Deepecho

The WHO recommends at leat one fetal ultrasound during a pregnancy. However quality fetal ultrasound is far from available in countries of the global south as well as the USA. This examination indepndently reduces fetal and maternal poor outcomes. Because of the flourishing of cheap portable ultrasound machines, the hardware is not the bottleneck amymore, it’s the software. At Deepecho we automate parts of the screening fetal ultrasound scan to streamline processes for trained sonographers and allow minimally trained operators perform part of the examination.

Our webapp supports ultrasound images and video from all sources and all types of ultrasound machines. Our AI algorithms are able to detect and segment relevant structures in real time to help the operator perform biometric measurements and take optimal images of the standard planes.

Our target population is the vast number of pregnant people all over the world who currently do not have access to quality prenatal ultrasound because of socio-economic or geographic reasons. Predominantly in countries of the Global South, but in the USA as well. We focus on the screening scan which is the most basic and vital ultrasound examination. By democratizing access to fetal ultrasound, we will enable early diagnosis of life threatening conditions such as fetal growth restriction, placenta previa, and preeclampsia.

We are a diverse team composed of medical doctors, ML and data engineers form Morocco and the USA. Our doctors both in managerial positions and annotation processes are all trained in fetal ultrasound and have deep knowledge of patient management and treatment. We are all committed to drastically reduce fetal and maternal poor outcomes in our country, the Global South, and the world at large.

At present, our team is working on the assembly of comprehensive documentation required for a 510k premarket approval planned for end of April.

We have secured an agreement to implement our solution across 20 healthcare facilities located in New York City. This deployment is scheduled to commence immediately upon receipt of the 510k premarket approval.

We have led prospective clinical studies in Morocco to test the robustness of the algorithm and published our results in Nature Communications on November 3rd 2023 under the name: "Fetal biometry and Amniotic Fluid Volume assessment end-to-end automation using Deep Learning"

We are currently gathering USA ultrasound data for the purpose of the FDA 510k premarket approval.

Finally we are currently leading prospective clinical studies in Morocco to automate the screening for preeclampsia and Post Part Hemorrhage 

DeepEcho is applying to Solve because we believe in the power of collaboration and the value of being part of a global community that is committed to solving the world’s most pressing challenges. We see Solve as a platform that can provide us with the opportunity to connect with a diverse network of leaders, innovators, and change-makers who can help us scale our solution and maximize our impact.

Specifically, we hope Solve can help us overcome the following barriers:

1. Market:  Expanding our solution to other markets poses a significant challenge. We hope to leverage Solve’s global network to gain insights into different markets, understand their specific needs and regulations, and build strategic partnerships.

2. Cultural: The adoption of AI in healthcare can be hindered by cultural barriers and lack of trust in technology. We hope to learn from the experiences of other Solver teams and partners on how to effectively address these issues and increase the acceptance of our solution.

3. Legal: Navigating the complex healthcare regulations in different countries can be daunting. We hope to gain guidance and support from legal experts in Solve’s network to ensure our solution complies with all relevant laws and regulations, especially for further more challenging solutions we are developing (breakthrough device denomination)

We understand that Solve is more than just a funding program. We are excited about the opportunity to be part of a community that is committed to driving social impact and look forward to contributing to and learning from this community. We believe that being a Solver team can significantly accelerate our journey towards ensuring every pregnancy is safe and every newborn has the best start in life.

Deepecho’s approach to addressing fetal health issues is both innovative and transformative. By leveraging the power of artificial intelligence (AI) and machine learning (ML), DeepEcho has developed a software that can analyze ultrasound images of fetuses and diagnose developmental issues. This is a significant improvement over traditional methods, which rely heavily on the expertise of medical professionals and can sometimes miss subtle signs of potential problems.

Our solution could catalyze broader positive impacts in the healthcare sector by setting a new standard for prenatal care. The use of AI in diagnosing fetal health issues could inspire other healthtech companies to explore similar applications of AI, leading to advancements in other areas of healthcare. This could result in improved patient outcomes, reduced healthcare costs, and increased accessibility to high-quality healthcare services.

In terms of market impact, our solution could revolutionize the prenatal care market. By providing a more accurate and efficient method for diagnosing fetal health issues, DeepEcho could potentially reduce the number of preventable poor outcomes. This could lead to increased demand for our product, driving growth in the healthtech sector.

Finally we, at Deepecho, are outcome focused: what matters to us is lowering fetal and maternal poor outcomes worldwide. That is why we recently launched clinical studies to allow non-trained operators screen for preeclampsia, and postpartum hemorrhage, the two largest causes of maternal poor outcomes on the planet. As far as we know, no other company has started work in that direction, focusing on ultrasound.

Goal: The ultimate outcome that Deepecho strives to achieve is to lower fetal and maternal poor outcomes worldwide. This is a significant improvement in health outcomes for both mothers and babies.

Activities: DeepEcho develops AI and ML software to analyze ultrasound images of fetuses and diagnose developmental issues. We also conduct clinical studies to screen for preeclampsia and postpartum hemorrhage. These activities are carried out by our dedicated team of researchers and developers.

Products or Services: The direct output of our activities is the AI and ML software that we develop. This software provides a more accurate and efficient method for diagnosing fetal health issues. It reduces the reliance on the expertise of medical professionals, making high-quality healthcare more accessible.

Changes in People’s Lives: The real changes in people’s lives that result from our outputs are improved patient outcomes, reduced healthcare costs, and increased accessibility to high-quality healthcare services. By providing a more accurate and efficient method for diagnosing fetal health issues, we can potentially reduce the number of preventable neonatal poor outcomes. Our focus on preeclampsia and postpartum hemorrhage could significantly reduce maternal poor outcomes. These outcomes will lead to the achievement of our goal of lowering fetal and maternal poor outcomes worldwide.

Our theory of change is supported by the growing body of research highlighting the effectiveness of AI and ML in healthcare. Studies have shown that these technologies can significantly improve diagnostic accuracy and efficiency. Moreover, our ongoing clinical studies provide direct evidence of our solution’s impact on maternal and fetal health. We measure our success by the lives we can potentially save and the positive changes we can bring about in global health.

DeepEcho’s solution directly addresses the United Nations Sustainable Development Goal 3 (SDG 3) for Good Health and Well-being. SDG 3 aims to ensure healthy lives and promote well-being for all at all ages, with a specific target (3.2) to end preventable poor outcomes of newborns and children under 5 years of age.

Deepecho’s AI-powered software is designed to screen for fetal and maternal abnormalities and address issues related to preterm birth and low birth weight, which are significant contributors to neonatal poor outcomes. By diagnosing Fetal Growth Restriction (FGR), one of the leading causes of neonatal poor outcomes, DeepEcho is directly contributing to reducing neonatal poor outcomes, thereby aligning with the objectives of SDG 3.

Furthermore, DeepEcho’s solution could potentially increase access to quality healthcare services, particularly in low-resource settings where access to skilled healthcare professionals may be limited. By leveraging AI, DeepEcho can provide accurate diagnoses regardless of the location or resources available, thereby promoting health equity - another key aspect of SDG 3.

In the future, DeepEcho plans to continue its work on in vivo clinical trials and publish our results, specifically on preeclampsia and postpartum hemorrhage, further contributing to advancements in prenatal care and reinforcing antenatal diagnosis. This ongoing commitment to innovation and improvement further underscores DeepEcho’s alignment with SDG 3.

Deepecho leverages a novel and proprietary dataset of annotated fetal ultrasound images, from 7 facilities in Morocco and 20 in the state of New York, which is one of the world’s largest of its kind. This dataset is crucial for training our AI and ML algorithms to recognize standard views and detect key standard planes with high accuracy. The algorithms are also able to measure the main fetal structures with extreme precision.

In terms of acquiring good, curated data, Deepecho partners with medical professionals and healthcare institutions from around the world to collect ultrasound images. These images are then annotated by our team of 12 medical doctors to ensure the accuracy of the data. This process ensures that the data used to train our AI models is both high-quality and relevant.

We also developed a proprietary way to gather video data directly from live ultrasound scans to augment data density which is currently deployed in 5 machines in Casablanca and 7 in New York.

Finally we are conducting in vivo prospective clinical studies providing us with more data that is tailored to our needs and those of our patients.

The underlying models vary from traditional computer vision models to transformers depending on the performances.

For more details on our technology, please refer to ou Nature Communications paper:Nature communication paper by Deepecho (Slimani et al.)

Full time staff: chief executive officer, chief technology officer, chief medical officer five data scientists, two data engineers, three ML engineers

part time: Chief science officer, clinical and data lead, 12 medical doctors for data annotation

Since August 2020

Deepecho is committed to incorporating diversity, equity, and inclusivity into our work. This commitment is reflected in various aspects of the organization.

Diversity in Leadership: We believe in the power of diverse perspectives and experiences. The leadership team is composed of individuals having grown and for some educated in Africa, bringing a wealth of knowledge and unique insights to the table. This diversity fosters innovation and drives the development of a solution that caters to a wide range of needs. 

Equity in Access: Our mission is to reinforce antenatal diagnosis and prevent poor outcomes, particularly in vulnerable populations. By leveraging AI, Deepecho can provide accurate diagnoses regardless of the location or resources available, thereby promoting health equity. This is particularly important in Morocco, and other countries of the Global south, where access to quality healthcare services may be limited.

Inclusivity in Solution Development: Our solution is designed with inclusivity in mind. The AI models are trained on a diverse dataset of annotated fetal ultrasound images from Africa, North America and Europe, ensuring that the solution is effective across different populations. 

Actions Taken: To achieve these goals, we have taken several actions. These include actively recruiting individuals from diverse backgrounds, collaborating with local healthcare providers to reach underserved populations, and continuously refining our AI models to ensure they are inclusive and effective

Deepecho’s business model is designed to create value in the prenatal healthcare sector through the application of AI and ML technologies. Here’s how it works:

Revenue Generation: DeepEcho is currently raising institutional capital to cover its expenses. The commercialization of our solution is planned to start in Q2 of 2024, with one client already lined up awaiting FDA 510 clearance.

Partnerships: We are partnering with Butterfly Network to offer our solution directly to consumers as part of their AI marketplace, Butterfly Garden. This partnership expands our reach and allows us to serve a larger customer base.

Pricing: The cost per API call is low, which allows us to offer our services to low and middle-income countries. This pricing strategy not only makes our solution more accessible but also aligns with our mission to lower fetal and maternal poor outcomes worldwide.

Value Proposition: Our solution uses AI and ML to analyze ultrasound images of fetuses and diagnose developmental issues. This offers a more accurate and efficient method for diagnosing fetal health issues compared to traditional methods, potentially reducing the number of preventable neonatal deaths and birth defects.

Grants: DeepEcho is actively seeking additional funding opportunities to further our mission. We are currently in discussions with the Bill and Melinda Gates Foundation for a potential grant. This potential funding could significantly support our efforts to improve prenatal care and reduce fetal and maternal poor outcomes worldwide. It’s important to note that while these discussions are promising, the grant is not yet confirmed. We are hopeful for a positive outcome that will enable us to continue and expand our impactful work in the healthcare sector.

Our primary revenue stream is through service contracts with healthcare facilities. We have a partnership with a chain of clinics in New York, comprising 20 facilities, which is a significant client awaiting 510k clearance. This partnership is expected to generate substantial revenue once the clearance is granted.

In addition to service contracts, we are actively pursuing grants and venture capital. We have been successful in this regard, having recently raised a seed round.

While we are not at liberty to disclose specific monetary amounts, we can confirm that the funds raised so far have enabled us to advance our technology and expand our partnerships. This progress is a testament to the viability of our financial plan.

In the long term, we anticipate that the growth of the share of t service contracts, over grants, and venture capital will cover our expected expenses, ensuring our financial sustainability.