The specialized magazine Big Data Magazine publishes an article by our business development manager Andrés Visús on AI in healthcare.
There is a lot of talk about Artificial Intelligence and its applications in medicine, and it is usually associated with, and the cases in which it is applied to the detection of cancers by means of diagnostic imaging are the loudest. In this field, fortunately, much progress is being made and also in many other areas of medicine and health that are not so well known and that have a great impact on society.
To address the potential of AI in healthcare with a broader perspective, it is necessary to understand, albeit in a very simple way, how Machine Learning works.
Machine Learning is the most important activity of AI, and it focuses on interpreting data to make decisions. Basically, Machine Learning models look for correlations and patterns in the data, and learn from them. With this information they are able to segment, detect anomalies in the patterns, and predict future results.
In short, they learn from a history, and when faced with new information they are able to interpret it based on the patterns they have identified in the data they have learned from.
For example, in the detection of cancers by diagnostic imaging, the models interpret the images as data, and are able to identify patterns that correspond to other cases in which a tumor was identified. Being able to “see” beyond what a person can do, no matter how experienced he or she may be. Leading companies such as IBM and Google have been major drivers of these advances.
COVID-19 is accelerating all types of scientific research, and AI is playing a very relevant role. Millions of PCR tests are performed every day around the world, but it is critical to minimize false negatives, i.e., those cases in which the PCR result is mistakenly determined to be COVID-19 negative. Data Science companies such as PredictLand are developing Machine Learning models that have learned from PCR histories and are able to determine results with accuracies that traditional techniques do not achieve. Undoubtedly, having the conclusion of a specialist that is backed by AI solutions of this type, provides a rigor and confidence, which are vital and not trivial.
Beyond identifying a “positive/negative” cancer diagnosis or a PCR, Artificial Intelligence is helping to predict neural diseases by noninvasively monitoring a patient’s vital signs and behaviors. Massachusetts Institute of Technology (MIT) scientist and professor Dina Katabi has developed a wireless technology that detects and records a person’s vital signs and movement without the need to carry a device. With the information it collects, and using Machine Learning techniques, it detects anomalies in behavioral patterns and vital signs, being able to identify and predict certain physical and neurological diseases.
The graph shows the record of a person’s behavior over the course of a month and during 24 hours. Machine Learning models analyze this data, providing highly relevant information for specialists that they would otherwise not be able to determine with the same precision and anticipatory capacity.
In short, Artificial Intelligence is driving the value proposition of specialists, providing information of great importance for decision making. It identifies, diagnoses and predicts with a capacity that goes beyond human capability.
The demand for Artificial Intelligence solutions is growing by leaps and bounds, bringing new opportunities to companies and society, opening the door to new professional profiles related to Data Science and AI in general.
