Tuberculosis (TB) is a serious infectious disease that affects the lungs, it is caused by the bacteria Mycobacterium tuberculosis. According to the World Health Organization (WHO), TB is the 13th leading cause of death worldwide, and is responsible for an estimated 1.6 million deaths each year. It is the second leading infectious killer – after COVID-19. (Source: World Bank) The majority of these deaths occur in developing countries where access to healthcare is limited, which is why it is imperative that we start finding ways to combat the spread and impact of this deadly disease.
Fortunately, with the advancements in technology, Artificial Intelligence (AI) can play a crucial role in the detection and treatment of TB. This can be done in numerous ways, including using AI to analyse large amounts of data to identify patterns and develop algorithms that can accurately diagnose TB. Additionally, AI can help predict the progression of the disease, and identify patients who are at high risk of developing severe complications.
“Ending the TB epidemic by 2030 is among the health targets of the United Nations Sustainable Development Goals (SDGs).” – World Bank
This is where our latest portfolio company; AI Diagnostics comes in. AI Diagnostics is a home-grown, South African, health-tech company whose technology has the potential to change the face of TB screening in Africa, and across the globe.
Dramatically improving TB diagnosis in Africa
Nurses carry a huge diagnostic responsibility in public healthcare yet they are, for the most part, ill equipped to do so efficiently. Even marginal improvements to the rate of misdiagnosis at the patient’s first exposure to the healthcare system, can make a significant impact, resulting in huge medical cost savings, while reducing delays to essential patient care.
AI Diagnostics, a Savant portfolio company, and part of the Savant BUILD programme’s 2022 cohort, has set out to reinvent diagnosis at patients’ first encounter with the healthcare system. Their custom-developed artificial intelligence algorithms analyse digitised diagnostic inputs obtained from low-cost and point-of-care diagnostic hardware to make key diagnostic decisions. AI Diagnostic’s portable digital stethoscope and accompanying AI model has been clinically tested for TB detection, demonstrating the potential to halve the number of TB-infected patients currently misdiagnosed or overlooked. The TB screening tool’s end users feel comfortable handling the technology because stethoscopes are already part of the nurses’ general clinical training.
So how does it work?
AI Diagnostics has harnessed the power of AI to develop a wireless digital stethoscope that detects TB by capturing and analysing patients’ lung sounds. AI Diagnostics has trained a state-of-the-art deep neural network on thousands of validated TB-positive and negative lung sounds. When new patients are screened, these recordings are processed by AI Diagnostics’ proprietary neural network which provides immediate TB diagnosis instructions to the nurse conducting the screening. These TB insights ensure that TB-negative patients are appropriately filtered out, ultimately reducing the cost and time spent on superfluous lab testing which could potentially save South Africa R300M annually in reduced lab costs alone. This tool has also been shown to identify up to 40% of patients that are currently being completely missed by the predominantly-adopted symptom-based TB screening method. AI Diagnostics has achieved a screening accuracy surpassing the WHO’s TB targets, setting them on the path for WHO endorsement.
In qualified hands, AI Diagnostics portable screening tool has the potential to halve the number of TB-infected patients currently misdiagnosed or overlooked.
Consumable-free, portable, wireless, and easy-to-use, the unit is perfectly suited to the African reality. The display guides users on how and where to record a patient’s bodily sounds, making it extremely user-friendly and intuitive for users with limited knowledge – without jeopardising the integrity of the results. And, it comes complete with its own local processing unit, allowing offline use. All of this means that patient care and diagnosis is unaffected by power outages. Data is simply backed up to a cloud-based patient registry when back online.
“We are very fortunate from a Savant perspective, to work with a number of deep tech start-ups. Savant has a specific mandate to concentrate on deep tech start-up developments – and this is why our investment into AI Diagnostics is so relevant to what we do.” says Savant CEO; Nick Allen. “This type of application is a brilliant example of how AI and deep tech can be used to solve one of Africa’s biggest problems. Savant is thrilled to be working with the AI Diagnostics team to help them on their development journey. Africa has numerous significant challenges to face and a majority of these challenges are going to be solved through the use of deep tech solutions, such as AI Diagnostics’ solution,” says Nick Allen.
Braden van Breda, AI Diagnostic’s CEO had the following to say: “Joining Savant’s BUILD Programme has been a game changer. We’ve been exposed to some phenomenal sessions with industry specialists, some of whom are now partners. I’ve enjoyed a host of stimulating ideas and options presented by the Savant team and cohorts. The Savant team has helped us focus, distill and frame our vision in a communicable and executable way.
Working with the Savant team has also exposed us to both essential customers and investors, I have started to realise the weight of being “Savant endorsed”. Through their extensive portfolio, we continue to benefit from introductions to relevant suppliers and manufacturers. They have been essential in co-developing our go-to-market strategy and supporting that strategy with their network. I believe Savant has understood our vision and supports it. I feel fortunate to have them alongside us in this venture.”