If you step back and really look at it, healthcare has come a long way. The idea of washing your hands to stay disease-free used to be a foreign concept. Now it’s so important that we’ve added a whole feature to the bathrooms to address just that. Let’s take a look at 2020 and see how Covid has caused enormous damage – but it has also accelerated the development of much-needed technological innovations such as virtual consultations and rapid medical care. Now we are experiencing a new revolution: AI healthcare. It’s real, it’s there and it changes everything.
At the Africa Health Congress 2024, AI healthcare was a key focus of the conference, with discussions including its impact on accessibility, cybersecurity and benefits. The conversation is crucial because the WHO has said the world will lack 10 million healthcare workers by 2030. Additionally, South Africa is a country with limited access to quality healthcare. A 2022 study (approved by the President) shows that a staggering 84% of South Africans are without medical care and rely solely on public healthcare. But mobile technology is changing healthcare. This means AI can help close this gap by making life-saving technology accessible to more people. Here you can find out what happens next for us – and what considerations health politicians need to take into account.
How AI in healthcare can make things better
AI can speed up examinations and diagnoses
AI diagnosis
Numerous companies have emerged to address the problem of how long it can take for healthcare workers to diagnose illnesses – and for patients to seek help. Take tuberculosis for example. According to the World Health Organization (WHO), it is the 13th leading cause of death worldwide and is responsible for an estimated 1.6 million deaths each year. Still, studies indicate that patients typically wait between 30 and 180 days before seeking medical attention for tuberculosis.
A new AI-powered diagnostic tool, AI Diagnostics, is a digital stethoscope that detects lung sounds to find TB-positive cases. According to AI Diagnostics, this could save money and time on conducting laboratory tests on TB-negative cases. They say it can also speed up care for TB-positive patients. It has also been shown to identify up to 40% of patients who are currently completely missed by the predominantly symptom-based tuberculosis screening method.
AI and mental illness detection
AI machine learning models have been shown to be able to detect fear. How? By only analyzing facial expressions in calm and fear scenarios. According to a study of the technology in Cureus, “AI algorithms are having a significant impact on detecting and effectively removing biomarkers associated with anxiety and depression.” Additionally, personalizing treatments using AI could improve the treatment of these conditions .
AI in healthcare can automate tasks
At the Africa Health Congress 2024, Solventum presented its AI management system, a kind of Alexa on steroids for doctors and healthcare workers. It records patient sessions, takes detailed notes, adds task notes and task prompts, and even manages the billing process. In an environment where doctors are overburdened with paperwork and administrative burdens, this AI solution aims to simplify and streamline the entire process. Ultimately, this improves the quality of care for patients.
Data can (finally) be centralized
Have you ever thought about how strange it is that your doctor has no idea what your psychiatrist knows about you? Or when you switch to a new doctor and need to pass on all your health information – every allergy and surgery – to the new doctor? Or that an emergency room doctor doesn’t know about your chronic medications unless you tell them about it – but what if you’re unconscious or can’t speak?
Patient records are not centralized and stored in a location accessible to healthcare professionals. Countries with centralized health data using AI are producing better health outcomes for doctors and professionals. And it can improve adherence to medications and treatments. This technology would be used to store all patient health data from all areas including mental, dental, surgical and vital health.
The specific performance with AI? It can interpret medical images and data, track illnesses and determine claims for medical care, according to a document from the Policy Action Network.
AI health means improved treatment
Since artificial intelligence can speed up medical problems, this could lead to faster treatment response times. AI-powered solutions are already being introduced. It gives healthcare workers insights into patient treatment, uncovers hidden indicators, and even predicts genetic mutations and provides prognosis.
Then there are AI doctors and surgeons. AGI4Health, the “world’s first true generalist AI doctor,” is being developed by minoHealth, an AI laboratory for the medical industry. Artificial intelligence is also being explored as a way to aid operations by assisting surgeons and improving surgical decision-making.
It could manage health
AI is already being used to monitor embryos in fertility clinics. Since experts value its potential as a holistic health management system, it is also suitable for helping with chronic diseases. With machine learning (and potentially wearables), AI could use data to personalize treatment recommendations, consider lifestyle and socioeconomic data, and monitor overall health while offering care options.
…but AI healthcare needs to be rethought
Before we skip off into the sunset with our freshly pimped smartwatch that tells us exactly when to eat next to optimize our health, there are serious considerations.
AI is expensive
On the one hand, who should pay for these expensive services? A good system first requires internet access. However, Wi-Fi and data coverage remain a major issue in South Africa. And they are still far more expensive than in other countries. When it comes to implementing AI technology in healthcare, a project to develop appropriate systems can cost millions of dollars. Building the processing units is also energy intensive and South Africa is still struggling with electricity. Finally, generative AI in South Africa and other multilingual countries would require advanced language skills – something that costs more.
Data may not be secure
As technology advances, hackers also evolve. They can obtain sensitive data and either reveal confidential health information or demand ransom from people or institutions. Some of the “largest and most damaging data breaches in recent history” occurred in 2024 alone. SA’s own National Health Laboratory Services was hacked and it took months to secure and reactivate the system. Data breaches can cause more harm than slowing down laboratory services. They can also expose sensitive health information and potentially endanger lives.
How ethical is AI?
Of course, there are also ethical considerations: If someone is treated incorrectly, who is to blame, the robot or the doctor? Can we trust AI to do the work we ask of it? In addition, AI laws, like internet regulations, are not implemented very quickly – precisely because there is so much to consider. However, slow implementation and rapid adoption open the door to uncertain legal foundations.
AI has a diversity problem
A now-famous study from 2022 discovered just how detrimental AI can be. In it, the researchers found that the intelligence community was more likely to associate black men with crime and women with being a housewife. In 2023, a black American woman sued the city of Detroit. Facial recognition technology incorrectly identified her as a criminal, leading to her arrest. In healthcare, such discrimination could have an impact on lives. Experts widely agree that local expertise and cultural nuances are critical when implementing this technology.
That doesn’t mean we can’t use AI to improve our health. We already use chatbots to quickly scan our symptoms like coughs and pain and figure out what might be bothering us. It is used to automate tasks and organize large data sets. And every time you shop online, AI personalizes your experience. Over time, we could see these uses skyrocket, problems being solved, and profits being made. Because who doesn’t want to skip the line at the doctor’s office?
