Solving drug-resistant TB using AI

We harness the power of big data and artificial intelligence to solve for drug-resistant TB.

The global incidence of drug-resistant TB, which experts describe as the greatest threat to human health, is on the increase. This deadly bacteria can either be acquired:

·      Through transmission in the air;

·      When patients are given sub-optimal treatment plans; or 

·      When they fail to adhere to their anti-TB medication. 

Developing countries, such as India, China and South Africa, are most burdened by the disease due to its high prevalence in these areas. The high costs associated with anti-TB medication further exacerbates the burden on these nations already constrained public health budgets as drug-sensitive (DS) , multidrug (MDR) and extreme drug resistant (XDR) TB costs $250, $ 6,700 and $26,000 respectively per patient to treat.

Secondly, complete adherence to TB medication is pivotal to the cure of this infectious disease. TB patients are often prescribed with treatment regimens which do not effectively cure their disease. AI techniques which make use of big data and predictive algorithms for the identification of drug resistant strains and the recommendation of the optimal treatment plans has been demonstrated for diseases such as cancer. To solve for the complex challenge of drug resistant TB, we propose an all-encompassing system that will harness the power of big data on TB and AI to offer a truly personalized approach to the diagnosis and treatment of the disease.

The current project is innovative in comparison to others proposed by other similar organizations as it will be one of the first attempts to make use of historical datasets on TB and artificial intelligence (AI) to optimize the diagnosis and treatment of the disease. 

Our solution is driven by technology. We make use of historical datasets on TB and artificial intelligence (AI) to optimize the diagnosis and treatment of the disease. Our AI system will be able to effortlessly analyze these large datasets and enable medical practitioners to uncover hidden patterns in these datasets that are pivotal to diagnosing and treating drug-resistant TB. 

Our goal is to further validate our solution by inputting more data into our AI system so that the accuracy of its predictions increases. We then aim to pilot the system at one of the local TB healthcare facilities for doctors to use.

Our vision is to build a robust, user-friendly AI platform that can be used by doctors to treat TB patients. We aim to first implement the technology in South Africa, then scale into other countries such as China and India where the disease is most rampant.

Once we have built a robust AI platform that is able to diagnose and treat TB patients accurately, the predictive models will be integrated into a mobile app that doctors can use in a hospital setting. The adoption of smartphones which make use of mobile apps is on the increase, particularly in the developing countries we are targeting. Our solution will, therefore, be available for doctors to use in urban, semi-urban and rural communities who have access to a smartphone or a computer.

There are over 400 000 cases of TB in South Africa every year, approximately 20,000 are drug-resistant.  Once we have built a robust AI platform that is able to accurately diagnose and treat TB patients, our primary target market will be this region which has one of the leading global incidences of the disease. 

As a start, we aim to serve approximately 200,000 TB patients over the next 12 months which represents 50% of the total TB population in South Africa. We believe that this target is achievable given that the TB research, diagnosis and treatment community both locally and internationally is relatively integrated. We have already started working with local researchers who would like to engage with our product. Over the next 3 years we aim to enter into markets such as China and India where the disease is also rampant.

Our team is currently composed of an entrepreneur who has prior experience developing medical devices, a genomics researcher, a programmer specialising in deep learning,  2 software developers and a former clinician-turned software developer. Together, we believe that our team has the required skill set to bring our proposed solution into being and to make a meaningful impact in society. 

Our primary customers will be government and private hospitals, health care related NGOs and health insurance companies who are the major dispensers of the medication. The key value proposition for this customer segment will be the cost savings they acquire from diagnosing and recommending the correct TB treatment when drug resistance occurs. Our revenue streams will come from an annual subscription payment to make use of our assisted decision-making platform. 

I believe that there is much synergy between the work conducted by our startup company and the solutions that Solve supports. We would like to make a positive contribution to solving for effective, accessible and affordable healthcare in the world.

Our team would like to start building global networks for when we scale our solution. Solve attracts many international businesses and organizations which can assist our team in making these connnections.