The AI will see you now

The AI will see you now

By Phil Massie

The global healthcare industry sits atop vast swaths of untapped data opportunities. Sadly, as is the case in so many other industries, these data are securely siloed away from each other. One of the reasons for this segregation is the need for confidentiality in patient records. Another is the array of different data sources, for example research and development, physician or caregiver records and patient collected data. 

Despite the familiar challenges, artificial intelligence (AI) and data science initiatives within healthcare are on the rise. Teams are looking at disease identification and diagnosis, personalised medicines, drug discovery and manufacturing, clinical trials, radiology and radiotherapy result analytics, disease outbreak prediction, smart health records and processing data from wearable electronics to name just a few. Clearly there is a lot of interest in and a need for big data and data science across the field. In fact, in response to the increasing number of data driven healthcare solutions and the growing amount of data available, Yale University (very) recently launched the Center for Biomedical Data Science which aims to “… enhance research at the medical school and at Yale as a whole in the broad area of biomedical data science from complex machine learning models, to simulations of molecular, cellular, and organismic systems.”

The breadth of scope of these types of projects and initiatives like the CBDS at Yale are obviously very encouraging. Thankfully, healthcare and big data are being successfully juxtaposed in the developing world too. For instance, since 2010 the Indian government has been in the process of issuing Aadhaar cards. The cards assign unique numbers along with biometric identification to every Indian citizen. The hope is that the system will become the backbone of a broad public health data generation and monitoring system. It will link to electronic medical records and health insurance information and is expected to translate into faster, more targeted public health interventions.

The United Nations Global Pulse is an impressive project looking to take advantage of big data for public good. It lists numerous humanitarian projects in developing countries focussed on everything from gender equality to Fintech solutions as well as public health programs. For instance their Pulse lab in Kampala has developed a solution using big data analytics to compile and analyse data collected centrally from health centres across Uganda. The system incorporates risk factors such as rainfall and population density and mobility, producing visualisations that map the spread of infections. The solution utilises open software systems used by numerous countries in the region and should scale well to other nations. The goal is to allow policy makers to track infection outbreaks in different areas, in near real-time, leading to much faster responses to disease. The tool is still under development but looks promising.

Another Ugandan Pulse Lab initiative monitors the rollout of their Option B+ project, an initiative aimed at preventing the transmission of HIV from mother to child. Using data from the country’s network of healthcare centres, the system visualises data such as the number of patients going for regular antenatal care, the number of HIV/AIDS cases and how many patients are receiving Option B+ treatment.  By looking at correlations between medical supplies and relevant factors, the project hopes to address bottlenecks in Option B+ rollout in Uganda.

There are several NGO backed big data and data science driven healthcare initiatives underway in the developing world. While the nature of these projects may be subtly different, they appear to reflect regional requirements. As the developed world delves deeper into drug discovery and genomics, the lessons learned will no doubt filter into and enhance data driven projects in the developing world also. Now is clearly the time for the governments of developing countries to invest heavily in data science education. Missing this opportunity now will mean endless missed opportunities later.

Share This Artcle :