minoHealth

Democratising Quality Healthcare with Artificial Intelligence, Data Science and Cloud Computing.

In Africa, there’s as high as 50,000 patients per doctor ratio in some nations and 9.2 million deaths mainly from communicable and non-communicable diseases. Quality, affordable and accessible healthcare has been a great challenge for Africa. And so in order to democratize some quality and affordable healthcare for all in Africa, we developed minoHealth where we use Artificial Intelligence for Medical Forecasts, Diagnoses and Prognoses. We've developed AI(Deep Learning) systems for conditions like Breast Cancer, Pneumonia, Fibrosis, Hernia, Edema, Cardiomegaly, Emphysema, Effusion and Pneumothorax. The system simply takes a medical image and then determines if the chosen condition is present or not. minoHealth also uses Cloud Computing to store patient health data and make it easily accessible to physicians in a scalable fashion. It also uses Data Science to analyse collected data in order to generate visualized important health statistics that can inform decisions and policy making. Collected data and visualized health statistics are aggregated from various health facilities and presented in a central portal as either regional or national health statistics which is collected and generated in real time.

And our research lab, minoHealth AI Labs also researches applying Artificial Intelligence to fields like Regenerative Medicine/Tissue Engineering, Biotechnology, Optometry, Nutrition/Dietetics, Epidemiology and Agriculture/Agricultural Economics.

With minoHealth, many regions in places like Africa without qualified radiologists for thoracic diseases, breast cancer and other conditions can still enjoy expert and above expert level forecasts, diagnoses and prognoses at an affordable cost and outbreaks would become easier to detect, which would then save numerous lives. 

We are using Deep Learning/Neural Networks algorithms like Multilayer Perceptrons and Convolutional Neural Networks for these Forecasting, Diagnostics and Prognostics system. We train these AI systems with various architectures that we fine tune during research and experimentation. The datasets used in training our AI systems are a combination of open sourced, closed sourced and data collected by our reseach collaborators and us. 

The entirety of our solution and system is based on how innovative technologies like Artificial Intelligence, Data Science and Cloud Computing can simplify and democratize healthcare. We believe such innovative technologies can complement physicians to become more efficient and for places without qualified radiologists, they can provide diagnoses and prognoses for the populace. 

Our system uses trained Convolutional Neural Network models that take medical images as input, be they mammograms or Chest X-Ray. It then propagates it through its layers and outputs the class, ie. whether it's a positive or negative case, Benign, Benign Without Callback or Malignant case.

We are currently in the field test phase. In the next 12 months, we'd complete testing and start deployment of our system in major areas in Ghana. We'd also train additional AI systems to cover some more conditions we're collecting data on. In the latter part of the next 12 months, we'd also start deploying our solutions in other African countries especially within the West African region. 

We'd train AI systems for genomics. We'd reach milestones in our research of forecasting and detecting outbreaks in Africa with AI and Data Science. Our system would be used in most African countries and some other foreign countries. We'd develop AI systems that can diagnose certain conditions with ubiquitous technologies like smartphones so that areas that don't have specialized medical technologies can still enjoy quality healthcare. Our systems would use data from patients' daily lives, collected by smartphones coupled with their medical records to make accurate forecasts about patients' health as well as further inform physicians in delivering personalized healthcare. 

We target health bodies and government agencies as our top down approach and target individual private hospitals. By working with these various bodies and agencies, we get all the healthcare facilities underneath them as customers and we would then give various portal access to them and give the main bodies a central portal. Since our systems are based on cloud computing, each healthcare facility's staff can access their facility's portal and use them to cater to the communities they interact with. 

We are currently in the field test phase, we are working with the National Catholic Health Service(NCHS) in Ghana to further test out these AI systems beyond the test sets used during modelling. The NCHS also shows interest in the deployments of our solutions after trials. 

In the next 12 months, we'd be serving an average of 3 million people. NCHS and its parent body, Christian Health Association of Ghana(CHAG) serve over 5.7 million outpatients annually within their 302 healthcare facilities. Our systems would be used to digitize their health records which would then be analysed by our Data Science systems to generate health statistics. And our AI systems would be used to diagnose and prognose some of their conditions. 

In the next 3 years, we'd be serving an average of 150 million people as we venture into other African countries and partner with their governments.

We are innovative, we are always challenging ourselves to come with ways we improve healthcare and even improve our solutions. Our different backgrounds from Industry and Academia, from fields like Software Development to Biotechnology, Optometry and Epidemiology gives varied unique views that all comes together to create something spectacular. Our familiarity with the Healthcare sector and African Healthcare helps us better determine what actions to take. 

We are also good at attracting talents and developing healthy collaborations with Academia and Industry.

We'd charge each health facility an average of $300 per month to use our full system. Which would translate to over $1 million annually with just 300 facilities and over $3.6 million dollars annually with 1000 facilities. As we add more AI systems and expand, we'd charge larger hospitals more. There are several healthcare facilities to work with, Nigeria alone has 23,640 public and private hospitals, working with just half of them would generate about $40 million annually. Our server cost and maintenance cost would only be a fraction of that allowing us to continuously expand with our profits.  

With grant funding from MIT Solve, we can drastically speed up deployment and acquire additional hardware to train a lot more AI systems at a faster pace. With the relationships and networking that we would derive from MIT Solve, it'd be easier to establish stronger relationships with the various African governments and health bodies.

Partnerships with African governments and health bodies. We have established relationships and are partnering with some health bodies and government agencies in Ghana. However, MIT Solve's global influence can be used to get us access to other African governments and health bodies around the globe. 

Early cost of Hardware. MIT Solve's can help us speed up training of AI systems and deployment with grant funding.