From disease diagnosis to the creation of new drugs, artificial intelligence (AI) is transforming healthcare. Our expert team is pioneering the application of AI for improving health outcomes for those in the earliest stages of life.
What is artificial intelligence?

Artificial intelligence (or, AI for short), is the demonstration of qualities of human (natural) intelligence in machines. Today, computers have become extremely efficient at analyzing large quantities of data to learn how to carry out specific tasks. Machine learning (ML) is the subbranch of AI that centers around the development of these algorithms that learn and adapt with experience.

It can help to think of machine learning in (metaphorical) terms of the cognitive development of a newborn baby. While it can't be said that machine learning models learn in the same way the human brain does (machine learning models are constructed with mathematics, not biology), these models do try to imitate behaviors of human intelligence as much as possible.

For instance, parents of a newborn will teach their baby facts like its own name. This is called supervised learning, because the baby learns words by listening to you repeat them over and over again. Machine learning models try to mimic supervised learning behaviors by learning from training datasets.

When your baby discovers new things without your input it is called unsupervised learning. For example, a child creating their own artistic drawings is an act of discovery. When left with a set of paints, many children will learn on their own how to express their thoughts and feelings through art.

But what happens if your child decides to paint a pretty picture all over the wall of your home? You will tell them this is not ok and that they should not do that again! In machine learning this process is called reinforcement learning.

The ability of machines to learn and adapt is a central concept in artificial intelligence, much like it is for a child's cognitive development. Machines can crunch numbers so fast that given the right data and algorithm, this learning process can take place in a matter of seconds or minutes. Since healthcare involves making medical decisions based on a patient's health information (test results, genetics, medical history, past experiences with similar patients, etc.), there are many areas where AI can be used to improve the delivery of healthcare.
Applications of artificial intelligence
Artificial intelligence is useful in a wide range of healthcare applications. This is reflected in the many diverse ways Velmio uses machine learning
Simplifying lifestyle tracking

Lifestyle tracking is helpful for better understanding your health, but can be time consuming if you need to manually record everything. Velmio simplifies the lifestyle tracking process with automation and clever AI features. For example, Velmio's AI image recognition model identifies a meal from a photograph and automatically records it nutrition content.

Furthermore, our team has built a sophisticated machine learning system that helps pregnant women understand what foods are safe during pregnancy. This system automatically analyzes the ingredients and macro-/micro-nutrient composition of meals to provide information on the benefits and potential risks of consuming certain foods during pregnancy.

Dig deeper and you'll find many other ways in which Velmio applies machine learning to make lifestyle tracking easier. Velmio's sleep tracking feature for Apple Watch, for instance, uses machine learning to automatically calculate the amount of time spent in each stage of sleep (light sleep, deep sleep, REM sleep and interrupted sleep). This helps Velmio users understand how to improve their sleep quality during pregnancy.
Understanding pregnancy symptoms

Velma is a powerful chatbot that helps pregnant women better understand symptoms they experience during pregnancy.

Under the hood, Velma is a complex piece of technology that leverages the latest advancements in statistical modelling and machine learning to understand patterns between symptoms, risk factors, lifestyle data (from your connected smartwatch devices, nutrition logs, etc.) and conditions that can arise during pregnancy.

Velma was created to help address one question we kept hearing over and over again: "Is this symptom normal?" As researchers we were frustrated that the existing medical research didn't fully address this question, leaving pregnant women confused and anxious. Velma was designed to fix this problem. Digital health tools provide a platform for collecting research data from a large and diverse cohort of pregnant women. With this data we can definitively understand what symptoms are "normal" and provide further advice and support for pregnant women through digital channels.
Personalized health insights

Every pregnancy is different and this is reflected in the unique genetic makeup of every human being. Velmio creates personalized health insights by analyzing your individual needs for a healthy pregnancy.

Other pregnancy health apps show women a series of generic articles at each stage of pregnancy, failing to address many questions that pregnant women have. Velmio instead identifies health information relevant for you, backed by evidence-based clinical guidelines. This is achieved through Velmio's unique Health Fabric technology, which analyzes patterns in your health and lifestyle. This system is extremely flexible and adapts to your needs and interests. For example, some women connect their smartwatch device to the Velmio app for fitness-related insights, whereas others like to use the app to assist with gestational diabetes management. The Velmio app also generates detailed reports that you can share with your doctors to help inform their clinical decision making.
Personalization in healthcare

Sophie Wharrie, Velmio's co-founder and CTO, is an applied mathematician and computer scientist conducting research on the development of new machine learning methods for personalized healthcare. Applications of this work includes the development of personalized drugs to improve health outcomes for cancer patients. Machine learning is used to select the best treatment based on the patient's unique genetic profile. This in turn optimizes the likelihood of the cancer treatment being successful.

Velmio applies a similar approach in a pregnancy context, by analyzing patterns in your health and lifestyle. Sophie says that the emerging research field of epigenetics is particularly exciting within the context of data collected by Velmio, to better understand how lifestyle choices and environmental factors influence the development of the fetus. Velmio's technology can help us better quantify risks (e.g. exactly how much alcohol is harmful during pregnancy?) and shift away from a one-size-fits-all approach (e.g. what is the optimal diet or exercise regime for you as an individual?).

In the future, everyone may have a digital health record that includes your individual DNA sequence. Technology such as machine learning may be used to identify segments of your DNA that indicate properties of your health, such as diseases you or your children may be predisposed to. Artificial intelligence can then help create personalized medicines to prevent and treat these diseases. There are already research labs around the world working on aspects of these technologies and Velmio's expert team are bringing their technical know-how in this field to applications in pregnancy healthcare.