Free scoping read. A written assessment of your manuscript in two business days, at no cost.

Start one
Contents

Academy · Research Data, Analytics & AI · 3D Web-Based Device Training

Design and Implementation of Instructional 3D Web-Based Interactive Medical Devices: A Case Study and the Randomised Evidence

A London South Bank University project developed web-based interactive 3D virtual user manuals for two research instruments. Separate randomised studies of 3D anatomy models indicate the circumstances in which this approach has been shown to improve learning outcomes.

Pubrica AcademyWritten by a medical education technology research writer Reviewed bya biomedical engineering educator Published10 October 2026 Last checked10 October 2026 Reading time7 minutes Sources10 peer-reviewed

The construction of an interactive 3D model and its delivery on the web are separate design problems, and the published literature on instructional medical-device training treats them as such. One directly relevant case study, a web-based interactive 3D virtual user manual for two skin-measurement research instruments, documents how such a system was developed and the reasons for the design choices. Separate randomised studies of 3D anatomical models, although not studies of device training, constitute the best available evidence on whether interactive 3D visualisation improves measurable learning outcomes in medical education, and they are therefore relevant alongside the device case study.

Learning objectives

  • The design and development of a web-based interactive 3D virtual user manual (VUM) for two physical research instruments (AquaFlux and Epsilon)
  • The technology choices adopted (3ds Max modelling and Flash-based interactivity) and the reasons the authors excluded other tools
  • The randomised evidence on the effect of interactive 3D models on learning outcomes in medical education more broadly, including reported effect sizes

Common misconceptions

  • “Any 3D product visualisation constitutes an interactive virtual learning environment.” The Al Hashimi and Xiao case study on AquaFlux and Epsilon developed button-linked, narrated image sequences that simulate device operation and fault-finding rather than a passive 3D render; in this literature, interactivity means that the user can act on the model and receive a device-specific response, rather than merely rotate a static mesh.
  • “Interactive 3D training content requires a real-time 3D engine (WebGL).” The documented AquaFlux and Epsilon VUM used 3ds Max for modelling and rendering and Adobe Flash CS6 with ActionScript to sequence and link pre-rendered image frames at 15 fps, a deliberate, lower-overhead choice the authors made in preference to the real-time engines they had also evaluated (Google SketchUp, Blender, Unity).

Case study: web-based 3D training for AquaFlux and Epsilon

Al Hashimi O and Xiao P documented “Developing a web based interactive 3D virtual environment for novel skin measurement instruments,” based on work carried out at London South Bank University and published as part of Al Hashimi’s 2020 doctoral thesis, “Web-Based 3D Interactive Virtual Systems for Medical Instruments (AquaFlux and Epsilon)”1. AquaFlux measures water vapour flux density, including transepidermal water loss, and Epsilon images the dielectric permittivity of soft materials; both are specialist research instruments that normally require on-site training to operate correctly.

The stated motivation was explicit: on-site training for this kind of instrument is expensive and difficult to schedule, and the authors argued that there was “a genuine need for an interactive virtual training environment” so that users could train “whenever and wherever they want”1. The resulting web-based virtual user manual (VUM) walked users through device operation and troubleshooting using an interactive, button-driven interface layered over the 3D models.

Rationale for the design adopted

The development pipeline is documented in sufficient detail to be instructive in its own right. The authors modeled and rendered the instruments in Autodesk 3ds Max, using V-Ray materials and HDRI lighting for photorealism, after first evaluating and rejecting Google SketchUp, Blender, and Unity1. Interactivity was added with Adobe Flash CS6 and ActionScript: rendered image sequences ran at 15 frames per second at 1920x1080 resolution, with buttons linked to specific frames and audio narration layered over the sequence1.

This architecture differs substantially from a real-time 3D engine that renders geometry in the browser (via WebGL, for example). The authors chose pre-rendered, high-fidelity image sequences linked by Flash because this approach achieved photorealistic rendering quality without requiring the end user’s browser or hardware to perform real-time 3D computation. This was a reasonable trade-off for a training tool aimed at laboratory staff rather than game-engine developers, although it ties the deliverable to Flash-compatible delivery, a significant constraint for anyone replicating the approach today given the discontinuation of Flash.

Randomised evidence on interactive 3D learning outcomes

Because device-specific randomised trials are scarce, the strongest outcome evidence for interactive 3D instructional content comes from medical anatomy education, where the method has been tested more rigorously. Nicholson DT, Chalk C, Funnell WRJ and Daniel SJ conducted a randomised controlled study of a computer-generated three-dimensional ear model built from an MRI scan, used within a web-based anatomy tutorial, published in Medical Education in 20062. Of 57 medical students, 28 used the tutorial with the interactive 3D model and 29 used the tutorial alone; on a 15-question quiz of 3D ear relationships, the 3D-model group averaged 83% versus 65% for the control group, a statistically significant difference (P < .001)2.

A separate randomised study, Battulga B, Konishi T, Tamura Y and Moriguchi H, “The Effectiveness of an Interactive 3-Dimensional Computer Graphics Model for Medical Education,” published in Interactive Journal of Medical Research in 2012, tested an interactive 3D shoulder model (rotatable, with hideable and semitransparent structures) against a textbook-only control in 100 first-year students3. The 3D-model group reported significantly higher perceived content quality (4.26 vs. 3.85, P = .001) and teaching-method quality (4.33 vs. 3.74, P < .001), with the largest gap on self-reported motivation (4.52 vs. 3.72, P < .001)3. The authors noted that this study measured perceived effectiveness by questionnaire rather than objective anatomical test performance.

Implications for the design of instructional 3D device content

Taken together, the case study and the two randomised trials suggest a practical design brief: interactivity should be built around specific user actions (operating, troubleshooting, rotating, and hiding or revealing structures) rather than a passive walkthrough, and the result should be measured against a genuine control condition wherever feasible. The AquaFlux/Epsilon project demonstrates a working, documented development process for a genuine medical or research device; the Nicholson and Battulga trials demonstrate that, at least for anatomical structures, interactive 3D visualization can improve an objective quiz score (Nicholson) and can shift self-reported engagement and perceived teaching quality (Battulga) relative to text-only or non-interactive instruction.

Limitations of both strands of evidence

None of the three sources this article relies on is a perfect fit for "does 3D web-based device training work," and each limitation is worth stating rather than glossing over. The Al Hashimi and Xiao case study documents a working build, not a controlled evaluation of whether the finished VUM actually improved training outcomes for AquaFlux or Epsilon users relative to the prior on-site method — it is evidence of feasibility and design process, not of measured learning gain.1 Nicholson et al.'s trial, while randomized and outcome-measured, used a 3D ear model built for anatomy instruction in a classroom setting with 57 students, a context with different stakes and a different user population than a research-instrument operator training on their own time.2 Battulga et al.'s shoulder-model study, meanwhile, measured self-reported perceived quality and motivation rather than an objective skill or knowledge outcome, so its strongest claim is about engagement, not demonstrated learning gain.3 Citing all three together is appropriate for building a design brief, as this article does, but citing any one of them alone as proof that "interactive 3D training works for medical devices" would overstate what that individual source actually measured.

Development sequence derived from the AquaFlux and Epsilon project

Read as a build process rather than just a technology list, Al Hashimi and Xiao's documented sequence has four stages worth separating out for anyone replicating it. First, the physical device itself was modeled in 3ds Max with attention to dimensional accuracy, since a training tool that misrepresents a control's position or a panel's layout actively works against the training goal.1 Second, photorealistic rendering was layered on using V-Ray materials and HDRI lighting, a step the authors treated as necessary specifically because trainees needed to recognize the real device's actual surface details and indicator states, not a stylized abstraction of them.1 Third, the rendered frames were sequenced and linked through Flash's button and timeline model to represent discrete device states — operating normally, mid-fault, in a specific troubleshooting step — rather than a single continuous 3D scene a user freely rotates.1 Fourth, narration was layered over the linked sequence to carry the procedural explanation that the visuals alone could not. For a team building an equivalent tool today with a WebGL-based real-time viewer instead of Flash, the same four-stage logic still applies even though the specific tooling has changed: accurate modeling, a deliberate fidelity decision, state-based (not free-rotation-only) interactivity, and narration tied to specific states rather than a generic walkthrough.

Weighting of this evidence in practice

Because this article already named the limits of its device-specific and anatomy-education sources, it is worth setting those two randomised trials against the larger body of evidence on 3D visualization in medical education. Wang, Li, Dun, Zhong and Ye's 2024 meta-analysis in BMC Medical Education, pooling 39 randomized controlled trials and 2,959 participants, found that 3D visualization technology improved test scores overall (standardized mean difference 0.69) and student-reported interest (0.70), but the pooled effect was driven almost entirely by China-based studies (1.72) while studies from other regions showed no significant effect (0.29) — a reminder that the Nicholson and Battulga results above, each a single-institution trial, sit inside a literature whose overall effect size is highly heterogeneous (I² often above 75%) and partly region-dependent rather than a single, settled finding.4 Harris, McGregor and colleagues extend the broader methodological caution to medical informatics settings specifically, cataloguing which validity threats are hardest to rule out when the intervention being evaluated is a piece of educational or clinical software rather than a drug or procedure — exactly the caution this article's limitations section above already applied to the Nicholson and Battulga trials.5

On how interactive 3D training is evaluated once it moves from a single study to a body of evidence, Kleinert and colleagues' 2015 thematic review in the Journal of Medical Internet Research surveyed studies of 3D immersive patient simulators and found learning-success outcomes reported inconsistently across studies, with few directly comparable effect sizes — a pattern that matches the heterogeneity Wang et al. later quantified across the wider anatomy-education literature.6 McInerney, Nally, Khan, Heneghan and Cahill's 2022 systematic review in GMS Journal for Medical Education, covering simulation training for medical students more broadly, likewise reports that most included studies measured self-reported confidence or satisfaction rather than objective clinical performance, the same gap Battulga et al.'s shoulder-model trial above is explicit about.7

The practical design brief this article sets out earlier also needs to be checked against the standards and tooling that actually govern it today. IEC 62366-1:2015, the international standard for applying usability engineering to medical devices, is the relevant reference for any interface — including a 3D web-based training tool treated as part of a device's labeling or accompanying documentation — that needs to demonstrate it was designed and tested for safe, effective use rather than just functional completeness.8 On the delivery technology itself, the Khronos Group's WebGL 1.0 specification, finalized in 2011, is what replaced the Flash-based, pre-rendered image-sequence approach Al Hashimi and Xiao documented: it exposes hardware-accelerated 3D rendering directly inside the browser without a plugin, which is why a team building an equivalent training tool today would default to a WebGL-based real-time viewer rather than Flash.9 And on the instructional side rather than the rendering side, Locatis's chapter on instructional design and technology in healthcare remains a reference point for why health professions training adopts new visualization technology unevenly — the forces he identifies, including clinical workflow constraints, credentialing requirements, and uneven institutional investment in educational technology, are the same forces that explain why a documented, working build like the AquaFlux/Epsilon VUM can exist without a published randomised evaluation of its own.10

Key points

Key points

A documented web-based interactive 3D virtual user manual for research instruments (AquaFlux and Epsilon) illustrates a feasible development approach, and separate randomised studies of 3D anatomy models (83% vs. 65% quiz scores, P < .001) provide the strongest available outcome evidence, which indicates that interactivity, rather than 3D rendering alone, is associated with the learning benefit.

Practical application

Practical application

  1. Specify the interactivity as well as the 3D asset. The actions available to the user, such as operating a control, triggering a fault state or hiding and revealing a structure, should be defined before a rendering or delivery technology is selected.
  2. Select the rendering pipeline according to the principal constraint. Pre-rendered image-sequence interactivity (as in the AquaFlux and Epsilon VUM) sacrifices real-time flexibility for consistent visual fidelity on any device, whereas a real-time 3D engine sacrifices some fidelity for live manipulation; the choice should depend on the training objective.
  3. Evaluate against an appropriate control condition. Where possible, the interactive 3D version should be tested against a text or static-image control using an objective quiz, following the design of Nicholson et al., rather than relying on satisfaction scores alone.

Frequently asked questions

The most directly relevant documented project is the AquaFlux/Epsilon web-based virtual user manual, which is a design-and-build case study rather than a controlled trial. The strongest controlled-trial evidence available is in adjacent 3D anatomy education (Nicholson et al. 2006; Battulga et al. 2012), which should be cited as supporting evidence for the learning-outcome mechanism, not as device-training-specific proof.

3ds Max for 3D modeling and rendering (with V-Ray and HDRI lighting) and Adobe Flash CS6 with ActionScript for interactivity, delivered as linked, narrated image sequences at 15 fps — a pre-rendered approach chosen over real-time engines the authors evaluated and rejected (Google SketchUp, Blender, Unity).

References

  1. Al Hashimi O, Xiao P. Developing a web based interactive 3D virtual environment for novel skin measurement instruments. Based on: Al Hashimi O. Web-Based 3D Interactive Virtual Systems for Medical Instruments (AquaFlux and Epsilon). PhD Thesis, London South Bank University, 2020. read 10 October 2026. View source
  2. Nicholson DT, Chalk C, Funnell WRJ, Daniel SJ. Randomized controlled study of a computer-generated three-dimensional model for teaching ear anatomy. Medical Education. 2006;40(11):1081-1087. read 10 October 2026. View source
  3. Battulga B, Konishi T, Tamura Y, Moriguchi H. The Effectiveness of an Interactive 3-Dimensional Computer Graphics Model for Medical Education. Interactive Journal of Medical Research. 2012;1(2):e2. read 10 October 2026. View on PubMed Central
  4. Wang J, Li W, Dun A, Zhong N, Ye Z. 3D visualization technology for learning human anatomy among medical students and residents: a meta- and regression analysis. BMC Medical Education. 2024;24:461. read 10 October 2026. View on PubMed Central
  5. Harris AD, McGregor JC, Perencevich EN, Furuno JP, Zhu J, Peterson DE, Finkelstein J. The use and interpretation of quasi-experimental studies in medical informatics. J Am Med Inform Assoc. 2006;13(1):16-23. read 10 October 2026. View on PubMed Central
  6. Kleinert R, Wahba R, Chang D-H, Plum P, Hölscher AH, Stippel DL. 3D immersive patient simulators and their impact on learning success: a thematic review. Journal of Medical Internet Research. 2015;17(4):e91. read 10 October 2026. View on PubMed Central
  7. McInerney N, Nally D, Khan MF, Heneghan H, Cahill RA. Performance effects of simulation training for medical students - a systematic review. GMS Journal for Medical Education. 2022;39(5):Doc51. read 10 October 2026. View on PubMed Central
  8. International Electrotechnical Commission. IEC 62366-1:2015 - Medical devices - Part 1: Application of usability engineering to medical devices. read 10 October 2026. View on ISO.org
  9. Khronos Group. WebGL 1.0 Specification (final release). read 10 October 2026. View on Khronos.org
  10. Locatis CN. Instructional design and technology in healthcare. In: Reiser RA, Dempsey JV, eds. Trends and Issues in Instructional Design and Technology. Upper Saddle River, NJ: Merrill Prentice Hall; 2002. Chapter 17. read 10 October 2026. View source

Every source above is a peer-reviewed article, a publisher’s record of one, or a standard issued by the body that owns it. Each carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.

How to cite this article

Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.

Pubrica Academy. (2026, October 10). Design and Implementation of Instructional 3D Web-Based Interactive Medical Devices: A Case Study and the Randomised Evidence. Pubrica. https://pubrica.com/academy/medical-research/design-and-implementation-of-instructional-3d-web-based-interactive-medical-devices/

This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.

Design checklist for instructional 3D web-based device training

A checklist to use before building or commissioning an interactive 3D training module for a medical or research device.

[ ] Defined the specific user actions the model must support (not just 'view in 3D')
[ ] Chosen pre-rendered vs. real-time 3D delivery against actual device-fidelity needs
[ ] Checked browser/device compatibility for the chosen delivery technology
[ ] Planned an objective outcome measure (quiz, task completion) alongside any satisfaction survey
[ ] Documented the build pipeline for future replication or handover

Pubrica AcademyWritten and maintained by Pubrica’s subject matter experts. Articles are revised when the underlying standards change, and each article states the date on which it was last reviewed.

How Pubrica can help

Pubrica’s medical education and instructional design researchers scope, reference and support interactive 3D training content with the published outcome literature rather than with general e-learning claims.

Find something else

Search the Pubrica Academy

Access is free and requires no account. Enter a topic, or open the full library to filter by category, sub-topic or resource type.

Evidence-based design of 3D instructional modules for medical devices

A scoping assessment is free and takes two business days. It provides a written assessment of what the work requires and which specialist would undertake it.

Get a free quote

Regulations

Check your project against this rule

  1. 1Your work
  2. 2Your details
  3. 3Send
What do you need?

Material you send is held under signed confidentiality, with access limited to the assigned team and encrypted transfer. It is not used to train anything and is not passed to third-party tools. Ask for the agreement before you send anything you would rather cover first.