For HMOs
An HMO (a health maintenance organisation) is the company that pays for its members' care. When a claim includes an ECG, what reaches you may be a phone photo of a printout. TraceRead gives your medical officer an AI first read of that photo, to compare with what the claim says.
TraceRead does not decide a claim
It does not pay, refuse or score a claim, and it does not tell you whether a claim is true. It shows what the tracing shows, so a qualified doctor on your team can compare it with the claim. Every AI first read is for that doctor to check.
What changes
Four things change for the medical officer with a claim in front of them. The numbers behind them are further down this page.
01
Your medical officer gets the rebuilt trace, the measurements (heart rate, PR interval, QRS, QT, QTc and heart axis) and the patterns TraceRead AI sees, each with a score from 0 to 100, to set beside what the claim says.
02
The rebuilt trace is drawn back over the photo as a red line, so the medical officer checks it against the page and sees which parts of each lead the paper shows.
03
When the page is not found, or the rebuilt trace looks unreliable, TraceRead gives no AI first read and asks for a new photo. It does not guess.
04
Before a photo is sent, the header, where the machine prints the person's details, is covered with a box on the phone. TraceRead asks for no names, only age and sex.
Words used on this page
Your day
Your claims process stays as it is. Four steps follow.
From the clinic or diagnostic centre, as it always does.
Someone on your team sends it through the capture app, which runs in a phone's web browser, after covering the header where the machine prints the person's details.
It finds the page, rebuilds the 12 tracings, measures them and lists the patterns it sees. If the page is not found or the rebuilt trace looks unreliable, there is no AI first read.
A qualified doctor on your team compares the AI first read with what the claim says, and decides what to do. TraceRead does not decide the claim.
The numbers
Each number says how many ECGs it was measured on and where it came from. They come from public research ECGs and from photos made by computer, not from patients at a clinic.
0.908
median match between the rebuilt trace and the original recording
Measured on 500 photos made by computer of printed PTB-XL test ECGs. Source: Spike S1.
99.0%
page layouts read correctly
Measured on 500 photos made by computer of printed PTB-XL test ECGs. Source: Spike S1.
7 of 500
photos where the page was not found, so no first read was given
Measured on 500 photos made by computer of printed PTB-XL test ECGs. Source: Spike S1.
5.9%
photos held for a retake because the rebuilt trace looked unreliable
Measured on 500 photos made by computer. Source: Spike S3.
0.902
how well 71 patterns are told apart (AUROC, where 1.00 is perfect and 0.50 is a coin toss)
Measured on 2,198 PTB-XL test ECGs (fold 10), original recordings. Source: Spike S2.
0.009
drop in that score when TraceRead AI reads photos made by computer instead of the original recordings
Measured on 500 PTB-XL test ECGs, photos made by computer. Source: Spike S3.
These are engineering measurements, not clinical claims. The page on how TraceRead AI works shows every step of a real run with its numbers. See the evidence
Features
TraceRead has ten named features. These six are the ones that matter when the photo comes from someone else.
TraceRead AI finds the four corners of the paper and how it is turned, even in a photo taken at an angle.
TraceRead AI follows the ink of every lead and rebuilds it as a digital signal, 500 points a second.
Marks exactly which seconds of each lead the paper shows, because most printouts show each lead for only 2.5 seconds.
When the rebuilt trace looks unreliable, TraceRead gives no first read and asks for a new photo.
TraceRead AI checks the tracing for 71 patterns of rhythm, conduction, thickened heart muscle and ST and T changes, each with a score from 0 to 100.
Before sending, the technician drags a box over the printed header where a machine prints the patient's details.
Safety
TraceRead helps a doctor read an ECG. It does not stand in for one.
An AI first read is not the last word.
Pricing
HMOs and groups get one volume price across all their sites, priced per AI first read. Talk to us for it. Prices are in naira, with dollars as approximate.
HMOs and groups
Volume pricefor many sites
priced per AI first read
Pay as you go
₦1,500per AI first read
about $1.10
Dollar figures are approximate, at ₦1,370 to the dollar.
Questions
An AI first read: the rebuilt tracing drawn over the photo, the measurements and the patterns TraceRead AI sees, each with a score from 0 to 100, and a map of which parts of each lead the paper shows. It is for a qualified doctor to review.
No. It does not pay, refuse or score a claim. A qualified doctor on your team compares the AI first read with the claim and decides.
When the page is not found or the rebuilt trace looks unreliable, TraceRead gives no AI first read and asks for a new photo. In testing it held 5.9% of 500 photos made by computer.
Printed 12 lead ECGs from adults (18 and over), at the usual 25 mm a second and 10 mm per millivolt, laid out as 3x4, 6x2 or 12x1, on A4 or US Letter paper. Photos of monitor screens, rhythm strips on their own and ECGs from children are not supported.
Patient data stays in the AWS Cape Town region, and it is never sent to an outside AI service. TraceRead needs only age and sex, never a name, and the header is covered on the phone before the photo is sent. TraceRead staff cannot see patient data by default. The HMO is the data controller, and TraceRead signs a data processing agreement with each customer.
We show TraceRead reading public research ECGs, and talk through how claims with ECGs reach your medical officers.