Extended reality and web-based imaging in case-based anatomy education: a two-year study of knowledge retention on anomalies and variations

Abstract

Background

This prospective repeated-measures quasi-experimental educational study, conducted across two consecutive first-year medical school cohorts, aimed to develop and implement case-based learning (CBL) modules using three-dimensional (3D) models and magnetic resonance imaging (MRI) within an extended reality (XR) environment. The intervention integrated anatomical variations and anomalies to enhance immediate learning and long-term knowledge retention.

Methods

MRI data from selected donors at the John A. Burns School of Medicine (JABSOM) were used to create XR-CBL modules. First-year medical students from the classes of 2024 (Cohort 2021) (n = 61) and 2025 (Cohort 2022) (n = 67) participated. Before the session, students received an email detailing the activity. During the gross anatomy laboratory, they accessed a website to explore cases featuring learning objectives, clinical and radiological data, 3D models, and MRIs. The XR experience utilized desktop-based XR technology. An assessment was administered before, immediately after, and eight weeks post-intervention. Students also completed a mixed-format feedback survey.

Results

A repeated-measures ANOVA revealed a significant effect of the intervention on student performance across both cohorts. For Cohort 2021, Wilks’ Lambda = 0.384, F (2, 59) = 47.27, p < 0.001, partial η² = 0.616. For Cohort 2022, Wilks’ Lambda = 0.30, F (2, 65) = 72.72, p < 0.001, partial η² = 0.691. Follow-up pairwise comparisons with Bonferroni adjustments indicated significant improvements from pre-test to post-test and from pre-test to post-post-test for both cohorts (p < 0.01). There was no significant difference between the post-test and post-post-test scores (p = 1.00). In Cohort 2021, mean scores significantly increased from pre-test (3.9 ± 1.9) to post-test (6.2 ± 2.6) and were maintained at post-post-test (6.3 ± 1.7). Similarly, Cohort 2022 demonstrated significant gains from pre-test (4.3 ± 1.6) to post-test (7.6 ± 2.2), with high retention at post-post-test (6.3 ± 2.4). Pairwise comparisons confirmed that improvements from baseline were significant (p < 0.01), while no significant decay occurred between post- and post-post-assessments (p = 1.00).

Conclusions

Incorporating multimodal resources, including anatomical variations and anomalies through CBL, into the gross anatomy laboratory was associated with significant score improvements and positive student perceptions. Learners reported that the interactive, imaging-based tools helped them better understand complex anatomical relationships, particularly when anatomical variability was highlighted.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12909-026-09470-7.

Introduction

Mastering anatomy presents unique challenges, particularly due to the need to comprehend complex three-dimensional (3D) spatial relationships between anatomical structures [45]. It involves both the memorization of detailed content and the ability to visualize and reason through spatial configuration. These skills are often difficult to develop through traditional teaching methods. Educators and students struggle to convey and grasp these intricate concepts effectively. To overcome these obstacles, educators must implement diverse, multimodal strategies that promote spatial understanding and encourage deeper engagement with the material [1, 10, 37, 45]. Students gain from active learning approaches that enhance spatial reasoning and support retention and meaningful application of anatomical knowledge [53].

One such learning approach is the use of extended reality (XR). XR refers to the use of digitally rendered environments that enable interactive visualization within a defined space. This technology has created a paradigm shift in medical education by introducing innovative, immersive learning opportunities [19, 22, 30, 31, 49]. XR is used as an umbrella term encompassing technologies that enhance interaction with digital 3D environments, including augmented reality (AR), virtual reality (VR), and related visualization platforms. The XR experience implemented in this study utilized desktop-based XR technology (zSpace system) rather than head-mounted displays. Unlike fully immersive VR environments delivered through head-mounted displays, this platform provides an interactive stereoscopic 3D visualization environment on a dedicated workstation, enabling spatial exploration and manipulation of the specimens without full immersion. In anatomy education, XR allows instructors to create engaging, spatially rich learning environments that deepen student involvement and comprehension [16, 32, 37]. The applications of XR in medical training continue to expand, including its use in teaching anatomical variations, a crucial yet often underrepresented aspect of anatomical education [33, 41, 44]. XR tools support active learning by allowing students to manipulate 3D models, explore structures from multiple perspectives, and engage with rare or clinically significant variants that may not be encountered during traditional cadaveric dissection. Furthermore, XR simulations can replicate realistic clinical scenarios, providing experiential learning opportunities for identifying and managing anatomical variations [21, 34].

Understanding anatomical variations is a core part of medical education and a requirement for successful practice [3, 26, 46]. While many variations rarely require direct treatment, they can still impact patient outcomes by making a diagnosis or procedure more complex. Being able to recognize these differences is vital for healthcare professionals, as it allows them to perform medical procedures more effectively. Despite this importance, anatomical variations are often overlooked in medical training [46]. Weaving these studies into the curriculum enriches a student’s perspective, leading to better diagnostic and treatment results in the clinic [3, 43]. The terminology describing deviations from typical anatomy remains inconsistent, with ‘anatomical variation’ and ‘anomaly’ variably defined and sometimes used interchangeably across the literature [23]. In the context of this study, we do not impose a rigid distinction between these terms; rather, both are used to represent departures from what is considered typical anatomy, consistent with our educational focus on recognizing anatomical diversity.

Integrating radiology into anatomy helps prepare future physicians for diagnosis and clinical decision-making. The anatomy lab serves as a natural bridge, allowing students to connect physical 3D structures with their 2D radiologic counterparts. Brown et al., [5]. Research shows that adding post-mortem CT and MRI scans to the anatomy lab really helps students. It gives them a better sense of how body parts relate to each other in space and makes what they learn feel more relevant to real-world medicine [15, 38].

Finally, Case-Based Learning (CBL) offers numerous benefits in medical and health professional education, primarily by engaging students as active learners in collaborative small groups to solve clinical cases. Hansen and Krackov exemplify how CBL can be effectively implemented in traditional courses like gross anatomy. Through small-group case discussions, their program fostered active learning, strengthened clinical integration, and enhanced student engagement [18]. Similar outcomes were reported in an interventional study by Sangam et al., [42] where CBL was introduced into the anatomy curriculum for first-year medical students. Compared to traditional lectures, CBL significantly improved both immediate learning outcomes and knowledge retention. The authors attributed these gains to the active, self-directed nature of the CBL sessions [42]. This learner-centered approach helps students develop crucial skills such as communication, critical thinking, clinical reasoning, and problem-solving by applying theoretical knowledge to real-world scenarios, thereby linking theory to practice [10].

Given these challenges, together with recent advancements and evidence-based strategies in anatomy education, this study introduces an integrated approach. Specifically, it combines emerging technologies with sound pedagogical methods to strengthen learners’ comprehension and retention of complex anatomical concepts. This study highlights four key components: the use of XR technology, integration of web-based MRI, emphasis on anatomical anomalies and variations, and delivery through a CBL approach. The study developed educational modules that integrate these elements into an innovative, immersive learning experience. Researchers evaluated the modules’ impact on immediate knowledge and long-term retention, alongside student feedback on how effectively the tools supported their mastery of anatomy.

Materials and methods

Study participants

This study was institutionally reviewed by the John A. Burns School of Medicine (JABSOM) Institutional Review Board and approved (IRB00891). This is a prospective repeated-measures quasi-experimental educational study, conducted across two consecutive first-year medical school cohorts. Volunteer first-year medical students from JABSOM matriculating in 2021 (n = 61) and 2022 (n = 67) accounted for the two cohorts. An overall participation rate of 81% (128/158) was achieved. Student performance was individually tracked over the study period. Students were informed that participation in the study was voluntary and would not affect their course grades or evaluations. Prior to participation, all students were provided with information about the purpose of the study and informed consent to participate was obtained from all the participants. Due to the design of this course-based intervention, all students participated in the module as part of their standard curriculum, making it infeasible to include a traditional control group. It was not possible to discriminate students into separate experimental and control arms without withholding a key learning experience. This limits the ability to draw causal inferences regarding the effectiveness of the intervention, as observed improvements may be influenced by confounding factors such as concurrent coursework, maturation effects, or prior knowledge.

Educational context

This study was conducted during the gastrointestinal and endocrine problems module at JABSOM, a medical school that uses a problem-based learning (PBL) curriculum. The anatomy curriculum is delivered through a flipped classroom model complemented by in-person laboratory sessions. To meet the primary objective of integrating anatomical case-based learning, a structured workflow was developed featuring two clinical conditions: Chilaiditi Syndrome and Mobile Caecum.

Development and delivery of instructional materials

Anatomical specimens utilized in this study were drawn from permanent body donations to the JABSOM, University of Hawaiʻi at Manoa. All activities conformed to standard operating procedures of the Willed Body Program [27]. The imaging protocol followed the methodology described by Nakamatsu et al., [32]. In brief, MRI scans were obtained from routine formalin-embalmed body donors. These images were reviewed collaboratively by a radiologist and an anatomist before being uploaded to a web-based Digital Imaging and Communications in Medicine (DICOM) viewer. Workflow for each case included clearly defined learning objectives and was supported by clinical, radiological, and pathological reports, along with anatomical correlations (Fig. 1). These were further enhanced with 3D photogrammetric reconstructions, segmented models, and corresponding 3D donor MRI data (Figs. 2 and 3).

Fig. 1.

Workflow illustrating the sequential development of XR-integrated CBL modules and their implementation during instructional delivery

Fig. 2.

XR-compatible web-based MRI viewers used to facilitate interactive radiologic exploration of the two case modules. Representative MRIs from the Chilaiditi syndrome case (A) and mobile cecum syndrome case (B)

Fig. 3.

XR-compatible 3D anatomical models generated through photogrammetry (A1, B1) and segmentation (A2, B2) for the Chilaiditi syndrome (A1–A2) and mobile cecum syndrome (B1–B2) cases. These multimodal visualizations were designed to enhance spatial understanding of clinically relevant anatomical relationships

All participants received an email detailing the purpose and procedures of the activity three days before the scheduled gross anatomy laboratory session. During the 2-hour gross anatomy laboratory session, students participated in a faculty-facilitated CBL activity using the Anatomy XRCore web platform integrated with the zSpace desktop-based XR system. The XR hardware consisted of a stereoscopic display, infrared-tracked polarized 3D glasses, and an infrared-tracked stylus pen, enabling learners to actively manipulate three-dimensional anatomical models and donor-derived MRI datasets. Student engagement was active. Working in small laboratory groups (n = 5–6 per zSpace desktop-based XR system), learners independently navigated the XR cases through the we platform, which provided access to case objectives, clinical prompts, donor MRI scans, and 3D reconstructions of anatomical variations and anomalies. Using the stylus, students could rotate, zoom, and see the labeled structures in three-dimensional space, as well as dynamically visualize the axial, coronal, and sagittal planes by virtually “cutting” through the MRI datasets. The instructional team included two faculty facilitators and three teaching assistants, whose role was to guide discussion, clarify anatomical relationships, and support technical interaction with the XR platform when needed. However, because JABSOM students had prior longitudinal exposure to the zSpace display and MRI viewer in earlier modules, the session emphasized self-directed exploration and peer discussion, with instructors serving primarily as facilitators rather than lecturers. During this laboratory session, students were first presented with the subject’s background, including history and physical examination findings. Working in small groups, they were expected to generate initial hypotheses based on this information and continue refining them as the case progressed (Appendix 1). Responses collected via Google Forms indicated that students were able to formulate reasonable and clinically relevant hypotheses, as assessed by the instructors. The remainder of the case was revealed, and students were given time to interpret the accompanying images on the XR-compatible computers. Within their lab groups, students alternated between performing the routine dissection of the abdomen and engaging with the XR display, taking turns throughout the session.

Data collection

A quantitative 12-item multiple-choice question (MCQ) assessment incorporating text-based, image-based, and case-based items was developed to evaluate learning outcomes. Items included interpretation of MRI scans, 3D anatomical renderings, and short clinical case vignettes, allowing assessment of both anatomical knowledge and clinical reasoning. The MCQs were created by two faculty members with expertise in anatomy and medical education using the framework described by Guaraná et al., and were aligned with Bloom’s taxonomy [14, 16]. Learning objectives included identifying and naming specific abdominal regions and retroperitoneal structures (remember), explaining the embryological basis of gut return and its impact on adult peritoneal relationships and clinical pain localization (understand), identifying anatomical structures, variations, and pathological findings on MRI scans and 3D images (apply), and differentiating among causes of quadrant-specific pain to determine the most likely diagnosis (analyze/evaluate). The MCQ assessment tool was not formally evaluated using reliability metrics. To support content validity, the assessment blueprint was mapped directly to the session learning objectives and reviewed independently by the two faculty experts to ensure alignment with intended cognitive levels and anatomical content coverage. Face validity and item clarity were further established through pilot testing with a small group of students, whose feedback informed minor revisions in wording and item presentation.

The same MCQ set was administered at three time points: before the case presentation (pre-test), immediately after (post-test), and eight weeks later (post-post test), to evaluate knowledge acquisition and retention (Fig. 1). Reuse of identical items across time points enabled direct score comparison. Although a testing effect cannot be completely excluded, this was minimized by randomizing both the order of questions and the sequence of answer choices at each administration. In addition, students completed a qualitative feedback survey comprising bimodal and 5-point Likert-scale items, as well as open-ended questions. Seven perception-based questions were developed by two experts and validated by a small student group and one expert. The survey was evaluated on a 5-point Likert scale, with 1-strongly disagree, and 5-strongly agree. The survey questions provided further insights into perceived attitudes toward CBL modules.

Data analysis

A repeated-measures design was used, with participants serving as their own controls across three assessment time points: before the case presentation (pre-test), immediately after (post-test), and eight weeks later (retention test). MCQ scores (12 items) were analyzed using R version 4.2.1 (R Core Team, 2022). For each cohort, descriptive statistics including means and standard deviations (SDs) calculated for all time points. A repeated-measures ANOVA was conducted to examine the effect of the case study on student test scores across three time points (pre, post, post-post). Mauchly’s test indicated that the assumption of sphericity was violated (p < 0.05). A multivariate approach using Wilks’ Lambda was adopted to determine significance.

Results

Performance outcomes

Descriptive statistics, including means and standard deviations (SDs), were calculated. Table 1 summarizes the mean MCQ scores and standard deviations for both cohorts at the three assessment points.

Table 1.

Mean scores, SDs for Cohorts 2021 and 2022 across three assessment points: pre-, post-, and 8-week retention

Cohort	Assessment	Mean	SD
2021 (N = 61)	Pre	3.9	1.9
	Post	6.2	2.6
	Post-post (8-week retention)	6.3	1.7
2022 (N = 67)	Pre	4.3	1.6
	Post	7.6	2.2
	Post-post (8-week retention)	6.3	2.4

A box and whisker plot analysis was performed to determine the distribution of the data for each cohort of the study (Fig. 4). The analysis revealed a significant effect of the case study on student test scores in both cohorts. For Cohort 2021, Wilks’ Lambda = 0.384, F (2, 59) = 47.27, p < 0.001, partial η² = 0.616. For Cohort 2022, Wilks’ Lambda = 0.30, F (2, 65) = 72.72, p < 0.001, partial η² = 0.691. Follow-up pairwise comparisons with Bonferroni adjustments indicated significant improvements from pre-test to post-test and from pre-test to 8-week retention test for both cohorts (p < 0.01). Notably, there was no significant difference between the post-test and 8-week retention test scores (p = 1.00), suggesting that the knowledge gained was successfully retained over time.

Fig. 4.

Pre-, Post-, and Post-Post-Test scores from the CBL activity. Box plots depict student performance on pre-test, post-test, and post-post-test for two academic years: (A) 2021 and (B) 2022. Significant improvements were observed between pre-test and post-test and pre-test and post-post-test (8-week retention test) scores (p < 0.01) in both cohorts. (Cohort 2021: n = 61; Cohort 2022: n = 67)

Assessment scores were normalized to percentages to evaluate the magnitude of change (Fig. 5). Both cohorts showed a positive average score increase between the pre-assessment and the post-assessment and 8-week retention assessment. While the 2022 cohort demonstrated an 11.7% numerical decrease between post- 8-week retention assessments, this change was not statistically significant (p = 1.00).

Fig. 5.

Mean score differences between Pre-, Post-, and 8-week retention tests in 2021 and 2022. Stacked bar graphs show the percentage changes in test scores between the pre-test and post-test (light green), pre-test and post-post-test (green), and post-test and post-post-test -test (black). (Cohort 2021: n = 61; Cohort 2022: n = 67)

Student perceptions

Figure 6 summarizes student feedback on the perceived impact of the CBL modules. Responses were categorized into 3 groups: “agree” (ratings 4–5), neutral (ratings 3), and “disagree” (ratings 1–2). Students in both cohorts responded positively, noting that the cases were well-organized and clearly emphasized the significance of normal and variant anatomy. Although the structure, content, and delivery of the CBL modules remained consistent across both cohorts, Class of (CO) 2025 reported higher levels of satisfaction and perceived learning gains compared to CO 2024. This discrepancy may reflect cohort-specific factors such as group dynamics, prior preparation, learning preferences, or varying levels of engagement. These findings highlight the influence of contextual and learner-related variables on educational outcomes, even in standardized instructional settings.

Fig. 6.

Student feedback on learning modules for both cohorts. (Cohort 2021: n = 61; Cohort 2022: n = 67)

A thematic analysis [8] was conducted on the open-ended responses from both student cohorts to the question, “How could case presentation be improved?” This qualitative research method involved systematically identifying, analyzing, and reporting themes within the data (Table 2). Table 2 highlights the strengths and areas for improvement of the case presentation system. Students particularly appreciated the interactive 3D models for facilitating visualization of anatomical concepts and valued the concise, well-organized case presentations and associated materials. The inclusion of clinical context was highlighted as a significant positive, helping to reinforce anatomical relevance. The consensus points to a strong foundation for learning, with students valuing the opportunity to connect anatomy with clinical imaging. However, the feedback also outlines several key areas where improvements could significantly enhance the learning experience. Recurring themes for enhancement include providing labeled structures and centralized step-by-step guidance on navigating the platform. Students also suggested more structured guidance within cases and the integration of short, focused videos with optional walkthroughs. On the technical front, improving the DICOM viewer’s load speed, fixing bugs, ensuring device compatibility, and offering assistance for common errors were identified as crucial. Finally, the integration of embedded quiz questions and the highlighting of key learning points were proposed, suggesting a desire for more active learning and assessment within the presentations (Table 2).

Table 2.

Summary of student feedback to the question ‘How could case presentation be improved?’

Theme	Positive Highlights	Suggested Improvements
3D Model Integration	Students appreciated interactive 3D models for visualizing anatomical concepts.	Provide labeled structures, clearer controls, and guided usage instructions for the DICOM viewer and Sketchfab.
Instructional Clarity	Students valued concise case presentations and linked materials.	Add centralized, step-by-step written or video instructions on using the platform and navigating the cases.
Multimedia Content	The topography of the colon video was frequently cited as helpful and engaging.	Include more short, focused videos with labeled content and optional walkthroughs for each case.
Clinical Relevance	Students enjoyed the real-case clinical context to reinforce anatomical relevance.	Have more structured guidance while progressing through each case.
Technical Usability	3D models worked well for most students.	Improve DICOM viewer load speed, fix bugs, and ensure compatibility across devices; offer help for common errors.
Learning Integration	Students appreciated the opportunity to connect anatomy with clinical imaging.	Add embedded quiz questions, highlight key learning points.

Discussion

Findings presented here align with a growing consensus that XR tools outperform traditional methods in fostering spatial understanding [1, 11, 20, 30]. These benefits stem from XR’s capacity to integrate digital content with real-world learning contexts, thereby reshaping established pedagogical practices. Research on XR models generated from radiological datasets, paralleling our CBL, demonstrates that learners regard such models as highly valuable and engaging, promoting meaningful participation in the learning process [25, 28]. Specifically, our study reinforces the “pedagogical value of immersion” noted by Nakamatsu et al., [32]. While Nakamatsu reported that 79.7% of students improved their grasp of anatomical relationships through MRI scans, our study found a comparable 81% approval for the XR-CBL format. This convergence suggests that regardless of the specific software, the integration of radiological imaging into anatomy curricula consistently stabilizes a student’s mental map of the human body [32]. Diegritz et al., [9] demonstrated that integrating immersive virtual reality (VR) with Cone Beam CT (CBCT) in a dental curriculum led to a substantial improvement in objective knowledge outcomes. Students effectively doubled their average scores on multiple-choice questionnaires. Notably, this performance gain persisted at the two-week follow-up, suggesting that the XR-based approach may support longer-term knowledge retention. In addition, participants consistently reported high levels of presence when using the VR/CBCT system, rating control, realism, and sensory immersion above 8 on a 10-point scale [9].

The technological barriers encountered with the initial DICOM viewer underscore a crucial pedagogical implication. The success of any advanced educational technology is contingent upon its user-interface design and accessibility. As first-year medical students, many faced a steep learning curve, suggesting that future users may benefit from early, informal exposure to the platforms. Despite these challenges, the majority of students indicated that the case studies had a positive impact on their learning, emphasizing the educational value of the content even in the face of technological limitations. In response to the feedback gathered from these experiences, we moved away from the earlier DICOM viewer and adopted a more user-friendly, web-based platform (shown in Fig. 2) to improve accessibility and enrich the overall learning experience. Future implementations must prioritize minimal cognitive load from technology itself to maximize the focus on the anatomical content.

Recent evidence underscores the value of three-dimensional visualization technologies (3DVT) in improving anatomical understanding. A recent meta-analysis of 36 studies, including 28 randomized trials and over 2,200 participants, demonstrated that 3DVT significantly enhances both factual and spatial anatomy knowledge, while also improving user satisfaction and learners’ perception of effectiveness [50]. In our proof-of-concept study conducted at a single institution using selected abdominal cases, we explored the use of various 3DVTs (such as 3D MRIs and 3D Photogrammetric and segmented models) to teach not only standard anatomical structures but also anatomical variations and anomalies, an area often underrepresented in traditional teaching formats. Within this context, XR-based approaches may offer a promising tool not only for knowledge acquisition but also for facilitating conceptual understanding of anatomical variability. By incorporating 3DVT to illustrate variation, this approach may help expand students’ understanding of the spectrum of human anatomy and better prepare them for the diagnostic uncertainty encountered in clinical practice. Further studies across institutions, anatomical regions, and learner populations are needed to determine the broader applicability of these findings. Studies examining medical imaging-enhanced anatomy curricula generally report positive outcomes [5, 24, 32, 36, 48]. Kawashima et al., reported that an integrated anatomical practice combining cadaver dissection with CT/MRI image analysis significantly enhanced student engagement, with over 74% reporting it as a useful introduction to anatomical imaging and nearly all submitting voluntary imaging reports [24]. Regardless of image source, repeated exposure to cross-sectional imaging tended to bolster spatial reasoning, and immersive modalities such as augmented-reality viewing of cadaver CT offered further gains [39, 47]. These findings suggest that providing students with opportunities to study anatomical imaging through approaches such as CBL, as implemented in the present study, the integration of imaging into anatomy training appears to support student learning and may represent a valuable component of anatomy curricula.

In line with previous research, there is growing advocacy for a stronger emphasis on incorporating anomalies and variations into the curricula of medical and allied health education programs [3, 35]. Early VR approaches were primarily developed to create limited anatomical databases [34] and to facilitate the transmission of anatomical knowledge to locations geographically distant from traditional classroom settings [6]. Within problem-based learning contexts, such environments can be used to design modules that illustrate anatomical variations and the diagnostic and treatment challenges they may present. In this study, we sought in part to examine whether presenting anatomical variations through XR visualization supports anatomical knowledge retention.

The abdomen is a common site for surgical procedures and exhibits a high prevalence of anatomical variations. These variations can have a significant impact on surgical approaches and clinical outcomes. The mobile cecum, defined as the failure of the cecum, terminal ileum, and/or right colon to properly fuse with the posterior parietal peritoneum during embryological development [40], is a condition that occurs in approximately 10–20% of the population. It is typically identified as a congenital anomaly in children but may also be identified incidentally in adults. Clinically, a mobile cecum can present with recurrent right lower quadrant pain [29] and, in severe cases, may lead to volvulus and subsequent strangulation ileus [13]. Chilaiditi’s syndrome involves the interposition of a segment of the bowel between the liver and diaphragm. This anatomical arrangement may produce a wide spectrum of symptoms, including abdominal pain, bloating, nausea, vomiting, altered bowel habits, and, in some cases, more atypical manifestations such as substernal pain, arrhythmia, dyspnea, and respiratory distress [52]. These two cases were presented and discussed during the anatomy course to illustrate the clinical relevance of abdominal anatomical variability. Students engaged in critical discussion, recognizing the importance of understanding these less common presentations for accurate diagnosis. By comparing these variations to typical anatomy, the session aimed to enhance students’ ability to identify and appreciate variability while reinforcing core anatomical knowledge.

Although the present study focused on the mobile cecum and Chilaiditi’s syndrome, the combined pedagogical framework integrating imaging, XR visualization, and CBL may be adaptable to other organ systems and anatomical regions. The increase in student engagement observed here aligns with reports from other complex anatomical domains. Recent studies, for example, document strong learner engagement in neuroanatomy [12] and in mastering the intricate spatial relationships of the eye and ear [2]. Similarly, Guaraná et al., demonstrated positive outcomes of a comparable CBL model in teaching veterinary imaging anatomy [16]. As summarized in Lang et al.,’s review, XR applications in radiology education span a broad spectrum of anatomical domains, demonstrating the versatility and generalizability of these technologies [28]. Studies employing XR platforms commonly target foundational regions such as the head and neck, cardiac vasculature, breast anatomy, and whole-body structures for introductory anatomy teaching. Beyond these core areas, XR has been used to enhance understanding of more complex spatial relationships, including cranial bone anatomy, abdominal and vascular structures, and detailed neuroanatomical and urologic regions such as the kidney and prostate. Patient-specific 3D models derived from CT and CBCT have further supported instruction in congenital cardiac lesions and intracranial tumor locations. Within procedural training, XR has also been widely applied to interventional radiology, supporting skills development for renal artery catheterization, percutaneous nephrostomy, biopsy guidance, and endovascular techniques [28]. Furthermore, the capacity of XR to visually archive and convey variations [34, 54] overcomes the limitations of traditional labs, where such specimens are rarely preserved. Collectively, these studies validate that the core principle of the XR-CBL model, using immersive 3D content to master complex, patient-specific anatomical variation, is not only applicable but is actively being used in nearly all major anatomical systems, from basic structural anatomy to highly specialized clinical domains [28].

The integration of XR technologies in anatomy education offers transformative benefits for spatial understanding and learner engagement; however, it also introduces important ethical and data-privacy considerations that must be addressed to ensure responsible implementation [4]. Beyond the traditional ethical principles associated with body donation, which continue to apply when digital resources are derived from donor images, the use of “digital cadavers” raises additional questions regarding informed consent, ownership, and respectful representation of donor bodies. Previous studies emphasize that explicit consent and transparent governance are essential when cadaveric images or anatomical datasets are digitized and distributed in educational platforms. These challenges highlight the need for robust ethical frameworks and institutional policies that address both the respectful use of anatomical materials and the secure handling of user data in immersive learning environments [7]. In our institution, informed consent documentation for body donation includes permission for medical imaging and three-dimensional reconstruction for educational purposes. In addition, access to digital anatomical materials is restricted to registered students, and institutional guidelines include clear usage disclaimers prohibiting unauthorized downloading, copying, or distribution of educational images.

There are limitations to the present study. Only two cases, both limited to the abdominal region, were included in this study, and the conditions presented are relatively rare. Another limitation was the small sample size, consisting of two cohorts from a single medical school. The absence of a negative control group presents another limitation since, all students received the same teaching. However, since these cases were part of their required curriculum, including a negative control was not feasible. From a technical standpoint, students encountered several usability difficulties with the initial DICOM viewer, such as slow loading, limited device compatibility, and complex navigation. In response, the system was later replaced with a more intuitive, web-based platform designed to enhance ease of access and overall usability. Such technical barriers highlight potential challenges to replicating and implementing similar approaches in other educational settings.

The distinction between the terms anatomical variation and anomaly remain ambiguous [23]. The terminology used to describe certain anatomical findings often varies across the literature. For instance, Chilaiditi syndrome and mobile cecum are alternately labeled as anomalies, variations, or both, depending on the source [13, 17, 51]. This inconsistency underscores a broader challenge in anatomical nomenclature. However, our study did not attempt to establish a formal definition or draw a sharp distinction between these terms. Rather, our purpose was to teach anatomical variability in comparison to what is considered ‘typical’ and to evaluate knowledge retention in this context. To better understand the broader educational impact of case-based presentations emphasizing variant anatomy, future studies should incorporate a wider range of anatomical regions, more varied case types, and larger, more diverse student populations.

Conclusion

This study validates a structured workflow for integrating donor-specific pathologies and XR technologies into a CBL framework. Unlike standard XR implementations that rely on idealized digital models, our approach utilizes authentic anatomical variations derived directly from our donors. By transitioning from generic anatomical models to donor-derived imaging, we provided students with authentic clinical scenarios that mirrored the complexities of real-world practice. While student perceptions highlight high engagement, the quantitative correlation with improved learning outcomes suggests that this iterative, multimodal workflow may offer a scalable model for modernizing gross anatomy laboratory education.

Abbreviations

ANOVA

Analysis of Variance

AR

Augmented reality

CBCT

Cone beam computed tomography

CBL

Case-based learning

CO

Class of

CT

Computed tomography

DICOM

Digital Imaging and Communications in Medicine

IQR

Interquartile range

JABSOM

John A. Burns School of Medicine

MCQ

Multiple choice question

MRI

Magnetic resonance imaging

PBL

Problem-based learning

3D

three-dimensional

3DVT

three-dimensional visualization technologies

XR

extended reality

VR

Virtual reality

Authors’ contributions

Conceptualization: Güneş Aytaç, Scott Lozanoff, Rebecca Romine; Methodology: U-Young Lee, Christoph Rettenmeier, Cemil Oktay, Scott Lozanoff, Güneş Aytaç; Formal analysis and investigation: Kalāwena Kalehuawehe, Shawnea Pagat, Güneş Aytaç, Rebecca Romine; Writing - original draft preparation: Kalāwena Kalehuawehe, Shawnea Pagat, Güneş Aytaç; Writing - review and editing: Güneş Aytaç, Scott Lozanoff, Rebecca Romine; Supervision: Scott Lozanoff. All authors read and approved the final manuscript.

Funding

This project was supported by grants from the National Institutes of Health (NIH), National Institute of General Medical Sciences (NIGMS), IDeA Networks of Biomedical Research Excellence (INBRE), Award number: P20GM103466, and National Science Foundation (NSF) Tribal Colleges and Universities Program (TCUP) Targeted Stem Infusion Project (TSIP), Award number: 1855379. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Data availability

The data that support the findings of this study are available on request from the corresponding author.

Declarations

Ethics approval and consent to participate

This study was institutionally reviewed by the John A. Burns School of Medicine Institutional Review Board and approved (IRB00891). All participants were given an electronic informational document and the researcher explained the study to them. Informed consent to participate was obtained from all of the participants. The study maintained participant anonymity and data confidentiality. It was conducted in full compliance with international ethical standards, including the Declaration of Helsinki.

Consent for publication

Obtained from all the participants.

Competing interests

The authors declare no competing interests.