Ah, always good to get a user's experience @Aaronfzr !

Just to make sure we're on the same page, — are you maybe thinking of EchoGo by Ultromics? Echo IQ’s product is called EchoSolv, and it works a bit differently. Instead of analyzing echo images or loops, EchoSolv interprets echocardiogram measurements (something Philip pulled me up on today too!).

I was thinking of EIQ echosolv, but this shows how little I actually understand of the product (and believe me, Ive tried to figure it out). Probably they aren't marketing to potential end-users.

This is an interesting and important clarification though.

To humans, Echo studies just videos. The measurements come from the sonographer or cardiologist who looks at the images and makes various measurements (shape of this, diameter of that). There is a surprising amount of inconsistency between users deciding which particular bit counts and which doesnt (hence why echo is "user-dependent" or subject to "inter-individual variability). If EIQ is taking the measurements created by sonographers, and plugging them into an algorithm, Id be worried the results are susceptible to user error and Im not that excited.

If he meant though, that they use the raw echo data used to create the images/loops, and generating their own measurements, then that is very interesting. It takes out what has always been a limitation - the human eye's ability to differentiate between various shadows created as a visual representation of the underlying data!

Exactly @Aaronfzr !

The whole point about such radiological studies and measurements is the “kappa” - inter-observer variability. I’m afraid I haven’t done the work on this one to offer an informed view, but it is usually an under-reported statistic. Often because a low Kappa basically undermines the whole thesis. So they don’t publish it.

c

Hi everyone @Aaronfzr @Scoonie @Strawman @Chagsy , I finally got to watching this meeting back having been busy doing some exams on, you guessed it, echocardiography! Now I'm absolutely not an expert but I am interested enough to take a look at this business.

I didn't find Philip as vague as you @Scoonie in that he answered the question about ICardio clearly. He explained that their software is analaysing the images themselves and deriving measurements as opposed to EchoIQ whose tech is entirely about assessing the data points (as collected by the songrapher) using a proprietary algorithm. He also made the point that a software which improved the acquisition of images would likely benefit EchoIQ as they would feed better data into the algorithm. I can conceive of a world whereby there was a health insurance reimbursement code that reimbursed for 1) Software based assessment of Echo images and a separate one for 2) Software bases assessment of Echo datapoints. I could also see a world where both of these small providers might be swallowed up by a much bigger business (Phillps, GE, Siemens etc).

One point which resonated me with regards heart failure is the potential difficulties in diagnosing it. Clearly cardiologists are in a better position than I (as an anaesthetist) to make this call, however, it can become very complex very quickly looking at different data points with regards diastolic dysfunction for example and I can absolutely believe an AI algorithm would do it better than most humans and the two combined would be even better as discussed. Heart failure is a big big deal and even a small improvement in detection and classification would be valuable from a human and financial perspective.

For me it's not quite investable yet having been burnt by biotechnology in the past (looking at you paradigm) but I will watch with interest.