Clinical sketch → FEA: closing the gap between periodontology and MedTech engineering

I do not trust a colourful plot that cannot survive a caption.

That is not hostility to engineering. It is periodontology. Every week I draw a defect on the back of a chart: walls, a likely flap, a papilla I may or may not be entitled to, a root that is either a pier or a passenger. That sketch is already a model. It has geometry, an implied material (whatever remaining attachment still exists), and an implied boundary condition (whatever that patient will actually bite on at eight in the morning).

Finite-element analysis is the same activity with better manners. Geometry, materials, boundary conditions, a field. The scandal is not that FEA exists. The scandal is how rarely the clinical sketch and the mesh are asked to confess to each other.

Digital dentistry has spent a decade getting very good at the first term. CBCT, intraoral scan, a fused virtual patient, surgical guides that print in the office. Groups working from Semmelweis — the Anatomic Modeler line, the virtual periodontal patient, 3D charting — have earned the right to talk about geometry as a solved-enough problem for planning. I mean that as a compliment, not as a eulogy. Solved-enough geometry makes the next absence obvious.

The next absence is mechanics.

A virtual regenerative plan that never meets a constitutive model is a very expensive photograph. It can tell you where the wall is. It cannot tell you whether the remaining attachment can carry the load after you close the flap. It cannot tell you whether an oblique vector will put the papilla in tension. It cannot tell you what happens to a graft when the patient is not a museum piece.

MedTech engineering, for its part, often arrives with a mesh and a catalogue of moduli and a quiet contempt for the chart. That is the other incomplete model. A PDL that is a linear spring, a mandible fused to a rigid basement, a load that is a single vertical newton — these are not conservative assumptions. They are different surgeries.

Closing the gap is not a startup. It is a translation habit:

1. Keep the sketch. It is the clinician's prior.
2. Make the virtual patient carry a ligament, two bones, and a cementum, not a painted socket.
3. Write the boundary conditions in sentences a surgeon can refuse.
4. Label every heatmap as a teaching field until it has been through a validation story that a journal, not a landing page, would accept.

I work in a clinic in Zagreb and I think about this because I have to suture things that will be loaded tomorrow. FEA that cannot change a suture is decoration. FEA that can argue with a sketch is a colleague.

PerioMechanics is the educational layer of that argument. It is not a device. It is the missing caption.