Hedgehog Medical Inc.

Lung ultrasound analysis is extremely qualitative in nature and while there are established protocols for acquiring the images the interpretation of the images remains difficult. It takes substantial training and experience to interpret the images correctly making it less accessible for non specialists to use.

We increase access, reduce costs and improve patient outcomes by using computer vision to analyze emergency lung ultrasound videos in real-time and automatically generating easy to understand reports.

Hedgehog Medical has developed a proof of concept that can both count the B-lines and measure the thickness of the pleural line. Our next steps include working with doctors to improve the algorithm and classify different types of lung diseases.

Our software then aggregates that data into a simple to share Lung Ultrasound Report. Further, this report will improve patient monitoring over time, enhance communication across teams and ensure effective documentation with no duplication of work.

In the Emergency room, minutes and even seconds count. Task saturation for emergency physicians is common. A survey by Emergency Physicians Monthly magazine published that of those respondents who had under utilized emergency ultrasound machines in the ER; 70.2% were either too busy or did not have the proper training/skills for interpretation.  

The related problem is that Emergency rooms are by nature non-routine and this leads to challenges regarding proper documentation and billing, which can easily be missed. The same survey indicates that 52.8% of these procedures are not billed.   

Long term care facilities typically do not have CT or X-ray imaging capabilities. However do have registered nurses on staff. The cost of annual patient transfers just in Ontario is 283 million per year. Further a bed in the hospital costs $842 per day while a long term bed costs only $126 per day.

The statistics above indicate that it takes substantial training and experience to interpret the images correctly making it less accessible for non specialists to use. 

We increase access, reduce costs and improve patient outcomes by using computer vision to analyze emergency lung ultrasound videos in real-time and automatically generating easy to understand reports.

We serve emergency physicians, sonographers, registered nurses of long term care homes, and emergency medical technicians. With the use of this technology lung ultrasound analysis will be more accessible for non specialists, it also provides faster and more accurate triage information directly impacting patient outcome and saving lives. 

Our solution is directly in line with the accurate detection and rapid response aspects of this challenge. Using AI and computer vision we are able to provide quicker triage and decision making for patients suffering from lung diseases, this includes COVID-19, pneumonia and other diseases like COPD. 

This technology can also be used in low resource settings such as as Africa and India where people do not have the capabilities to pay for expensive machinery such as X-Ray and CT scans.

Currently there is no other FDA approved software solution in the market that uses AI for lung ultrasound analysis. We would be the first movers, therefore, our competition is primarily the current standard of care. We are better than the current standard because our solution:

• Reduces the time required for scanning and interpreting from 10 mins to less than 5 mins 

• Is significantly cheaper than current standard of care alternatives 

• Increases access to a life saving tool by lowering the training barrier 

• Reduces the number of patient transfers due to respiratory diseases, particularly from long term care homes

• Improves outpatient monitoring that may allow for early release from ICU 

One significant existing limitation of current lung ultrasound analysis is the fact that it is qualitative in nature. Doctors have told us that creating an AI system to automatically track and measure different features in a lung ultrasound can rapidly improve their workflow. This is because an AI-assisted tool to automatically quantify features from lung ultrasound data will decrease both inter- and intra-operator variability. Therefore we are using computer vision and deep learning to automatically extract features from lung ultrasound data. These features include: the number of B-lines, the number of A-lines, the pleural line thickness, any pleural line irregularities, and the percentage of the pleural line that contains B-lines. Then using machine learning models, such as deep learning, we can map these features to common lung ultrasound diseases such as pneumonia, COPD, or covid-19.

We already have a working proof of concept that can quantify b-lines and detect the thickness of the pleural line which is absolutley crucial in detecting lung diseases like COVID-19.

The link below has a 3min pitch about our company:

https://hedgehogmedical.com/lu...

Our proof of concept can be found at 2:22 mark of the video.

https://docs.google.com/presentation/d/1_cmx_gUEJZsMxNgAJ0eyroYoXcoVVrEaTgNyrpdqwn4/edit?usp=sharing

Currently we are not serving anyone. Our solution is still in the product development phase. 

One year from now, we probably would have just clear FDA approvals and other regulatory approvals and will be serving close to 5 hospitals in the US. 

Five years from now we will be in roughly 860 hospitals around North America which roughly equates to $8.6million in revenue and will be directly improving patient outcome. 

Within this next year our primary goals are as follows:
-Product development
-Clinical trials
-FDA & Regulatory Approvals

Within the next 5 years we plan to be in our growth phase and would have expanded to hospitals around the world. We will deploy this technology with low resource settings around the world. We have already begun discussions with Doctors Without Borders so they can use our product in remote locations around the world. 

The biggest barrier to entry for us is our FDA and regulatory approvals. Furthermore, this specific market/solution is getting very popular and therefore time is another crucial factor for us, we need to move fast with our development and this prize money will help us a lot in accelerating our growth. 

We are overcoming this barrier by making strategic partnerships with emergency room doctors who use ultrasound on a daily basis. We have secured 3 doctors from the US who will be on our clinical advisory board and will help us develop this technology. These same doctors will help us with clinical trials to prove that our solution works, these trials will be extremely helpful in passing FDA approvals.  

NA

Currently the core team of five is working half their time at Hedgehog Medical. Upon raising funds the team will be moving to full-time. When the global pandemic started, the team met and asked how we can contribute to the solution. Our plan to go full-time as soon as possible. In addition to the core team, we have a clinical advisory board of 3 medical doctors who specialize in lung ultrasound.

Our team consists of two PhDs in computer vision, two masters in business, and a full stack software engineer. Before starting on our lung ultrasound tool, in 2016 we built a computer vision tool to analyze arteries from ultrasound. This arterial software tool was sold to universities around North America and is still actively being used. Given the coronavirus pandemic, the team convened and decided to build another ultrasound tool. We started this new tool in March 2020 when the coronavirus was declared a pandemic by the World Health Organization.

We are currently setting up collaborations with two hospitals, one in Canada and one in the states. The first hospital is Cooper Health University in New Jersey and our contact there is Dr. Sharad Patel (MD). Cooper Health University has one of the largest ultrasound clinics in the USA. The second hospital is William Osler Health System in Ontario Canada. We are currently finalizing the contract with William Osler Health System in order to receive anonymous patient data from covid-19 patients. Finally, we are also currently working with a point of care ultrasound company called Clarius. Clarius is providing us access to their low-cost ultrasound probes.

Our main customers will be hospitals. The way we reach these hospitals is 2 ways. Firstly, we will partner with point of care ultrasound manufactures and license our software to them. The second method is to directly integrate into the hospitals picture archiving and communication systems known as PACS. When we integrate directly into the PACS system we can charge the hospitals on a per report basis. According to our research we can charge the hospitals around $10 per report. A small scale hospital in the US performs roughly 2,500 lung ultrasound exams every year. 

Currently we are looking to raise investment capital. The money we are raising will help us get to FDA and Regulatory approvals. Once we have those approvals we can start charging for our product and selling it to hospitals. As mentioned above we can charge $10 per report and an small scale hospital in the US conducts roughly 2,500 exams every year. In 5 years we project we could be making close to $8.6 million in revenue. 

In a time where COVID-19 has changed the world, we need more solutions that allow for quicker triage and diagnosis of patients who show signs of lung disease. Doctors around the world have told us that they need faster lung ultrasound tools using computer vision. We have a history of building reliable tools for ultrasound. Given we are the team to build this product we need to Solve for funding and mentorship. Together we can directly improve patient outcome and help our essential workers make better and faster decisions.

We believe that we can't succeed alone, therefore we are actively looking for collaborators to bring this technology to market to save lives. We are aware that we are an early stage company and we know that partnering with other organizations will help scale faster. 

- The Bill and Melinda Gates Foundation would be a great partner for the work we will be doing in low resource settings with Doctors Without  Borders. 

Our solution directly helps doctors and nurses in the ER who are working non-stop during the COVID pandemic. This solution allows for quicker triage with safer methods significantly reducing task saturation for these doctors and nurses while also improving patient outcome. The doctors we have spoken to so far have mentioned time and again that a solution such as ours would be a "game changer" triaging lung diseases using ultrasound around the world. 

We will use the money for product development, data acquisition and our IP strategy.