Skip to main content

The first three winners of the MEDICA Start-up 2022 competition were selected from 12 finalists by an expert jury on November 15 in Düsseldorf, Germany. First place was awarded to health technology company Idoven for its powerful and scalable artificial intelligence (AI) platform for cardiology. Second place went to AlgoDX for its precision medicine AI platform targeting critical care. Finally, third place went to Reactive Robotics for its AI-based solution for robotic intensive care. Here is a presentation of these three promising companies.

1/ Idoven (Spain)

Idoven is a health technology company focused on the early detection and precision medicine of cardiovascular disease. Rika Christanto, COO of Idoven, explained:

“About a third of deaths worldwide are caused by cardiovascular diseases, which represents a trillion dollars in terms of global impact on the cost of the health system.

Today we tackle this problem by focusing on the electrocardiogram, the ECG. The ECG is the entry point for diagnosing any cardiovascular condition in any specialty area. Of course, doctors spend about a million hours a day interpreting and analyzing the ECG in order to diagnose their patients.

Willem collects ECG data from both hospital ECGs and home ECGs. He synthesizes it and assembles it. Then the deep learning algorithm processes it and provides an output that will be integrated into all DMEs. (Credit: iStock)

Idoven’s solution to address these challenges is Willem, an AI ECG analysis platform. Rika Christanto added:

“We believe we will build a digital cardiologist in the cloud to automatically detect the patient’s most clinically relevant cardiac events. Our system is already working with the precision of a cardiologist and it can deliver results in seconds. or minutes, depending on the duration of the ECG.

How it works? Idoven collects ECG data both from ECGs taken in the hospital as well as ECGs taken at home by implanted heart monitors, Holters, patches and even from the ECG-enabled smartwatch, etc. The company synthesizes the data and assembles it. Then its deep learning algorithm processes it in a secure cloud and provides an output that will be integrated into any electronic medical record (EMR) or cardiovascular information system.

Rika Christanto clarified:

“It’s not an ‘AI black box’: we can show the specific patterns in which we identify the condition.

We provide information to the cardiologist, primary care physician or oncologist. We aggregate and highlight the most clinically relevant information about the patient. But of course we believe that for AI to be trustworthy we need to have a human in the loop, at least for now. It is therefore not about autonomous decision-making.

Willem, the AI ​​ECG analysis platform from Idoven.  (Credit: Idoven)
Willem, the AI ​​ECG analysis platform from Idoven. (Credit: Idoven)

What is Idoven’s secret sauce?

The company claims to have one of the most diverse, largest and longest running ECG databases of symptomatic and asymptomatic patients, which is essential to creating the solution it has.

Idoven is already monitoring hearts around the world in collaboration with three groups: medical device companies to make devices smarter, pharmaceutical companies to identify cardiovascular diseases that are often underdiagnosed (particularly in primary care settings ) and high risk. patient groups.

2/ AlgoDX (Sweden)

AlgoDX has built an AI platform for precision medicine that targets critical care. What does it mean? David Becedas, Founder and CEO, explained:

“We’re focusing first on sepsis, which is one of the deadliest conditions we still have today and one of the most common causes of hospitalizations, deaths, and deaths in the hospital. hospital. Obviously, this is a major cost driver for our systems at all levels.

The challenge with sepsis is that this clinical syndrome can deteriorate patients very quickly: you can go from mild, treatable sepsis to life-threatening sepsis within hours. For every hour that antibiotic treatment is delayed, the risk of death increases by 7.6%.

Early detection is therefore really essential and this is where we want to make a difference with the use of machine learning in intensive care. Because sepsis is also very common in intensive care.

In this context, AgoDX has developed a machine learning algorithm that uses routinely collected intensive care and vital signs data. Around 2,000 data points are collected every day for an ICU patient, and AlgoDX uses this data in real time.

David Becedas said:

“These vital signs and data are all clues that together say nothing about a patient’s sepsis status, but when combined and processed by our algorithm, they can indicate whether a patient is trending toward sepsis. or not.”

AgoDX's machine learning algorithm uses routinely collected critical care and vital signs data.  (Credit: Siqui Sanchez/Getty Images)
AgoDX’s machine learning algorithm uses routinely collected critical care and vital signs data. (Credit: Siqui Sanchez/Getty Images)

AlgoDX does not have its own hardware, it integrates its machine learning platform with the EMRs of intensive care units. It could therefore receive real-time data from patients directly from the beds where they are treated. The system can stratify patients into low-risk, moderate-risk, or high-risk categories for developing sepsis within three hours.

He added:

“We actually developed our models in a sort of transparent box instead of a black box. So behind each prediction, the nurse or doctor can get more information and can understand the key contributing parameters every time, so basically they can drill down further if they want.

Diagnosing sepsis is not easy.  Image of the intensive <a class=care unit; (Credit: Shutterstock)” class=”wp-image-50760″ srcset=”https://news.google.com/wp-content/uploads/sites/9/jn-sepsis-share-1200w-1.jpg 1200w, https://news.google.com/wp-content/uploads/sites/9/jn-sepsis-share-1200w-1-300×200.jpg 300w, https://news.google.com/wp-content/uploads/sites/9/jn-sepsis-share-1200w-1-139×93.jpg 139w, https://news.google.com/wp-content/uploads/sites/9/jn-sepsis-share-1200w-1-768×512.jpg 768w, https://news.google.com/wp-content/uploads/sites/9/jn-sepsis-share-1200w-1-320×213.jpg 320w” sizes=”(max-width: 1200px) 100vw, 1200px”/>
Diagnosing sepsis is not easy. Image of the intensive care unit; (Credit: Shutterstock)

He precised :

“This is a point-of-care clinical decision aid that we have proven to have excellent predictive performance in a large randomized clinical trial. We also conducted a large health economic study to examine health system benefits as well as social benefits. And obviously, reducing the length of stay and reducing mortality in hospitals are essential.

This is where we want to make a difference with our platform. Sepsis is the first product we have developed, and we have more in the pipeline. Currently we are fundraising so I would be more than happy to speak with investors.

3/ Reactive robotics (Germany)

Showing a photo of a hospital patient, Alexander König, founder and CEO of Responsive roboticshas explained :

“What you see here is a Covid-19 survivor with complete lung failure. Thanks to our VEMOTION robotic system, this man was able to receive therapy in what was probably the most critical moment of his life. He’s up, he’s up, he’s having therapy without leaving his bed.

Thanks to our robot, this man was able to recover. We have developed the world’s first AI-powered solution for robotic intensive care.

"What three nurses can do manually, one person can do with the robot.  And our AI turns even a non-professional user into an expert." (Credit: responsive robotics)
“What three nurses can do manually, one person can do with the robot. And our AI turns even a non-professional user into an expert. (Credit: responsive robotics)

Living in an aging society with a shortage of nurses, treating patients as quickly as possible could be a bewildering problem. But according to Alexander König, there is a solution to achieve this: it’s called early mobilization, which has been shown to significantly reduce the cost per treatment while improving patient recovery.

“The problem? It’s completely manual, you need three nurses to do the therapy and so most patients can’t benefit from it at all because you don’t have those three nurses.

That’s why we developed VEmotion. What three nurses can do manually, one person can do with the robot. And our AI turns even a non-professional user into an expert.

Watch a video here:

After more than seven years of research, Reactive Robotics has been able to launch the first commercially available patent-protected AI for critical care. In view of their situation, they have already carried out thousands of operations.

Last year, the company won the prestigious European Technology Award, the euRobotics Technology Transfer Award for that. It was also recently awarded by the European Commission with the EIC Accelerator—a grant of 10 million euros to deploy this innovation throughout Europe.

Konig said:

“The market is worth more than €1 billion a year just for critical care robotics and we see this as a blue ocean opportunity.”

Each year, the MEDICA Start-up competition seeks out the most significant digital innovations in the healthcare industry. The competition aims to promote the innovative solutions of start-ups through networking and global visibility. From healthcare applications and AI-assisted healthcare data analytics to robotic assistance systems and new diagnostic approaches, innovators around the world can already submit their solutions for the 12th edition of the MEDICA Start-up competition. up in 2023.