could be strengthened.

Figure 5.

(a) the co-authorship network of authors. (b) Enlarged View of Reza Rostami's Network (Pink Group). (c) Enlarged View of Paolo Brambilla's Network (Blue Group). Lines represent collaboration between authors, while different colors indicate distinct collaboration groups.

Table 3.

Top 10 authors by number of citations.

Rank	Author	Country	Total citations	Total publications	Total link strength
1	Uddhav Rajendra Acharya	India	413	5	0
2	Reza Rostami	Iran	374	6	8
3	Paolo Brambilla	Italy	330	7	0
4	Joseph Kambeitz	Germany	250	3	8
5	Nikolaos Koutsouleris	Germany	250	3	8
6	Peter Falkai	Germany	233	3	7
7	Xiaowei Li	China	201	4	5
8	Alice Othmani	France	188	5	3
9	Lang He	China	181	3	0
10	Roger C. M. Ho	Singapore	181	3	0

A total of 365 journals have published articles on the application of AI in insomnia, anxiety, and depression. In terms of publication volume (Table 4), the top three journals are the Journal of Affective Disorders (60 publications), Frontiers in Psychiatry (39 publications), and PLoS One (24 publications). Among the top 10 journals with the most publications, Psychological Medicine (Impact Factor—IF = 5.9) has the highest IF. Citations are a key indicator of a journal's academic influence. The most highly cited journal is the Journal of Affective Disorders (744 citations), followed by Frontiers in Psychiatry (396 citations) and Psychological Medicine (337 citations).

Table 4.

Top 10 most productive journals.

Rank	Journal	Total publications	Total citations	Impact factor (2024)	JCR (2024)
1	Journal of Affective Disorders	60	744	4.9	Q1
2	Frontiers in Psychiatry	39	396	3.2.	Q2
3	PLoS One	24	306	2.9	Q1
4	Scientific Reports	19	178	3.8	Q1
5	IEEE Access	16	147	3.4	Q2
6	Journal of Psychiatric Research	15	164	3.7	Q1
7	Psychological Medicine	13	337	5.9	Q1
8	Biomedical Signal Processing and Control	13	171	4.9	Q1
9	Psychiatry Research	12	74	4.2	Q1
10	Multimedia Tools and Applications	12	45	3	Q2

Analysis of references

The number of citations a publication receives reflects its significance within the field. Appendix A lists the top 7 most-cited articles. The most cited article is by Kroenke, K., published in 2001, which validated the Patient Health Questionnaire-9 (PHQ-9) as a brief and effective scale for assessing depression severity.²⁶ The second most cited article is by Acharya, U. R., published in 2018, which employed deep convolutional neural networks (CNNs) for automated depression screening using electroencephalogram (EEG) data.²⁷

Figure 6 highlights the top 20 references with the strongest citation bursts, with the most significant being a 2016 study by Chekroud AM.²⁸ This study utilized machine learning to predict the effectiveness of depression treatments, optimizing personalized treatment strategies. The timeline view of co-cited references illustrates the evolution of research hotspots (Figure 7), with the latest topics including “Convolutional Neural Networks” (#0), “Depression Detection” (#1), “Machine Learning” (#2), and “Deep Learning” (#3). This highlights the increasing application of AI technologies in the diagnosis and treatment of depression and other mental health disorders.

Figure 6.

Top 22 references with the strongest citation bursts. Burst strength indicates a significant increase in citation frequency for a publication during a specific time period. The timeline shows citation bursts from 2000 to 2024.

Figure 7.

Co-cited references timeline view. The size of each node represents the annual frequency of a keyword, with larger nodes indicating higher frequency. Dense areas represent keyword clusters, reflecting major research directions. The numbers and labels on the right indicate cluster IDs and themes.

Analysis of keyword

Keyword analysis is instrumental in identifying important topics and research hotspots within a field. Keyword co-occurrence refers to the occurrence of two or more keywords appearing together in the same or related publications. Figure 8(a) presents the keyword co-occurrence network, showing that “machine learning,” “depression,” and “anxiety” have the strongest overall connection strengths, indicating their central role as key focal points in this area of research.

Figure 8.

(a) the network map of keyword co-occurrence. The size of each node represents the frequency of a keyword, while dense regions indicate keyword clusters, revealing major research directions. (b) The top 22 keywords with the strongest citation bursts. The timeline on the right displays burst periods from 2000 to 2024, with red bars indicating periods of keyword bursts. (c) The timeline map of keyword co-occurrence. Dense regions indicate keyword clusters, highlighting the primary research directions.

CiteSpace was used to analyze citation bursts for keywords. Figure 8(b) displays the top 20 keywords with the highest burst strength, serving as key indicators of emerging research hotspots in this field. Among them, “feature extraction” is currently experiencing a burst, underscoring its significance in recent studies. The keyword timeline view illustrates the trends in research hotspots over time. In the timeline view (Figure 8(c)), all keywords are grouped into 14 clusters. The clusters “deep learning” (#0), “symptom” (#1), “postpartum depression” (#2), and “major depressive disorder” (#3) span almost the entire timeline, indicating their consistent importance within the field. In contrast, keywords like “artificial intelligence” (#8), “human–robot interaction” (#11), and “neural network” (#7) have gained prominence in recent years, highlighting current research trends.

Discussion

Global trends and Status of publications

This study provides a comprehensive review of articles published over the past 24 years using the WoSCC database, highlighting significant advancements in the application of AI in insomnia, anxiety, and depression. Notably, the past 5 years have seen a rapid increase in publications within this field. The study analyzed a total of 876 articles, authored by 4156 researchers from 1643 institutions across 74 countries/regions, published in 365 journals, and citing 35,713 co-cited references.

China (274 articles) and the United States (214 articles) lead in the number of publications, positioning them as the primary contributors to research on AI in insomnia, anxiety, and depression. Both countries are at the forefront of international collaboration, maintaining strong connections with other nations. The United States holds the highest total citation count (5004), underscoring its significant influence in the medical community. The University of Toronto stands out as the most productive and most cited institution. Furthermore, both the University of Toronto and the Chinese Academy of Sciences occupy central positions in the institutional co-authorship network, underscoring their pivotal roles in advancing this field.

Analysis of the co-authorship network revealed that Uddhav Rajendra Acharya's articles have the highest citation count (413), while Paolo Brambilla has published the most articles (7), reflecting their significant academic influence in this field. However, overall collaboration among researchers remains relatively limited, and future efforts should focus on fostering academic cooperation to expand the depth and breadth of research. Journal analysis indicates that the Journal of Affective Disorders has the highest number of publications (60) and citations (744), while Psychological Medicine has the highest IF (IF = 5.9), underscoring the academic significance and reference value of these two journals in the field.

Research hotspots

The analysis of cited references and keywords further highlights the evolution of research hotspots and emerging trends. Key terms such as “neural networks,” “machine learning,” “deep learning,” “feature extraction,” “depression detection,” “artificial intelligence,” and “human-robot interaction” occupy prominent positions in the timeline view, reflecting their widespread application in mental health research and the considerable attention they have garnered from researchers. The current research hotspots in this field primarily concentrate on the following areas.

Machine learning and personalized treatment

In recent years, machine learning has been extensively applied to the classification and prediction of mental health data, emerging as one of the core technologies in this field.^29–31 The co-citation analysis of keywords in this study reveals that terms such as “machine learning,” “feature extraction,” “depression,” and “anxiety” frequently appear in the literature and are strongly interconnected, underscoring their importance as key research directions. Feature extraction is a critical step in machine learning. Machine learning techniques, including support vector machines, decision trees, and random forests, are frequently used to identify clinical features in patients with anxiety or depression. By utilizing these technologies, researchers can extract the most diagnostically valuable features from complex biomedical data, thereby enhancing the predictive accuracy of models.^32,33 Personalized treatment represents a key application of machine learning in the mental health field.^34,35 Through the development of machine learning models, researchers can classify patients based on the severity of clinical symptoms and predict disease progression.^36,37 Xu et al.³⁸ investigated the neuroimaging characteristics of the negative emotion network in patients with chronic insomnia and applied machine learning to predict the severity of anxiety symptoms. The results revealed that neural activity in the negative emotion network is closely related to anxiety, offering new insights into the relationship between anxiety and insomnia. Furthermore, machine learning models can predict patients' responses to different treatment options by analyzing data, thus reducing the trial-and-error process and optimizing treatment strategies. Chekroud et al.²⁸ utilized a machine learning model to predict treatment outcomes for depression patients across multiple clinical trials. The study demonstrated that the model effectively predicted treatment responses in diverse clinical settings, highlighting its potential for personalized treatment in depression.

Deep learning and intelligent diagnosis

Deep learning, particularly CNNs, has become a core technology for processing complex medical data, such as EEG and imaging data.^39–41 These techniques have demonstrated excellent performance in the automated diagnosis and prediction of mental health issues, particularly in the early detection and monitoring of depression and anxiety, showcasing significant potential in these areas.^42–45 Among the key studies, Acharya et al.²⁷ published a highly cited article in 2018 titled “Automated EEG-based screening of depression using deep convolutional neural network.” This research developed an automated depression screening system using CNNs to analyze EEG data for detecting depression. The results showed that CNNs significantly improved the accuracy and reliability of depression screening. Additionally, Wang et al.⁴⁶ developed a deep learning framework based on diffusion models for diagnosing major depressive disorder using EEG data. This model effectively extracted key features of the disorder and demonstrated strong classification performance, providing a new tool for the early detection of major depression.

These studies highlight the application value of deep learning in the automated diagnosis of mental health conditions. Multimodal fusion is another important application of deep learning, integrating medical data from different sources into a single model to provide more comprehensive diagnostic information. This approach captures and analyzes complex patterns and relationships that cannot be revealed by a single data source, thus enhancing diagnostic accuracy and improving prediction reliability.^47,48. Liu et al.⁴⁹ developed a system using CNNs to assess depression severity based on facial expressions and body movements. The results demonstrated that the model exhibited high accuracy in predicting depression severity, underscoring the potential of deep learning technology in analyzing nonverbal behavioral features for depression assessment.

Human-Computer interaction and AI-assisted therapy

The emergence and advancement of human–computer interaction (HCI) technologies, particularly in combination with AI-powered smart devices and robots, have shown unprecedented potential in medical research.^50,51 These technologies not only aid in treatment but also enable real-time monitoring and adjustment of treatment plans through patient interaction, offering a more personalized treatment experience. The application of HCI technology in addressing common mental health issues such as insomnia, anxiety, and depression has been steadily increasing, suggesting that it may become a crucial component of mental health treatment in the future.^52,53

In recent years, AI and HCI-based smart devices, such as intelligent sleep monitors, have proven to be effective tools in assisting with the treatment of insomnia. Roberts et al.⁵⁴ conducted a comparative analysis of multisensor consumer wearable devices for sleep detection. The study found that these devices performed well in detecting total sleep time, providing valuable reference data for sleep monitoring in everyday environments. Additionally, HCI-based virtual reality therapy has gained widespread use in the treatment of anxiety disorders, emerging as an effective and innovative psychological therapy tool.^55,56 AI-enhanced virtual reality therapy simulates realistic scenarios, allowing patients to gradually acclimate to anxiety-inducing factors, reduce sensitivity to anxiety triggers, and effectively alleviate anxiety symptoms.⁵⁷

AI-driven chatbots have also introduced new therapeutic approaches for patients with anxiety and depression.^58,59 Fitzpatrick et al.⁶⁰ evaluated the effectiveness of an automated chatbot, Woebot, in delivering cognitive behavioral therapy to university students. The results showed that participants using Woebot experienced a significant reduction in depression and anxiety symptoms within two weeks. As a convenient and user-friendly tool, Woebot provides effective support for managing anxiety and depression symptoms.

Future development trends

With the rapid advancement of machine learning and deep learning technologies, personalized treatment in mental health is expected to see broader and more precise applications.^19,61,62 Future research should prioritize improving the accuracy and robustness of models, particularly through the integration and analysis of multimodal data. By combining data from diverse sources—such as facial expressions, body movements, voice characteristics, EEG, and other physiological signals—researchers can develop more comprehensive and detailed personalized treatment models.⁶³ These models not only enhance the capabilities of automated diagnosis and prediction but also provide more precise and customized treatment plans.

The application prospects for personalized models in treating anxiety and depression are especially promising. By analyzing multidimensional patient data, clinicians can more accurately predict responses to various treatment methods, including medication, psychotherapy, or behavioral therapy. This reduces the trial-and-error process and significantly optimizes treatment outcomes. Furthermore, integrating multiple data sources allows these models to detect early symptoms or risk factors that may be overlooked in traditional screenings, enabling earlier and more effective interventions, and thus improving long-term patient outcomes.

Comprehensive application of human-computer interaction technology in mental health

The maturation of HCI technology has expanded the application of AI-based smart devices and virtual reality therapy in mental health treatment.^64–66 These advancements not only provide patients with a more interactive therapeutic experience but also make the treatment process more convenient and personalized.^67–69 In the future, optimizing the interaction between HCI technology and patients will be crucial to enhancing treatment outcomes. AI-driven autonomous therapeutic tools, such as chatbots and virtual advisors, are increasingly becoming essential support tools in the mental health field. These tools can utilize technologies like natural language processing and sentiment analysis to more accurately assess patients' psychological states and deliver personalized interventions.^70,71

In resource-limited or remote areas, these technologies can be integrated into conventional treatment processes, offering continuous support and guidance to patients and helping address the shortage of mental health resources. Future research should focus on optimizing the interaction design and algorithms of these AI tools to improve their ability to recognize complex emotions and psychological states, thereby providing more accurate psychological interventions and emotional support. Overall, the continuous advancement of HCI technology will further broaden the scope of mental health treatment, making it more accessible, efficient, and personalized, ultimately improving patients' quality of life.

Focus on ethics and privacy

As AI technology becomes more widely applied in the mental health fields of insomnia, anxiety, and depression, concerns around data privacy and ethics have increasingly come to the forefront.^72,73 Balancing the protection of patient privacy with the effective utilization of big data and AI for personalized treatment remains a key challenge for future research. Potential measures include anonymizing data, implementing data encryption technologies, and strengthening access control mechanisms.^17,74,75.

Moreover, the application of AI technology raises ethical risks related to algorithmic bias and automated decision-making. Future research must ensure algorithm transparency and fairness, with regular reviews and updates to algorithms to prevent harm to patient rights and ensure that these technologies contribute meaningfully to improving mental health. Striking a balance between technological innovation and ethical safeguards will be crucial for the sustainable development of the mental health field, particularly in addressing insomnia, anxiety, and depression.

Limitations of this study

This study provides a comprehensive review of the application of AI in insomnia, anxiety, and depression through bibliometric analysis, but several limitations should be noted. First, the study only included English-language literature, potentially leading to language bias by excluding relevant studies published in other languages. Secondly, our analysis relied primarily on the WoSCC database. Although this database contains a large collection of high-quality international literature, exclusive reliance on it may result in the omission of influential studies from other databases. Additionally, the bibliometric methods employed in this study are based on indicators such as citation counts, which may not fully capture the academic value and broader impact of the research. Despite these limitations, this study offers valuable insights and guidance through a comprehensive and systematic analysis of AI applications in the fields of insomnia, anxiety, and depression.

Conclusion

In summary, this study is the first to comprehensively review publications related to the application of AI in insomnia, anxiety, and depression over the past 24 years using bibliometric analysis. The findings indicate that AI's role in addressing mental health issues is rapidly expanding, with significant potential in personalized treatment and intelligent diagnosis for these conditions. Machine learning and deep learning technologies are prominent research hotspots, widely employed to extract key features from complex data, optimize treatment plans, and enhance diagnostic accuracy. Multimodal data fusion offers more comprehensive support for the diagnosis and treatment of mental health disorders.

Looking ahead, the continued development of personalized treatment and HCI technology will be central research directions. However, as AI technology becomes more pervasive, concerns surrounding data privacy and ethics are increasingly pressing. Future efforts should focus on mitigating algorithmic bias and reducing the risks associated with automated decision-making, ensuring the robust and sustainable development of AI in the field of mental health.

Supplemental Material

sj-docx-1-dhj-10.1177_20552076251324456 - Supplemental material for The application of artificial intelligence in insomnia, anxiety, and depression: A bibliometric analysis

Supplemental material, sj-docx-1-dhj-10.1177_20552076251324456 for The application of artificial intelligence in insomnia, anxiety, and depression: A bibliometric analysis by Enshi Lu, Di Zhang, Mingguang Han, Shihua Wang and Liyun He in DIGITAL HEALTH

Footnotes

Availability of data and materials

The original data from the current study are available from the corresponding author upon reasonable request.

Contributorship

ESL conceived and designed the study, analyzed and interpreted the data, and wrote the manuscript. LYH responsible for the study design and instructions. DZ performed the data acquisition and analysis, and contributed to writing the manuscript. MGH participated in the data analysis and discussion of the results. SHW analyzed the results of the study and revised the manuscript.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

No patients were involved in this research, so the ethical approval is not required. The data sets analyzed during this study are available in the Web of Science core collections.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Capital Health Development Research Special Project (grant number: 2022-1-4301).

Guarantor

ESL and LYH

ORCID iD

Enshi Lu

Di Zhang

Mingguang Han

Shihua Wang

Supplemental material

Supplemental material for this article is available online.

References

GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: A systematic analysis for the global burden of disease study 2019. Lancet 2020; 396: 1204–1222.

Gustavson

Knudsen

Nesvåg

, et al. Prevalence and stability of mental disorders among young adults: Findings from a longitudinal study. BMC Psychiatry 2018; 18: 65.

Ben Simon

Rossi

Harvey

, et al. Overanxious and underslept. Nat Hum Behav. 2020; 4: 100–110.

Tsigos

Chrousos

. Hypothalamic-pituitary-adrenal axis, neuroendocrine factors and stress. J Psychosom Res 2002; 53: 865–871.

Riemann

Nissen

Palagini

, et al. The neurobiology, investigation, and treatment of chronic insomnia. Lancet Neurol 2015; 14: 547–558.

Palagini

Baglioni

Ciapparelli

, et al. REM Sleep dysregulation in depression: State of the art. Sleep Med Rev 2013; 17: 377–390.

Feld

Born

. Neurochemical mechanisms for memory processing during sleep: Basic findings in humans and neuropsychiatric implications. Neuropsychopharmacology 2020; 45: 31–44.

Riemann

Espie

Altena

, et al. The European insomnia guideline: An update on the diagnosis and treatment of insomnia 2023. J Sleep Res 2023; 32: e14035.

Krystal

Prather

Ashbrook

. The assessment and management of insomnia: An update. World Psychiatry 2019; 18: 337–352.

10.

Topol

. High-performance medicine: The convergence of human and artificial intelligence. Nat Med 2019; 25: 44–56.

11.

Heinz

Bhattacharya

Trudeau

, et al. Testing domain knowledge and risk of bias of a large-scale general artificial intelligence model in mental health. Digit Health 2023; 9: 20552076231170499.

12.

Miotto

Wang

, et al. Deep learning for healthcare: Review, opportunities and challenges. Brief Bioinform 2018; 19: 1236–1246.

13.

Chen

Yuan

Dong

, et al. Artificial intelligence significantly facilitates development in the mental health of college students: A bibliometric analysis. Front Psychol 2024; 15: 1375294.

14.

Huang

. Use of machine learning to identify risk factors for insomnia. PLoS One 2023; 18: e0282622.

15.

Kim

Lee

Park

. Machine learning for mental health in social media: Bibliometric study. J Med Internet Res 2021; 23: e24870.

16.

Bone

Simmonds-Buckley

Thwaites

, et al. Dynamic prediction of psychological treatment outcomes: Development and validation of a prediction model using routinely collected symptom data. Lancet Digit Health 2021; 3: e231–e240.

17.

Price

2nd Cohen

. Privacy in the age of medical big data. Nat Med. 2019; 25: 37–43.

18.

Tang

Fantus

. Medical artificial intelligence ethics: A systematic review of empirical studies. Digit Health 2023; 9: 20552076231186064.

19.

Shatte

ABR

Hutchinson

Teague

. Machine learning in mental health: A scoping review of methods and applications. Psychol Med 2019; 49: 1426–1448.

20.

Rudin

. Stop explaining black box machine learning models for high-stakes decisions and use interpretable models instead. Nat Mach Intell 2019; 1: 206–215.

21.

Wang

Liang

Cao

, et al. Application of artificial intelligence in geriatric care: Bibliometric analysis. J Med Internet Res 2023; 25: e46014.

22.

Zhang

Jin

, et al. Artificial intelligence applied in cardiovascular disease: A bibliometric and visual analysis. Front Cardiovasc Med 2024; 11: 1323918.

23.

Chen

Song

. Visualizing a field of research: A methodology of systematic scientometric reviews. PLoS One 2019; 14: e0223994.

24.

van Eck

Waltman

. Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics 2010; 84: 523–538.

25.

Chen

. Citespace II: detecting and visualizing emerging trends and transient patterns in scientific literature. J Am Soc Inf Sci Technol 2006; 57: 359–377.

26.

Kroenke

Spitzer

Williams

. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med 2001; 16: 606–613.

27.

Acharya

Hagiwara

, et al. Automated EEG-based screening of depression using deep convolutional neural network. Comput Methods Programs Biomed 2018; 161: 103–113.

28.

Chekroud

Zotti

Shehzad

, et al. Cross-trial prediction of treatment outcome in depression: A machine learning approach. Lancet Psychiatry 2016; 3: 243–250.

29.

Bzdok

Meyer-Lindenberg

. Machine learning for precision psychiatry: Opportunities and challenges. Biol Psychiatry Cogn Neurosci Neuroimaging 2018; 3: 223–230.

30.

Tai

AMY

Albuquerque

Carmona

, et al. Machine learning and big data: Implications for disease modeling and therapeutic discovery in psychiatry. Artif Intell Med 2019; 99: 101704.

31.

Tiffin

Paton

. Rise of the machines? Machine learning approaches and mental health: Opportunities and challenges. Br J Psychiatry 2018; 213: 509–510.

32.

Dwyer

Falkai

Koutsouleris

. Machine learning approaches for clinical psychology and psychiatry. Annu Rev Clin Psychol 2018; 14: 91–118.

33.

Winter

Blanke

Leenings

, et al. A systematic evaluation of machine learning-based biomarkers for major depressive disorder. JAMA Psychiatry 2024; 81: 386–395.

34.

Lin

Lane

. Precision psychiatry applications with pharmacogenomics: Artificial intelligence and machine learning approaches. Int J Mol Sci 2020; 21: 969.

35.

Vetter

Schultebraucks

Galatzer-Levy

, et al. Predicting non-response to multimodal day clinic treatment in severely impaired depressed patients: A machine learning approach. Sci Rep 2022; 12: 5455.

36.

Rajpurkar

Yang

Dass

, et al. Evaluation of a machine learning model based on pretreatment symptoms and electroencephalographic features to predict outcomes of antidepressant treatment in adults with depression: A prespecified secondary analysis of a randomized clinical trial. JAMA Netw Open 2020; 3: e206653.

37.

Maciukiewicz

Marshe

Hauschild

, et al. GWAS-based machine learning approach to predict duloxetine response in major depressive disorder. J Psychiatr Res 2018; 99: 62–68.

38.

Dou

Luo

, et al. Neuroimaging profiles of the negative affective network predict anxiety severity in patients with chronic insomnia disorder: A machine learning study. J Affect Disord 2023; 340: 542–550.

39.

Esteva

Robicquet

Ramsundar

, et al. A guide to deep learning in healthcare. Nat Med 2019; 25: 24–29.

40.

Anwar

Majid

Qayyum

, et al. Medical image analysis using convolutional neural networks: a review. J Med Syst 2018; 42: 226.

41.

Roy

Banville

Albuquerque

, et al. Deep learning-based electroencephalography analysis: A systematic review. J Neural Eng 2019; 16: 051001.

42.

Parsa

Rad

Vaezi

, et al. EEG-based classification of individuals with neuropsychiatric disorders using deep neural networks: Asystematic review of current status and future directions. Comput Methods Programs Biomed 2023; 240: 107683.

43.

Mumtaz

Qayyum

. A deep learning framework for automatic diagnosis of unipolar depression. Int J Med Inform 2019; 132: 103983.

44.

Yildirim

Talo

, et al. Automated depression detection using deep representation and sequence learning with EEG signals. J Med Syst 2019; 43: 205.

45.

Zhong

Ying

, et al. Depressive disorder recognition based on frontal EEG signals and deep learning. Sensors (Basel) 2023; 23: 8639.

46.

Wang

Zhao

Jiang

, et al. DiffMDD: A diffusion-based deep learning framework for MDD diagnosis using EEG. IEEE Trans Neural Syst Rehabil Eng 2024; 32: 728–738.

47.

Zhou

Chen

, et al. Multi-modal diagnosis of Alzheimer's disease using interpretable graph convolutional networks. IEEE Trans Med Imaging 2025; 44: 142–153.

48.

Suk

Lee

Shen

; Alzheimer's Disease Neuroimaging Initiative. Deep ensemble learning of sparse regression models for brain disease diagnosis. Med Image Anal 2017; 37: 101–113.

49.

Liu

Lin

, et al. Measuring depression severity based on facial expression and body movement using deep convolutional neural network. Front Psychiatry 2022; 13: 1017064.

50.

Garcia-Aracil

Zollo

Casals

, et al. New methods of human-robot interaction in medical practice. Comput Methods Programs Biomed 2014; 116: 49–51.

51.

Perez-Pozuelo

Zhai

Palotti

, et al. The future of sleep health: A data-driven revolution in sleep science and medicine. NPJ Digit Med 2020; 3: 42.

52.

de Zambotti

Goldstone

Claudatos

, et al. A validation study of fitbit charge 2™ compared with polysomnography in adults. Chronobiol Int 2018; 35: 465–476.

53.

Walch

Cochran

Forger

. A global quantification of “normal” sleep schedules using smartphone data. Sci Adv 2016; 2: e1501705.

54.

Roberts

Schade

Mathew

, et al. Detecting sleep using heart rate and motion data from multisensor consumer-grade wearables, relative to wrist actigraphy and polysomnography. Sleep 2020; 43: zsaa045.

55.

Wiebe

Kannen

Selaskowski

, et al. Virtual reality in the diagnostic and therapy for mental disorders: A systematic review. Clin Psychol Rev 2022; 98: 102213.

56.

Emmelkamp

PMG

Meyerbröker

. Virtual reality therapy in mental health. Annu Rev Clin Psychol 2021; 17: 495–519.

57.

Maples-Keller

Bunnell

Kim

, et al. The use of virtual reality technology in the treatment of anxiety and other psychiatric disorders. Harv Rev Psychiatry 2017; 25: 103–113.

58.

Fulmer

Joerin

Gentile

, et al. Using psychological artificial intelligence (Tess) to relieve symptoms of depression and anxiety: Randomized controlled trial. JMIR Ment Health 2018; 5: e64.

59.

Liu

Peng

Song

, et al. Using AI chatbots to provide self-help depression interventions for university students: A randomized trial of effectiveness. Internet Interv 2022; 27: 100495.

60.

Fitzpatrick

Darcy

Vierhile

. Delivering cognitive behavior therapy to young adults with symptoms of depression and anxiety using a fully automated conversational agent (Woebot): A randomized controlled trial. JMIR Ment Health 2017; 4: e19.

61.

Luedtke

Sadikova

, et al. Development and validation of a machine learning individualized treatment rule in first-episode schizophrenia. JAMA Netw Open 2020; 3: e1921660.

62.

Saqib

Khan

Butt

. Machine learning methods for predicting postpartum depression: Scoping review. JMIR Ment Health 2021; 8: e29838.

63.

Jiang

Zhi

, et al. Artificial intelligence in healthcare: past, present and future. Stroke Vasc Neurol 2017; 2: 230–243.

64.

Nahavandi

Alizadehsani

Khosravi

, et al. Application of artificial intelligence in wearable devices: Opportunities and challenges. Comput Methods Programs Biomed 2022; 213: 106541.

65.

Sanchez-Iborra

. LPWAN And embedded machine learning as enablers for the next generation of wearable devices. Sensors (Basel) 2021; 21: 5218.

66.

Apoorva

Nguyen

Sreejith

. Recent developments and future perspectives of microfluidics and smart technologies in wearable devices. Lab Chip 2024; 24: 1833–1866.

67.

Etzelmueller

Vis

Karyotaki

, et al. Effects of internet-based cognitive behavioral therapy in routine care for adults in treatment for depression and anxiety: Systematic review and meta-analysis. J Med Internet Res 2020; 22: e18100.

68.

Gaffney

Mansell

Tai

. Conversational agents in the treatment of mental health problems: Mixed-method systematic review. JMIR Ment Health 2019; 6: e14166.

69.

Wallin

Norlund

Olsson

EMG

, et al. Treatment activity, user satisfaction, and experienced usability of internet-based cognitive behavioral therapy for adults with depression and anxiety after a myocardial infarction: Mixed-methods study. J Med Internet Res 2018; 20: e87.

70.

Morris

Kouddous

Kshirsagar

, et al. Towards an artificially empathic conversational agent for mental health applications: system design and user perceptions. J Med Internet Res 2018; 20: e10148.

71.

Lindner

Hamilton

Miloff

, et al. How to treat depression with low-intensity virtual reality interventions: Perspectives on translating cognitive behavioral techniques into the virtual reality modality and how to make anti-depressive use of virtual reality-unique experiences. Front Psychiatry 2019; 10: 792.

72.

Larson

Magnus

Lungren

, et al. Ethics of using and sharing clinical imaging data for artificial intelligence: A proposed framework. Radiology 2020; 295: 675–682.

73.

Morley

Floridi

Kinsey

, et al. From what to how: An initial review of publicly available AI ethics tools, methods and research to translate principles into practices. Sci Eng Ethics 2020; 26: 2141–2168.

74.

Fiske

Henningsen

Buyx

. Your robot therapist will see you now: Ethical implications of embodied artificial intelligence in psychiatry, psychology, and psychotherapy. J Med Internet Res 2019; 21: e13216.

75.

Char

Shah

Magnus

. Implementing machine learning in health care: Addressing ethical challenges. N Engl J Med 2018; 378: 981–983.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

The application of artificial intelligence in insomnia,anxiety,and depression: A bibliometric analysis

Abstract

Background

Objective

Methods

Results

Conclusions

Keywords

Introduction

Methods

Data sources and search strategy

Research ethics

Statistical analysis

Annual growth and academic collaboration

Keyword analysis

Reference analysis

Results

Trends in annual publications

Distribution of countries and institutions

Analysis of authors and co-authorship networks

Analysis of references

Analysis of keyword

Discussion

Global trends and Status of publications

Research hotspots

Machine learning and personalized treatment

Deep learning and intelligent diagnosis

Human-Computer interaction and AI-assisted therapy

Future development trends

Comprehensive application of human-computer interaction technology in mental health

Focus on ethics and privacy

Limitations of this study

Conclusion

Supplemental Material

sj-docx-1-dhj-10.1177_20552076251324456 - Supplemental material for The application of artificial intelligence in insomnia, anxiety, and depression: A bibliometric analysis

Footnotes

Availability of data and materials

Contributorship

Declaration of conflicting interests

Ethical approval

Funding

Guarantor

ORCID iD

Supplemental material

References

Supplementary Material