Artificial intelligence (AI) has already had a significant impact on every industry sector where it has been adopted.

By Henry Adams, country manager of InterSystems

It is especially in healthcare where the technology has become critically important in not only improving outcomes but recognising patterns of patient behaviour to pro-actively help save lives. AI is capable of enabling applications to sense, think and act. It can see and hear and speak using natural language processing and machine vision. And with deep learning and neural networks it’s capable of a level of understanding and action.

AI is unlikely to replace the human component in most workflows any time soon, but it can provide a second, safe pair of hands that can support physicians when workloads are heavy and times are stressful.

This is particularly true of AI in the radiology sector where some radiologists are increased expected to interpret a high number of images a day, with one study suggesting that rate to be an image every three to four seconds in an eight-hour work day. These volumes are unsustainable and can potentially impact on patient care which is where AI steps in …

Healthcare providers must use the data at their disposal in smarter ways. AI provides the means to do so in a real-time manner critical for the health sector. Fundamentally, this technology is disrupting established processes and systems by injecting them with sophisticated, data-driven insights to deliver more value.

Even though local providers have shown interest in leveraging AI, its adoption is still very much at an early stage in South Africa. Certainly, there is an awareness of how it can be used to facilitate patient care.

However, other priorities in the sector, such as giving more South Africans access to quality and affordable healthcare, have pushed AI down the agenda. But by readily embracing the opportunities AI can provide, healthcare providers can more effectively deliver on their mandate and reach a larger pool of citizens with the critical care needed especially during times of crisis.

The lockdown South Africa and many countries around the world are experiencing, has forced industries to reassess traditional approaches and become more digitally-led. In part, this can be attributed to the need for self-isolation. In healthcare it is clear making sense of the vast amounts of patient data and unstructured information will help identify trends and maybe inform society how to ‘flatten the curve’ of infection and limit the spread of the coronavirus.

For example, AI can assist in predicting the number of potential new cases by areas, which population segments are most at risk, and identify the communities where the contagion is likely to move to next. While the technology will certainly prove its value during the current pandemic, it has significant potential beyond it.

African examples

In Nigeria, start-up Ubenwa is using AI embedded in a mobile app to analyse a baby’s cry to detect early sign of anomalies such as asphyxia or brain injury which could be fatal. Its automated cry analysis can also extract indicators of basic needs such as hunger and sleep.

Meanwhile in Ghana, the minoHealth system uses deep learning and AI to predict and diagnose medical conditions in patients. This includes detecting breast cancer, pneumonia, hernia, and fibrosis from just x-ray images. Using data analytics, it can collect and analyse health data from facilities to better population groups and start tracking and forecasting outbreaks.

AI can also be used to assist in the prediction of the amount of waste a hospital is likely to produce. This will be helpful in the storage, transportation, and disposal of hospital waste management, which is a notoriously expensive undertaking at any healthcare facility.

Closer to home, Dr Raymond Campbell developed a mobile health clinic solution focused on underserviced communities. Built in the form of a backpack that can be carried by a single health worker, it can screen patients at their homes, submit the data to the cloud for AI analysis, and have results available within approximately 20 minutes.

Optimisation

AI can also assist in improving the management of under-resourced facilities by identifying the most efficient and effective ways to use the available medical professionals. This improves the patient flow while also ensuring help is available to everyone who needs it.

All told, AI will be an essential part of the modern healthcare environment in South Africa and other countries beyond the lockdown. Using the technology will significantly optimise processes and systems and highlight more innovative ways quality healthcare can be provided to those who need it most.

Already, the application of analytics has evolved thanks to the massive amount of data available and the increasing number of specialised algorithms used to process it. Today it can help improve efficiencies and the overall impact of projects across the healthcare sector. It speeds the digital transformation of healthcare from the reactive care for the masses to preventive personalised care.

However, AI will remain a key driver only if society also develops and upholds the governance required for its ethical and transparent application. Avoiding bias and establishing trust are now the new hurdles AI must navigate, no more so than in healthcare.