In a recent study, the healthcare industry is estimated to be spending close to 36 billion US Dollars in Artificial Intelligence and Machine Learning by 2025. AI is redesigning the healthcare sector and helps to make diagnoses, monitoring, and drug discovery faster and more precise. Especially in these challenging times, AI can help us take effective measures and make data-based reasonable decisions on how to react.
How does AI work? And how can this technology help us get through a global health crisis?
In the second episode of our interactive web show Gymtalk, we invited Artificial Intelligence (AI) expert Christian Ehl to talk about applications of machine learning to a global pandemic. Christian is the founder and managing director of the digital agency HILLERT NEXT, and founder and investor in many startups. His major focus is on the use of AI and its impact on people, companies, and our planet.
Let’s quickly dive into the talk: The human brain collects information and creates a model on how to behave and react in certain situations. It is constantly learning and takes decisions based on those lessons learned. Like a brain, AI builds a model based on collected data. With a pandemic, this model shows how the virus spreads, what it takes to slow down this process, and increases the chances to find effective measures and a cure faster.
Within the last months, many people were watching the news and changed their behavior accordingly: They now wash their hands more frequently and thoroughly, keep a proper distance, and wear a mask in public places. Data points like how many people wear masks can be collected and fed into an AI machine. Then we can try to understand the data and the causing effect of it: Does the fact that more people wear masks lead to a decrease in spreading the virus? Based on those results, AI can make predictions on what will happen if we all wear masks. We can create scenarios based on the predictions: Is it possible to get all the masks that are needed? How do we make people wear them?
For humans, it is hard to understand and process all this data since it is too complex. AI has unlimited capacity to see all the details that matter in the data sets. It can make predictions based on data points which help us decrease uncertainty. This makes things plannable and enables us to take effective measures. People are part of creating that data, for example, by using the Corona App. They feed data into a neural network and train the AI machine. The machine then takes a decision based on the learnings.
Scientists are working on understanding this data.
Which data points can be used to detect and fight the virus? One way to detect the virus with AI is to collect and publish chest radiography images and to test and train the AI with images of healthy and unhealthy lungs. The AI can then recognize if a person has the virus and make a prediction on how severe the course of the disease will be.
AI can also help find a cure for the virus: Doctors are already publicly sharing data online like the genome of the virus. This and information about the structure of the virus helps to build a model on how the medicine or vaccine could look like. With this data, we can learn about mutations and create models on how we can respond to those and roll out effective protection matters.
If we create more data, models can be built, and we can continue to improve our knowledge.
If we have more data sets with anonymized people’s data like age, cases, and areas, we can match that data to the responses and measures and find out what effect it has on the curve. Like this, we can build better models and find the right measures to react to the crisis.
Watch the entire Gymtalk episode here and find out more about data privacy, biases in AI, and what you can do.
COVID-19 is only speeding up the application of AI in the healthcare sector.
AI can digitize and break down massive data sets to analyze disease vectors and identify the effects of treatment. Many companies already apply AI in fields like medical records and data management, digital consultation, disease detection, patient care, robot delivery, and drug design.