Free tools for medtech engineers

Three calculators I built because I needed them, and figured you might too.

No signup, no tracking, no upsell. Just the math, the logic, and a clear answer. If one of these saves you an hour, that's the whole point. If you're hiring and you find these useful, I'd love to talk.

Tool 01 / Attribute Sample-Size Calculator

How many units do you need to sample?

Attribute (pass/fail) sampling for process validation, batch release, incoming inspection. Uses the binomial zero-failure model most commonly seen in medical device validation protocols: how many samples, with zero failures, to demonstrate a given reliability at a given confidence.

Built because I kept grabbing the same lookup table and wanted something that showed the math next to it.

Inputs
How sure do you need to be? 95% is typical for OQ/PQ protocols.
Required proportion of conforming units. 95/95 is the validation default.
Zero-failure plans are cheapest on sample size but harsh on batch release. Try 1 or 2 to see the tradeoff.
Result
Minimum sample size
59units
At 95% confidence / 95% reliability with zero failures allowed.
The math
n ≥ ln(1 − C) / ln(R)
n ≥ ln(0.05) / ln(0.95)
n ≥ 58.4 → 59
Zero-failure form. With allowable failures, solves the binomial inequality iteratively.
Verify before use in a regulated protocol. This matches standard lookup tables (ANSI Z1.4 / MIL-STD references, Statistical Methods in Medical Research). But your quality system, your protocol, your signature.
Tool 02 / Use-Error Risk Scorer

Score a use-error hazard against the risk matrix.

Per IEC 62366-1 and ISO 14971, use-errors get scored on severity and probability of occurrence. This tool lays out a 5×5 matrix, highlights your cell, and tells you the risk priority and the next action (mitigate, validate, accept, or escalate).

Built because I spend a lot of time in URRA (use-related risk analysis) tables and wanted something I could show a clinician in 30 seconds.

Inputs
3 · Serious
1 Negligible · 2 Minor · 3 Serious · 4 Critical · 5 Catastrophic
3 · Occasional
1 Improbable · 2 Remote · 3 Occasional · 4 Probable · 5 Frequent
Result
Risk priority
9
Medium · Mitigate
Risk is tolerable but should be reduced if reasonably practicable. Apply design controls, interlocks, or labeling before validation.
Matrix
Rows = severity (1 bottom → 5 top). Columns = probability (1 left → 5 right).
This is a teaching tool, not a quality record. Real URRA tables include hazard ID, task, causal factors, mitigation evidence, and residual risk sign-off. Use this to think, then write it up properly.
Tool 03 / Air Leak Detection Demo

Live differential-pressure monitor. Adjust the threshold, watch it alarm.

A simplified, browser-based version of the detection logic from my pneumothorax air-leak monitor capstone. Two simulated pressure sensors, a differential signal, and a user-adjustable alarm threshold. Try dragging the threshold slider down and watching the alarm catch smaller leaks.

Built to show the alarm logic concretely. The real device samples at 50 Hz, has hysteresis, and a debounce timer. This demo keeps the logic honest without the hardware.

Controls
0.8 cmH₂O/s
Drag to change the simulated pressure differential. 0 = no leak. Clinical leak range is roughly 0.2 – 2.5 cmH₂O/s depending on wound and drainage.
1.2 cmH₂O/s
Trigger level for the alarm. Lower = more sensitive (catches smaller leaks, but more false positives).
0.2 cmH₂O
Sensor RMS noise. Affects how much smoothing (moving average) the alarm needs before it can trigger reliably.
Monitor
Pleural drainage monitor · v0.2 demo
NORMAL
−8.4
Chest P₁ (cmH₂O)
−8.6
Drain P₂ (cmH₂O)
0.2
ΔP (cmH₂O/s)
60-sec rolling trace · threshold line shown in amber
Not a medical device. This is a simulation of detection logic for demonstration and education. The real monitor requires verified sensors, hardware debouncing, hysteresis, battery backup, and regulatory clearance before it goes anywhere near a patient.