Web app · tangent method

QTc measurement app

Just paste an ECG image and the app measures the QT interval by the tangent method and calculates QTc (Bazett and Fridericia). The results come out as text you can paste straight into an electronic medical record. The interface is in Japanese.

Latest version 2.52026/9/8 update

Try it in the browser Download the zip
Example analysis screen (click to enlarge)

Features

Features

  • Measure just by pasting

    Paste (or drag in) an ECG image and the app measures QT for each beat in the image and shows the average. It is designed for images of about seven beats.

  • T-wave end by the tangent method

    The end of the T wave is where a tangent drawn on the downslope of the T wave meets the baseline. All markers are shown on screen and can be corrected by hand.

  • Works with Holter ECGs too

    In addition to 12-lead ECGs at 100% and 200%, there is a setting for Holter ECGs. It can also read Holter images with gray grids.

  • Works offline

    It runs from a single HTML file and does not use the internet. Copy it to an electronic medical record terminal and open it in Edge to use it inside the hospital system.

Download

Download

qtc.html is the app itself. The zip contains the app and nine sample ECGs to try right away.

PC version · recommended

The QTc measurement app

Unzip the file and open qtc.html in a browser. You can also try it right on this site without downloading (on a phone you can see what the screen looks like).

zip file (ver. 2.5) Open in the browser

Updated 2026.9.8

Phone version

Analyze from a photo

Take a photo of a paper ECG with your phone, drag a box around the two beats you want to analyze, and confirm. Accuracy is still not great.

Open qtc_mobile.html

How to use

How to use

  1. Open qtc.html in a browser

    Unzip the file and open qtc.html in Edge or Chrome.

  2. Paste an ECG image

    For a 12-lead ECG, the basic use is to cut out leads II and V5 and paste them, but any lead can be measured. Start by dragging a sample ECG from the zip into the drop area.

  3. Choose the display scale

    Choose the scale at which the ECG is displayed: 100%, 200% or Holter (ホルター). At 100% accuracy is a little uncertain; at 200% my impression is that it stays within about ±5 ms.

  4. Check the markers

    Check the markers on each beat. If they are off, you can correct them by hand.

    • R wave (blue ×)
    • QRS onset (green)
    • Tangent (red line)
    • Baseline (blue line)

    The baseline and tangent can be moved by dragging. Add an R wave with Shift + click.

  5. Paste the results into the medical record

    Press Copy results (解析結果をコピー) and text like the following is copied to the clipboard, ready to paste into an electronic medical record (the lead labels are in Japanese).

    2誘導 : HR 66, QT 344, QTcB 362, QTcF 356 (3beats)
V5誘導 : HR 67, QT 336, QTcB 355, QTcF 349 (4beats)
Example results screen (click to enlarge)

Movie

Video

Introduction video of the QTc app (in Japanese)

Calibration

Calibration

Calibration sets how many pixels in the image correspond to how many seconds. Normally it is read automatically.

Automatic

Each time you paste an image, calibration is read automatically from the grid. It used to be manual because automatic reading was not accurate enough, but the algorithm was rebuilt in ver. 2.4 and now reads within ±0.5% at 100%, 200% and Holter.

Manual

For images without a grid, when automatic reading fails, or when the value looks wrong, you can still adjust it manually.

Save once per computer

After adjusting manually, press Reset (再設定) to save the value on the computer. From then on, the saved value is used whenever you press the 100%, 200% or Holter button. The value depends on the display, so set it once on each computer.

Troubleshooting

Troubleshooting

The waveform is recognized wrongly
Choose Signal view (信号表示) to see the ECG waveform and QT range the app recognizes. If it looks wrong, check the calibration value and make sure Holter is not selected. A problem where normal ECGs were misjudged as Holter was fixed in ver. 2.3.
Still not fixed
Pressing the 100% or 200% button again can greatly improve recognition.

About

About this app

I have been hooked on AI lately and made an app that measures QTc by the tangent method. I started building it with Claude Opus 4.6, but I kept hitting the usage limit, so I also use Google Antigravity.

It is something like a toy made by an amateur, and its accuracy has issues, but please give it a try. I would be glad to hear what you think.

When I heard that Dr. Yoshinaga measures QT in three beats at each of five points in a Holter ECG, 15 beats in all, I thought I could never do that. With this tool, 15 beats can be measured with five drags.

Release notes

Version history

  1. ver.2.5Fixed the baseline floating above the waveform. Also fixed the Q marker being set on the downslope of the P wave and the last beat sometimes not being measured.
  2. ver.2.4Rebuilt automatic calibration. It also handles the gray grids of Holter ECGs.
  3. ver.2.3Fixed a bug where normal ECGs were misjudged as Holter and QT could not be measured at all. Also fixed the tangent being drawn at the T-wave peak, lengthening QT, and the Q marker moving back to the P wave.
  4. ver.2.2Fixed the Q marker sometimes being placed too late.
  5. ver.2.1Fixed the T-end label.
  6. ver.2.0Fixed a bug where the tangent was drawn on an atrial pacing spike.
  7. ver.1.8Fixed the baseline tending to drift down.
  8. ver.1.7Fixed the baseline tending to drift up.