Artificial Intelligence, or AI, is currently transforming nearly every single industry we know: From healthcare, to energy, to even the everyday technology that we use, like our YouTube recommendations. But the AI revolution is doing something truly spectacular for healthcare, and these computational leaps could be just what Bangladesh needs to bridge the inequalities here.
What is AI?
In simple terms, AI is just using data and a computer algorithm to make predictions about something. There are various ways in which to do this, and the field of researching these AI algorithms is currently highly active. These algorithms, after some fine-tuning of their parameters, have the ability to spew out predictions much faster -- and often even more accurately -- than humans.
How can it be applied to healthcare?
The reason that healthcare is such a lucrative field for AI is the data aspect. Hospitals globally generate about 50 petabytes of data per year, mostly from medical imaging. To put this into context, 50 petabytes of data is like the entire written works of mankind, from the beginning of recorded history, in all languages. And hospitals generate this much data every single year; but the best part is that 97% of this data is unused.
Machine learning and AI presents an opportunity to take advantage of this huge amount of data. And big-tech is already jumping on this. In 2019, pharmaceutical company Novartis and Microsoft announced a collaboration to “transform medicine with artificial intelligence.” Facebook partnered up with New York University (NYU) to use AI to do faster MRIs. Amazon has developed AI that automatically generates medical records from patient-doctor conversations. Google DeepMind created AlphaFold, which can accurately predict the shape of proteins, helping in medicine discovery.
In 2019 alone, over $4 billion was invested in AI healthcare startups, and that number seems to only be increasing. According to CB Insights, there are 90+ healthcare AI startups to watch currently, ranging from companies that work with imaging and diagnostics, drug discovery, genomics, fitness, virtual assistants, clinical trials, nutrition, compliance, mental health, remote monitoring, hospital decision support, predictive analytics and risk scoring, and so much more.
Stakeholders in AI and healthcare
Ultimately, healthcare is centred around improving the quality of life for patients, but there are many stakeholders who make this possible. The main “trio” of stakeholders in the healthcare space include patients, doctors and medical professionals, and administrators. AI and healthcare can impact every single stakeholder. For patients, AI can help with discovering new, better drugs that save their lives, as wearable devices that monitor health, through genomics, or as a therapist through mental health chatbots (such as Stanford’s WoeBot app).
But it isn’t just for patients -- it can help doctors and administrators too, and massively so. For doctors, it can help with improved diagnosis, medical imaging, identifying risk, robotics surgery, and telemedicine. For administrators, it can help with health insurance, electronic health records, scheduling procedures, predicting equipment orders (such as PPE, blood units, etc), and much more.
AI’s role in bringing down healthcare costs
Healthcare costs in the US have reached $4 trillion per year. Health insurance accounts for 3-10% of that spending, or $120 to $400bn per year. Health insurance companies don’t have good safeguards against fraud, so the cost is passed down onto its customers. AI can help remove much of this cost by fighting fraud; it can identify charges for services that didn’t happen, misrepresented dates/locations, false diagnoses, and unnecessary prescription drugs.
The ethics of AI in healthcare
Hearing about AI’s many possible applications in healthcare is as exciting an idea as it is daunting. This could pose a threat to replacing medical professionals’ jobs. And what happens if the AI makes a wrong diagnosis -- who do we blame? What about data privacy of patients?
There are still many things to consider before the wide implementation of these technologies; many laws, policies, and frameworks to pass, still. But one thing is quite clear -- this technology could be game-changing for patients, doctors, and administrators. Perhaps it should not be viewed as a replacement to doctors, but rather as a source of help to patients, clinicians, scientists, administrators, and everyone else working in the healthcare industry.
In a country like Bangladesh, where the doctor-patient ratio is 5.26 per 10,000 people (the second worst in South Asia) according to WHO and healthcare inequity is massive, perhaps it is time that Bangladesh’s healthcare administrators and doctors begin looking at the potential of AI -- not as a replacement, but as an assistant. An extra helping hand from computers may just be what we need to make healthcare accessible to all.
Farihah Ahmed is a freelance contributor.
