January 31, 2021
By Charles C Onu, on behalf of the Ubenwa team
When I discovered, six years ago, that we could screen for birth asphyxia using AI and the mic on a smartphone, I knew we had the potential of offering swift and reliable screening tools for one of the top 3 causes of newborn mortality today. Birth asphyxia yet causes over 2 million deaths and disabilities every year.
Realizing this vision continues to require extended work and the assembling of many individuals with diverse skill sets. Now more than ever, the world needs diagnostic solutions that are accessible and cater to the most vulnerable populations. At Ubenwa Health , we remain committed to our mission of saving lives at birth using AI.
Realizing this vision continues to require extended work and the assembling of many individuals with diverse skill sets. Now more than ever, the world needs diagnostic solutions that are accessible and cater to the most vulnerable populations. At Ubenwa Health, we remain committed to our mission of saving lives at birth using AI.
The year 2020 was an incredible one for many. With over 2.1 million deaths and 100 million infections globally, the world experienced, and is still going through, the worst pandemic in nearly a century. The pandemic (necessarily) led to an interruption in our own work for several months, especially our clinical studies. However, we have since forged ahead the best we can, moving our core team fully online and resuming the studies with public health guidelines at the fore.
In this article, I highlight some of the strides we made in 2020 towards our goal of deploying rapid diagnostic tools for birth asphyxia. In addition, I provide a window into our plans for 2021.
Data fuels machine learning. We need to acquire large amounts of clinically-annotated cry recordings of both healthy and asphyxiated infants in order to train our algorithms to accurately and robustly identify the patterns of asphyxia. Following the success of our proof-of-concept study (achieving 86% sensitivity and 89% specificity in screening for asphyxia), in late 2019, we initiated clinical studies in Nigeria and Canada to develop and validate our algorithms in a much larger and more diverse population of newborns.
Beginning in January 2020, neonatologists at the Enugu State University Teaching Hospital have been hard at work recruiting and enrolling patients. Our burgeoning software engineering team has also been hard at work, responding to the needs of the clinicians by building a mobile intake application and back-office dashboards. Our clinical support team has been closely monitoring the study, offering research support and ensuring that data standards are met throughout.
We are grateful to the 227 parents and newborns who have participated in between all the pauses necessitated by the pandemic. We are targeting the enrollment of 1500 patients by the end of 2021. To do so, we are formalizing agreements with neonatologists in other hospitals over the next few months. If you, or someone you know is interested in initiating a clinical trial, write to us at firstname.lastname@example.org.
In early 2020, I participated in the annual national conference of the Paediatric Association of Nigeria (PAN). This was an important highlight of the year as it enabled me to speak to a broader base of health professionals (beyond the ones we already collaborate with), to understand their experience of birth asphyxia and to further validate the need for Ubenwa.
I interviewed over a dozen clinical professionals working across 9 different Federal Medical Centres (FMC) and University Teaching Hospitals in the country. In speaking with these neonatologists, paediatricians, medical residents and paediatric nurses, the problem of birth asphyxia was underscored. They complained of the persistent issue of late diagnosis and referral of asphyxiated patients, and all expressed interest in seeing Ubenwa close that gap by providing a means for early screening.
When the COVID crisis forced a months-long pause in our clinical studies, our engineering lead, Innocent, proposed using the break to redesign our website and incorporate a demo of our cry detection technology. This way, we could continue our outreach efforts and instil confidence in future partners. Cry detection is the first step to diagnosing asphyxia since we must first detect and isolate segments of an audio recording that contain cry samples to avoid spurious results from analysing non-relevant audio. The demo shows how this is possible using Ubenwa’s cry detection and localisation technology.
We graduated from the Creative Destruction Lab (CDL) in June 2020 after 10 months of intense programming. CDL is a nonprofit organization that delivers an objectives-based program for seed-stage, science- and technology-based companies. The program attracts some of the best MedTech entrepreneurs from around the world as mentors. This resulted in several enriching conversations for us during which we were able to reflect deeply on our short-term goals, our long-term mission and a connecting strategy between both.
We thank the individual mentors who guided us at different stages: Alexandre Le Bouthilier, Jean-François Pariseau, Sarah Catanzaro, and Inès Holzbaur. Special thanks to Alexandre who continues to listen to us every time we call on him.
We are most proud of being named one of 35 social ventures out of over 2,600 applicants to the MIT Solve Class of 2020. MIT Solve, an initiative of the Massachusetts Institute for Technology, is a program that helps social entrepreneurs to scale their work and impact in solving the world’s most intractable challenges.
In addition to being awarded a Solver in the Maternal and Newborn category, Ubenwa also received the AI for Humanity Prize by Patrick J McGovern Foundation (PJMF) and the Health Workforce Innovation Prize by the Abu Dhabi Crown Prince Court (CPC).
It was really an honour to receive these recognitions from Solve, PJMF and the CPC. It compels us to reflect upon the trust placed on us and the responsibility that we have to bring our mission to fruition — neither of which we take lightly. You can find my pitch to the Solve community here.
We participated in the McGill Clinical Innovation Competition (CLIC) where we emerged one of 5 finalists (MedTech companies). Special thanks to the brilliant Andrew Churchill who helped us prepare our presentation. Following Samantha’s pitch at the finals, Steve Arless, biotech inventor and renowned Canadian entrepreneur commented that:
“I want to congratulate you (Ubenwa)…in my many years of competition I’ve never seen such a compelling presentation as you gave today.” — Steve Arless
You can find a recording of the presentation at CLIC here.
I accepted opportunities to present our work and share our learnings to both scientific audiences and the general public. These gave chances to get feedback, to learn, and to inspire others.
Following interviews conducted in 2020, McGill University recently published a biographical article about Ubenwa: The birth of an idea. Below are a selection of other media articles featuring Ubenwa in 2020:
We have our work cut out for us this year. We are aiming to get Ubenwa to the milestone of readiness for regulatory clinical trials. If you would like to contribute in any of the areas below, do write to us at email@example.com.
We are presently working with new hospitals/clinics to join our clinical study. Do reach out if you are a neonatologist and are interested in collaborating.
We are designing the framework for clinical trials and regulatory approvals for Ubenwa. Do reach out if you have any thoughts or experience.
We welcome the contributions of audio or AI scientists/engineers who would like to provide their expertise in digging into the rich data we are collecting.
We will be developing mobile support for iOS (in addition to Android) while expanding the capabilities of our API with respect to both edge- and cloud-based ML.
As always, I would like to thank our team and advisors for their commitment and everyone who continues to contribute to our mission of saving newborn lives. I thank our existing partners and warmly welcome the organisations that joined forces with us in 2020: MIT Solve, PJMF, Abu Dhabi CPC and Schmidt Futures, as well as our latest clinical partner in Canada: Clinique MultiSens, a network of clinics specialised in audiology and speech pathology.
We will be publishing preliminary results from the clinical study as part of our Q1 updates this April. To stay in touch and receive our quarterly updates, you can subscribe here or via our home page.
6666 St Urbain St #200,
Montreal, QC H2S 3H1
Maryland Mall 350/360 Ikorodu Road