This study used a scoping review methodology to systematically map and synthesize the existing empirical literature on technology-integrated art therapy. The review was conducted in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines [34].

A comprehensive search of peer-reviewed papers was conducted from January 1, 2016, through August 16, 2026. The search was limited to peer-reviewed journal papers published in English over the past decade to reflect contemporary technological developments and current clinical and research practices. A computer-based method was used to locate studies through 9 bibliographic databases and electronic search sources via the University of Haifa Library interface (MEDLINE, APA PsycINFO, ERIC, PsycNET, Cochrane Central Register of Controlled Trials, CINAHL, IEEE Xplore, Web of Science, and AgeLine) to identify peer-reviewed studies examining the use of digital and technological modalities in art therapy and art psychotherapy in which technology constituted a substantive component of the therapeutic process, rather than a peripheral or administrative tool. Using both terms, “art therapy” and “art psychotherapy,” broadens the search scope and reflects variations in professional terminology across countries, disciplines, and publication contexts.

Consistent with the multidimensional scope of art therapy practice, which extends beyond psychotherapy to include assessment, rehabilitation, prevention, wellness, and social action [35], this review included technology used in therapeutic interventions, assessment, rehabilitation, system development supporting therapeutic or assessment processes, and art therapists’ professional practices and perspectives. In addition, for the purposes of this review, we define digital art therapy as the use of digital technologies within an art-therapeutic context or in ways that directly support art-therapy practice, including digital art-making, creative self-expression, art-based assessment, therapist-mediated art therapy, or the development of technologies explicitly designed to facilitate these processes. Studies in which artistic or visual content functioned solely as a passive stimulus, aesthetic feedback, or generic emotion-regulation tool, without an art-making, expressive, assessment, or art-therapy practice component, as well as studies focusing solely on the education or training of art therapists or students, were excluded. Digital technology was defined as technologies that substantially mediate, augment, transform, or enable art-therapy practice beyond the conventional use of photographic or video media as artistic materials.

To ensure conceptual specificity, the search strategy required that art therapy or art psychotherapy be explicitly referenced in the publication title and/or abstract. Terms related to digital and technological modalities were searched in the title and/or abstract to capture a broad range of technology-mediated practices, including, but not limited to, digital media, computers, mobile devices, teletherapy, VR, and AI. The final search string combined technology-related terms with art therapy-specific terms in the title and/or abstract fields depending on the structure of each database and using Boolean operators as follows: ("technology" OR "computer" OR "tablet" OR "mobile phone" OR "smartphone" OR "internet" OR "digital" OR "teletherapy" OR "virtual reality" OR "artificial intelligence" OR "WhatsApp" OR "online therapy" OR "telehealth" OR "videoconferencing" OR "video conferencing" OR "augmented reality" OR "mixed reality" OR "immersive technology" OR "machine learning" OR "digital art") AND ("art therapy" OR "art psychotherapy"). The search yielded 697 records. Overall, papers were identified as follows: 188 from Web of Science, 291 from PsycINFO, 106 from MEDLINE, 64 from CINAHL, 24 from PsycNET, 11 from AgeLine, 10 from ERIC, and 3 from IEEE Xplore. No records were identified in the Cochrane Central Register of Controlled Trials.

Duplicated records were removed using normalized title matching, followed by an additional near-duplicate screening to capture records with minor title variations across databases. This process resulted in the removal of 232 duplicate records, leaving 465 records for title and abstract screening. The first 2 researchers, who are expert art therapists, conducted this stage collaboratively, screening titles and abstracts using predefined inclusion criteria requiring the title and/or abstract to explicitly refer to art therapy or art psychotherapy and to digital or technological modalities. Disagreements were resolved through discussion and consensus.

Publications were eligible for inclusion if they reported empirical qualitative, quantitative, and/or mixed methods studies on the use of technology in art therapy. Specifically, studies were included if they empirically examined the use or clinical integration of technology in art therapy practice, including applications for therapeutic intervention, assessment, or therapists’ reported use of and attitudes toward technology in treatment settings. Records that did not meet these criteria were excluded at this stage, with the rationale systematically documented.

In total, 353 records were excluded for the following reasons: (1) a technological or digital focus without an explicit or sufficient art therapy or therapeutic intervention component (n=226); (2) an art therapy focus without any technological or digital component (n=29); (3) neither art therapy nor technological modalities were addressed (n=62); (4) review, conceptual or practice-based papers without empirical data (n=21); and (5) conference paper/proceedings/corrigendum (n=12); and (6) retracted papers (n=3).

The first 2 reviewers collaboratively evaluated the remaining 112 records. Any disagreements were addressed through discussion and consensus-building. To determine the reproducibility of eligibility determinations, 25 full-text records were randomly selected and independently reassessed by the first 2 authors (Cohen κ=.91, P<.001).

Following this full-text review, records were excluded for the following reasons: (1) a technological or digital focus without an explicit or sufficient art therapy or therapeutic intervention component (n=34); (2) an art therapy focus without any technological or digital component (n=10); (3) neither art therapy nor technological modalities were addressed (n=2); and (4) review, conceptual, or practice-based papers without empirical data (n=19). At the conclusion of this selection process, 65 papers were excluded, leaving 47 empirical papers for extraction and analysis (see Multimedia Appendix 1 [2,3,15,20-22,25,36-74] for the coding of 465 papers).

Relevant information was systematically extracted in accordance with the PRISMA-ScR guidelines [34]. Given the exploratory purpose of the review and the field’s developing and heterogeneous nature, the review protocol was not prospectively registered. Instead, the review was designed to map the available evidence rather than evaluate a narrowly defined clinical intervention or treatment effect. However, eligibility criteria, databases, and search procedures were specified in advance and applied systematically by the first 2 authors collaboratively in accordance with PRISMA-ScR guidelines, as presented below, using a structured charting form that was pilot-tested on a subset of studies and refined iteratively to improve consistency and clarity. The form included the following classifications: (1) authors, year, and country; (2) study name and category; (3) aim; (4) research type (eg, qualitative, quantitative, or mixed methods); (5) setting, location, and duration; (6) technology modality/platform; (7) delivery mode (in-person, remote, or hybrid); (8) target population; (9) age; (10) sample size; (11) study design; (12) analytical approach; (13) outcome domains and measures; and (14) main findings.

In addition, the methodological quality of each study was assessed by the first 2 authors collaboratively using the 2018 version of the Mixed Methods Appraisal Tool (MMAT) [75], which enables evaluation of qualitative, quantitative, and mixed methods studies within a unified framework. Rather than calculating a numerical MMAT score, we used the pattern of methodological strengths and limitations across the applicable MMAT criteria. This approach is consistent with previous work published in JMIR Mental Health, in which MMAT ratings were treated qualitatively rather than converted into a numerical score [76]. Extracted information and classification decisions were reviewed and agreed upon through discussion and consensus (see Table 1 for paper coding).

To address the study questions, descriptive numerical summaries were used to map the characteristics of the included studies, while findings, particularly those derived from qualitative and mixed methods studies, were synthesized using an interpretive approach. The findings were read across studies to identify recurring convergences, tensions, and patterns in reported therapeutic experiences and processes. Through iterative comparison and joint discussion among all 3 authors, these recurring patterns were organized into broader conceptual domains in relation to the review questions. This process was intended to interpret and organize patterns across the literature rather than constitute a formal thematic analysis or establish empirically tested processes [77].

Of the 47 included studies, China contributed the largest number (n=10, 21.3%), followed by the United States (n=8, 17%), South Korea (n=7, 14.9%), Israel (n=5, 10.6%), and Italy and the United Kingdom (n=4 each, 8.5%). One study each was conducted in Canada, the Netherlands, South Africa, Australia, Portugal, and Hong Kong (n=1 each, 2.1%). Three studies were multinational (n=3, 6.4%), involving collaborations between Canada and China, South Korea and China, and Luxembourg and the Netherlands. The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram (Figure 1) delineates the systematic search and screening process.

Of the 47 included studies, qualitative designs constituted the largest single methodological category (n=19, 40.4%), followed by mixed methods studies (n=16, 34%), and quantitative studies (n=12, 25.5%). Sample sizes were generally small and highly variable. Median sample sizes were 44 (IQR 26‐81) for quantitative studies, 44.5 (IQR 14.5‐100.3) for mixed methods studies, and 8 (IQR 3.3‐14.3) for qualitative studies among studies reporting identifiable participant counts. Eight (17%) studies used case-study, case-report, or case-series designs, further reflecting the exploratory nature of much of the literature.

Methodological quality varied across study designs. Qualitative studies generally showed strong alignment between RQs, data collection, analysis, and interpretation, although reporting of analytic coherence was less consistent. Randomized and nonrandomized quantitative studies commonly reported complete outcome data and appropriate measurements, but limitations were frequent in randomization procedures, baseline comparability, blinding, representativeness, and control of confounding. Quantitative descriptive studies were similarly limited by sampling and representativeness. In mixed methods studies, qualitative and quantitative elements are commonly documented concurrently; however, the methodological rigor often varies, and explicit integration across these components is inconsistently reported.

The studies were classified into 4 main categories based on their objectives rather than their design. The majority examined technology-enhanced art therapy interventions (n=28, 59.6%). Four (8.5%) studies focused on technology-assisted art-based rehabilitation, primarily in neurological populations such as patients with stroke or spinal cord injury. Ten (21.3%) studies explored therapists’ perspectives on integrating technology into art therapy, while 5 (10.6%) focused on system development supporting creation or assessment. These findings suggest that although intervention-focused studies predominated, the field also encompasses a broader ecosystem, including rehabilitation, system development, and professional perspectives.

Twenty-eight (59.6%) studies involved in-person use of technology, 11 (23.4%) used remote delivery, and 5 (10.6%) used a hybrid format. In 3 (6.4%) studies, the mode of delivery was either not reported or not applicable.

Across the 47 studies, digital art platforms and tools were the most frequently represented technological modality (n=15, 31.9%), including tablets, smartphones, digital drawing and coloring applications, and other computer-based art-making tools. AI (n=13, 27.7%) and VR (n=12, 25.5%) were also prominent. AI applications included generative image systems, human-AI cocreation platforms, explainable assessment systems, deep-learning models, and recommendation systems, whereas VR applications included immersive art-making environments, VR painting and sculpting systems, and rehabilitation platforms. Telehealth and online communication technologies accounted for 12 (25.5%) studies, including videoconferencing, messaging, and other remote-delivery platforms. Five (10.6%) studies involved other technologies, including gesture-based interactive systems, physiological or neuroimaging technologies used alongside digital art tasks, accessibility technologies such as augmentative and alternative communication and eye-gaze interfaces, and metaverse or other multimodal interactive environments. As some studies incorporated more than one technological modality, these categories were not mutually exclusive.

The samples varied widely, with some involving multiple target groups. For descriptive purposes, studies were grouped by primary target population. Specifically, art therapists and trainees/students were the most frequently studied population (n=13, 27.7%), which suggests a growing focus on practitioners’ experiences and perspectives on technology integration. Clinical adult populations, including individuals with neurological, medical, or psychological conditions or undergoing rehabilitation, were represented in 12 studies (n=12, 25.5%). Children and adolescents, often in clinical or at-risk contexts such as autism spectrum disorder or attention-deficit/hyperactivity disorder (ADHD), were represented in 9 studies (n=9, 19.1%). Nonclinical adult populations, including students and community samples, were also represented in 8 studies (n=8, 17%). Older adults, including individuals with cognitive decline and chronic conditions, were represented in 4 studies (n=4, 8.5%). The remaining study (n=1, 2.1%) involved system development without a direct target population. Within these categories, trauma-focused populations were represented in 3 studies: veterans with posttraumatic stress disorder [36], adults with work-related injuries and psychological distress [37], and female adolescents with complex trauma [38].

The findings reported in the quantitative and mixed methods research fell into two categories: (1) preliminary results related to clinical and functional outcomes; and (2) results related to feasibility, engagement, and user experience. Of the 12 quantitative studies, 6 used randomized controlled or randomized experimental designs [25,39-43]. The remaining 6 used nonrandomized or other designs, including a within-participant repeated-measures intervention [44], a comparative mixed-design study [45], an observational single-group clinical study [46], a cross-sectional survey [47], a comparative modeling study [48], and a system-development/evaluation study [49]. Randomization was also reported in 3 of the 16 mixed methods studies [37,50,51], whereas none of the qualitative studies involved randomized allocation.

These studies reported improvements in cognitive functioning [25], mental health [25,37,50,52,53], parent-child relationships and maternal psychological well-being [39], and emotional regulation [41]. Several neurological rehabilitation studies reported functional and motor improvements [40,46,54], whereas Franzò et al [45] primarily supported the usability and feasibility of the VR-based rehabilitation system. Zhang et al [21] reported reductions in ADHD-related symptoms, while Sion et al [42] reported stress reduction following digital and traditional art-making.

Importantly, these findings were based on heterogeneous designs that often lacked a control group, and several studies focused primarily on feasibility, usability, or user experience rather than clearly defined clinical outcomes [6,45,47,50,55,56,78]. Even among the randomized and comparative studies, sample sizes were generally modest, and considerable heterogeneity in outcome measures further limited comparability across studies [25,39,40,42,45,54].

Most qualitative findings reported perceived client benefits associated with using digital media in art therapy, with engaging in digital art described as supporting deeper emotional expression, self-exploration, emotional relief, well-being, and reduced anxiety related to artistic performance [3,16,57]. Note that these findings reflected subjective experiences rather than empirically established clinical effectiveness. A similar pattern emerged for the mixed methods studies. In Newland et al [56], videoconference-based delivery was associated with high feasibility, acceptability, and positive participant experiences, as evidenced by qualitative reports of engagement and satisfaction. However, this study focused primarily on feasibility and delivery rather than deeper therapeutic processes, such as the meaning of artistic expression, the therapeutic relationship, and the therapist’s role beyond facilitation. In Kim [57], the use of familiar digital tools was described as supporting engagement and self-exploration. Similarly, Zhang et al [21], in a study of an AI-assisted art therapy intervention, reported reductions in attention-deficit/hyperactivity symptoms, along with qualitative observations of enhanced engagement and emotional expression. Overall, these findings indicate high feasibility and engagement but provide mixed evidence on clinical effectiveness.

The mixed methods studies suggested several potential contributions of technology to art therapy. VR sculpting was associated with reduced anxiety and negative affect, increased positive affect, and physiological relaxation, while participants described immersion, focused attention, and emotional release [44]. Likewise, VR art therapy was associated with reduced depressive symptoms and negative affect and increased positive affect, alongside participants’ descriptions of movement from grief and isolation toward hope and symbolic integration [58]. AI-assisted approaches were associated with reduced anxiety and emotion-regulation difficulties and were experienced as providing real-time feedback, emotional expression, engagement, and personalized support [50], while another AI-assisted intervention showed greater reductions in anxiety and posttraumatic stress disorder symptoms and was perceived as emotionally supportive and engaging [37].

Among participants with borderline personality disorder, VR-based digital art was associated with a significant reduction in anxiety, while qualitative responses emphasized the usefulness of therapist-supported reflection [52]. AI-assisted painting with children with ADHD was associated with reduced ADHD symptoms and increased engagement and emotional expression [21], whereas a human-AI recommendation system produced mood improvements comparable to expert recommendations while reducing therapist workload, although human oversight remained necessary [6].

Nevertheless, these findings remain preliminary because the mixed methods evidence was methodologically uneven. Several studies relied on very small or single-case samples, including a single-participant VR art therapy case in which concurrent traditional art therapy made it impossible to isolate the contribution of VR [58], and a VR intervention for participants with borderline personality disorder that included only 8 participants and no control group [52]. Other studies used nonrandomized or weak comparison designs: the VR sculpting study compared the same participants across control and intervention phases without randomization or a separate control group [44], the AI-assisted symbolic healing study did not report how participants were assigned to the intervention conditions [50], and the AI-assisted art therapy study was described as quasi-experimental but used block randomization [37]. The ADHD study also relied on a small single-group prepost design, limiting causal interpretation of the reported symptom reduction and increased engagement [21].

Across studies that combined quantitative outcomes with qualitative or open-ended material, limitations included weakly developed qualitative components or a lack of a clearly specified systematic qualitative analytic procedure [21,44,52]. Among mixed methods studies, integration between quantitative and qualitative components was often incompletely reported or difficult to evaluate [21,50,52,58]. Thus, although the mixed methods studies provided useful indications of both potential outcomes and participants’ experiences of technology, their designs often limited their ability to explain how or why technology contributed to the observed changes.

In addition to the outcome studies, several works examined technology’s contribution to rehabilitation and system development. Four studies examined technology-assisted art-based rehabilitation, primarily using immersive VR. These studies were conducted in neurological and physical rehabilitation centers to improve motor and functional rehabilitation, primarily in patients with stroke or spinal cord injury [40,45,46,54]. The results suggested that participation in VR-based art activities was associated with motor and functional improvements, including upper-limb function, movement fluidity, and activities of daily living. They also found high levels of engagement, usability, and treatment participation, suggesting that increased patient engagement may be one possible pathway through which virtual art therapy contributes to rehabilitation; however, this pathway was not directly tested.

Five studies dealt with the development and evaluation of digital systems [48,49,51,59,60]. These studies examined the extent to which AI-based or interactive platforms could facilitate digital artmaking, support therapist-client interactions, and/or contribute to assessment. The findings covered usability, user experience, accessibility, and system performance, and appeared to indicate that these systems enabled human-AI interaction, supported creative processes, and, in some cases, allowed therapists to monitor progress or extend therapeutic activity outside sessions [49,59]. Across these studies, limitations included questions regarding the reliability and interpretability of AI-based assessments in studies that used them, the need for integration with clinical expertise, and the need for future confirmatory work with larger datasets.

Ten studies explored art therapists’ perspectives on integrating technology into art therapy, primarily using qualitative designs. Several consistent themes emerged across these studies. The most prominent theme, reported in most studies (n=9), was cautious openness toward technology. Therapists recognized its potential benefits while emphasizing the importance of maintaining clinical control, professional judgment, and ethical standards, particularly regarding privacy, data security, boundaries, and the potential impact on the therapeutic relationship [2,22,51,61,63,67-70].

At the same time, therapists highlighted technology’s potential to enhance accessibility, flexibility, and client engagement, especially for populations with physical or communicative limitations [22,61,63,69,79]. In addition, studies pointed to changes in the therapeutic process itself, including shifts in presence, embodiment, and the nature of therapist-client interaction in digital or online environments, as well as challenges related to privacy in the therapeutic setting [22,51,68].

Practical and technical challenges were commonly reported [22,63,68,69], including the need for training, technological literacy, and adaptation of therapeutic techniques, alongside concerns regarding the therapist’s role, particularly system limitations, risks of overreliance, and the need to preserve professional judgment and maintain therapist control over assessment and decision-making [70].

One study used an experiential approach, in which therapists engaged directly with VR art-making rather than just reporting on attitudes or clinical use [67]. On the one hand, therapists found VR artistic creation to be immersive, engaging, and easy to use, offering high flexibility and control. It provided unique features such as 3D creation, embodied movement, perspective shifts, and the ability to enter the artwork, which helped clients externalize and explore their experiences while also supporting therapists’ perspective-taking and empathy. However, the lack of physical materials and tactile feedback, the potentially overwhelming immersive nature of VR, and the absence of eye contact, which can lead to detachment and reduced shared presence, were seen as possible drawbacks. VR was not viewed as a replacement for traditional art therapy but as an additional approach suitable for certain populations, especially after establishing a stable therapeutic relationship.

Across the included studies, ethical and safety concerns were mainly related to clinical and relational aspects rather than legal or technological issues. The main concerns reported involved privacy, data security, professional boundaries, therapist oversight, and the risk of overdependence on technology [2,22,69,71]. Specific challenges appeared in immersive and digitally mediated contexts, where overstimulation, distancing, reduced eye contact and shared presence, and attention disruptions and therapeutic containment issues were noted [20,22,67,68]. More broadly, these findings highlight the importance of maintaining therapist and client presence, attunement, and clinical control in technology-mediated therapy settings [64,68,71]. In studies involving AI, trust, explainability, transparency, and authorship were addressed but only in a few cases [2,47,51,71]. Broader issues such as data governance, AI-informed consent, algorithmic bias, legal ownership of AI-generated art, and reliance on commercial platforms were largely missing from the empirical research. Therefore, current ethical discussions mainly focus on clinical and relational aspects, with broader technological and governance concerns still underexplored.

This scoping review explored a growing body of research on technology-enhanced art therapy, which remains methodologically uneven and conceptually underdeveloped. Beyond documenting this expansion, the findings suggest that the field is at an early stage of theoretical consolidation, with empirical exploration outpacing conceptual clarity. Currently, technology may be understood as a mediating layer that reconfigures the conditions, processes, and relational structures enabling change. This discussion recapitulates the populations and methodological patterns that emerged from this review, draws conceptual implications from the findings, and presents future directions to advance the field.

Several consistent patterns emerged from the findings. First, most empirical work focused on adult clinical populations, particularly individuals diagnosed with neurological conditions, cognitive decline, or chronic health issues, as well as children and adolescents in clinical or at-risk contexts [21,40,45,52,54,65]. Relatively few studies examined nonclinical populations, and a substantial proportion of the literature focused on therapists’ perspectives rather than client outcomes [22,61-63,68,69,79]. This distribution suggests that empirical literature largely reflects clinical needs, such as rehabilitation and accessibility, and exploratory efforts to probe practitioners’ engagement with emerging technologies.

The included studies reported potential benefits of technology-based modalities for engagement, accessibility, and, in some cases, functional or emotional outcomes [21,25,50,52,56,59]. Digital and VR-based interventions appeared to reduce barriers related to artistic skill, increase motivation, and enable alternative forms of expression, especially for populations with physical, cognitive, or communicative limitations [3,15,19,63,64]. However, the small-scale and exploratory nature of many of these studies made it impossible to establish the mechanisms underlying these reported effects. Specifically, it remains unclear whether the reported benefits stem from the art-making process itself, technological affordances, or broader contextual factors, such as increased engagement and therapist involvement. Disentangling these contributions requires study designs that systematically vary the type and extent of technological mediation while holding other elements of the therapeutic context constant. In addition, studies did not consistently distinguish between outcome variables, reported therapeutic processes, and underlying mechanisms.

A further distinction emerged between intervention-focused studies and system development research. Whereas the latter introduced innovative platforms and tools for human-AI interaction, co-creation, and therapist support, these systems were typically evaluated through usability testing or small-scale exploratory designs rather than through rigorous clinical trials [48,49,59,60]. Although the emerging literature has noted several advantages of integrating digital technologies, such as enhanced accessibility, reduced barriers, timely therapeutic support, and expanded expressive possibilities [1,2,5,6,62], deeper conceptual tensions remain between the embodied, material-based traditions of art therapy and emerging digital modes of expression that reconfigure the roles of the body, the artwork, and the therapeutic space.

The limited representation of trauma-affected and other complex clinical populations represents an important gap in the current literature. Features such as graded exposure, aesthetic distance, sensory engagement, and control over the pace of interaction may be theoretically relevant for these populations; however, the current evidence is insufficient to determine whether they confer particular therapeutic benefits. Future research should therefore examine these possibilities cautiously within ethically supported clinical protocols.

The disciplinary distribution of the publications included in this review revealed a clear methodological pattern. Of the 47 included studies, 21 (44.7%) were published in art therapy journals, 6 (12.8%) in technology- or computing-oriented journals, and 20 (42.6%) in health, psychology, or interdisciplinary journals. Qualitative studies were concentrated in art therapy journals, with 12 of 19 (63.2%) appearing in these outlets, compared with only 2 of 12 (16.7%) quantitative studies. Mixed methods studies showed an intermediate pattern, with 7 of 16 (43.8%) published in art therapy journals, 4 (25%) in technology-oriented journals, and 5 (31.3%) in other journals. Notably, mixed methods studies published in art therapy journals tended to show stronger qualitative than quantitative components, with the latter often more limited in design or reporting.

These patterns suggest that empirical research on technology-enhanced art therapy has not yet been fully consolidated within mainstream art therapy scholarship and remains distributed across art therapy, technology, health, and interdisciplinary publication contexts. This distribution may reflect the substantial contribution of technological and interdisciplinary domains to the development of this emerging research area.

Importantly, across the 47 included studies, technology was often examined as an adjunct or exploratory component rather than as a fully integrated therapeutic medium. In particular, in a teletherapy context, technology use was often driven by necessity (eg, COVID-19) rather than a clearly articulated therapeutic rationale [1,4]. Within the empirical literature reviewed here, this pattern suggests that technology has not yet been consistently conceptualized as an inherent component of art therapy practice.

Nevertheless, the qualitative studies tended to converge on a proposed conceptual model of how technology may shape the therapeutic process. The findings pointed to a layered process in which accessibility serves as a primary point of entry into the therapeutic process and into clients’ internal worlds, thereby enabling participation by reducing barriers and creating the preconditions for engagement to emerge [3]. In this context, engagement was reflected in sustained, active involvement in artmaking and was often accompanied by heightened playfulness, increased agency, and reduced self-criticism.

Beyond engagement, digital affordances appeared to augment expressive capacities, enabling more flexible, iterative, and multimodal forms of symbolic expression. Crucially, these processes were not self-contained but appeared closely related to changes in the therapeutic relationship and the therapeutic frame. Technology was seen as reconfiguring the relational and structural matrix, and by extension, the dynamics of interaction itself. From this perspective, the therapeutic relationship and frame may constitute important relational and structural contexts through which accessibility, engagement, and sensory-embodied experience may acquire therapeutic significance. These relationships are proposed conceptual pathways derived from the synthesis rather than empirically demonstrated mechanisms (for an illustration of these processes, see Figure 2).

This scoping review has several limitations. Many studies used small samples, pilot designs, or case-based methodologies, which curtail generalizability. Reporting was often inconsistent, and considerable heterogeneity existed across populations and interventions. Moreover, unclear conceptual boundaries between “art therapy,” “art-based activities,” and broader digital mental health interventions complicated the interpretation. A substantial proportion of the literature relied on proximal indicators such as engagement, usability, and participation rather than robust clinical outcomes, making it difficult to draw robust conclusions about therapeutic effectiveness. Future research should increase methodological rigor, clarify how art therapy processes are operationalized in technological contexts, and move beyond feasibility to test proposed mechanisms and pathways of change. Greater conceptual clarity is also needed to distinguish between technology as a therapeutic medium and as a standalone intervention. Integrated training is also needed to combine technological competence with conceptual and clinical understanding of digital processes. Finally, excluding conference proceedings may have underrepresented emerging work in rapidly developing technology-oriented areas such as AI, VR, and human-computer interaction, where peer-reviewed conference publications may be an important dissemination route.

Overall, the field appears to be in a transitional phase. Across the included studies, technology-integrated approaches were associated with reported benefits in engagement, accessibility, and creative expression, while evidence regarding clinical effectiveness remained limited. The synthesis suggests that digital technologies may influence access, engagement, aesthetic distance, expressive capacities, and sensory-embodied experience; however, these proposed relationships should be regarded as hypothesis-generating conceptual pathways rather than empirically established mechanisms.