Before any medical student ever presses a biopsy punch into a living patient, someone has to decide whether their hands are ready. At the University of Augsburg, that judgment is now being made against an unlikely stand-in: a silicone replica of a human face that costs exactly 62 euro cents to produce. Writing in the journal 3D Printing in Medicine, a team of dermatologists, educational researchers and statisticians describes how they 3D-scanned a real face, digitally sculpted cancer-like lesions into it, printed a mold on a desktop machine and cast a skin simulator that second-year students used to rehearse punch biopsies and wound closure. Of 82 students who took the course, 58 completed the anonymous evaluation — and nearly all of them said the model felt like the real thing and left them more confident about performing the procedure for the first time on a patient.

The stakes are higher than the modest price tag suggests. Suturing is among the most fundamental skills in medicine: done poorly, it invites infection, excessive scarring and delayed healing; done well, it promotes faster recovery and better cosmetic outcomes. Yet a 2016 national survey of medical students in England found that 86 percent — 526 of 705 students across 16 medical schools — considered their practical suturing training inadequate. Forty-four percent reported failing to meet competency standards for a simple interrupted suture, 84 percent fell short on subcutaneous sutures, and one in five had paid out of pocket for additional courses. The materials available for practice compound the problem. Low-fidelity options such as silicone mats, foam blocks and fruit skins — bananas are a particular favorite — are cheap, portable and reusable, but lack the tactile realism of human skin. High-fidelity alternatives like pig feet, chicken skin and human cadavers feel right, but they are costly, require ethical clearance, pose infection risks and demand specialized storage.

The Augsburg team approached the problem from inside a curriculum. The university’s Faculty of Medicine introduced a competency-oriented model degree program in the 2019/2020 winter semester, built around a spiral curriculum that keeps linking theory back to practice and is aligned with Germany’s National Competency-Based Learning Objective Catalog, NKLM 2.0. Within the mandatory second-year skills course “Examination of the Skin,” the Department of Dermatology identified one procedure as a critical early milestone: the diagnostic punch biopsy closed with single interrupted sutures. Rather than bolting a simulator on as a technical afterthought, the researchers embedded it in a structured teaching sequence that mirrors real clinical logic — lesion assessment, disinfection, sterile draping, local anesthesia, verification of adequate numbing, punch placement, control of excision depth, hemostasis, wound closure and, finally, dressing. The model was conceived as a learning tool with defined competencies attached, not a gadget.

Everything began with light. Because 3D facial data count as sensitive biometric information, the team obtained separate ethics approval from Ludwig Maximilian University of Munich and scanned the face of a dermatological resident with consent. The instrument was an Artec Space Spider, a handheld structured-light scanner that projects light patterns onto a surface and triangulates their deformation into a digital mesh — a technology chosen partly because, unlike some alternatives, it emits no radiation harmful to the eyes, making it suitable for use near patients. The scan was imported into Blender, the free open-source 3D suite, where the team isolated a deliberately instructive patch of anatomy: the cheek, nasal ala and nasolabial fold. The choice was clinically motivated. These subunits are among the most common sites of basal cell carcinoma, the most frequent form of nonmelanoma skin cancer, and therefore territory students will most often need to biopsy; they also sit within facial aesthetic units, where cosmetic outcomes carry special weight.

Inside Blender’s sculpt mode, the researchers then shaped pathology into digital skin: two tumor-like lesions mimicking the morphology of basal cell carcinoma and a gaping wound above the upper lip. Flat air channels were engineered into the design to simulate the sliding of multiple skin layers — the subtle mobility that makes living skin feel alive under a needle. Because the final model would be cast in colorless silicone, color information was discarded. To fabricate the part, the team converted the positive model into a negative mold, digitally adding a one-millimeter outer shell with Blender’s solidify modifier, or alternatively subtracting the model from a cube using a Boolean difference operation. The mold was exported as an STL file, sliced in Cura with support structures enabled and printed horizontally on an FLSun Super Racer, a delta-style fused deposition modeling printer whose lightweight print head, driven by three jointed arms and a 48-millimeter stepper motor, can travel at up to 200 millimeters per second. Polylactic acid filament flowed at 200 degrees Celsius onto a 60-degree heated bed through a 0.4-millimeter nozzle at 0.2-millimeter layer height. One mold took 36 minutes and 9 grams of filament — roughly 18 euro cents of plastic.

Casting came next. The printed mold was filled with a two-component, room-temperature-curing silicone mixed in equal parts, degassed to drive out air bubbles and left to cure for several hours. After demolding and minor postprocessing — trimming excess with a scalpel, picking stray filament threads out with tweezers — the finished models showed a homogeneous surface free of layer lines and voids. The material choice was a small act of materials science in itself. Human skin measures roughly 21 on the Shore-A hardness scale, but because the simulator was cast as a solid block rather than a thin membrane, the bulk would feel firmer than the same silicone in sheet form. The team therefore selected a softer Shore-A 10 compound to restore a realistic tactile impression. Each block survives about 10 to 15 biopsy-and-suturing procedures before tearing makes it unusable. The silicone costs about 44 euro cents per model; adding the printed mold brings the material bill to 62 cents. The one-time design phase consumed roughly 30 hours of expert time — labor the per-unit figure deliberately excludes, and iterative problems such as bubble formation and insufficient skin mobility had to be solved along the way, including a material bridge built into the mold to allow realistic tissue movement.

The simulator debuted in the 2023/2024 winter semester, and the course numbers were striking. Of 82 participants, 58 completed an anonymous online questionnaire distributed by QR code at the end of the session. Fully 94.8 percent agreed the model closely resembled real anatomy, and the same share judged it an effective way to practice skin biopsies. Ninety-six point six percent reported feeling more confident about performing the procedure on a real patient for the first time after practicing on the model. Every single respondent wanted surgical exercises on 3D skin models to become an integral part of dermatological and surgical training. Students said the structured session helped them understand each procedural step and connect the hands-on exercise to the theory covered beforehand, and many reported feeling better prepared for the objective structured clinical examination at the semester’s end. The evaluation, approved by the ethics committee of Ludwig Maximilian University of Munich, carries honest caveats: there was no baseline measurement, no comparator group, and all outcomes rest on subjective self-assessment rather than objective tests of skill.

What the Augsburg results underline is a finding that recurs across simulation research: effective skills acquisition depends less on technological sophistication than on task-specific realism, functional anatomy and the chance to repeat a procedure until it becomes routine. At 62 cents a model, realism of this order scales to entire teaching cohorts — a sharp contrast with commercial trainers, whose validation is often missing and whose price is not. The digital workflow is also inherently adaptable: the same scan-and-cast pipeline can be retargeted to other anatomical regions or to procedures relevant to general surgery and plastic surgery without substantial redesign, and future versions could integrate a simulated blood source to raise fidelity for advanced training. There is a second-order benefit, too. Students who grow fluent with 3D-printed trainers are also rehearsing for a clinical world in which patient-specific printed models of organs are already used to plan operations; fluency with the technology is becoming a clinical skill in its own right.

The authors are candid about the limits. This was a single-center, exploratory pilot: the model has not undergone formal validation, the cohort size was fixed by the curriculum rather than chosen for statistical power, and no head-to-head comparison against fruit, foam, animal tissue or commercial pads has yet been performed — a randomized comparison is now underway at the institution. Students also received personal suture kits for home practice, a didactic bonus that muddies attribution, although the survey was administered immediately at the course’s end. The model lacks simulated bleeding and varied tissue textures, and the silicone itself is not easily recyclable, an environmental concern the researchers note applies equally to foam and fruit. Their sights are already on alternatives: polylactic acid is biodegradable under industrial composting conditions but too stiff to mimic soft tissue, while emerging biopolymer composites based on gelatin, alginate or chitosan could one day reproduce both the elasticity and the layered structure of skin with a greener footprint. Adoption also demands technical expertise in modeling, printing and mold preparation, and access to hardware — gaps the team hopes to close by openly sharing data and instructions; a step-by-step production protocol is available from the corresponding author on reasonable request.

None of that dims the central image: a face pulled from a silicone mold for the price of a piece of fruit, carrying two cancers and one wound, teaching hundreds of hands before a single patient is touched. The Augsburg team argues its simulator directly remedies the shortcomings of the banana peel and the foam pad — the haptic flatness and anatomical vagueness that generations of students have quietly endured while paying for extra practice out of their own pockets. If the ongoing comparative trials confirm what the first cohort reported, the 62-cent skin may travel well beyond one German faculty, and the first biopsy a young doctor ever performs may increasingly be one that, in every way that matters to the hand, has already happened before.