AI Researchers Design Recipe Book With Anti-Cancer Recipes

Recently, a book of recipes called Hyperfoods was released, comprised of recipes generated through the assistance of AI and machine learning algorithms. The recipes in the book are based on foods with anti-cancer properties.

Artificial intelligence is seeing increasing use in the creation of food recipes. For example, companies like Analytical Flavor Systems have been using AI to analyze the flavor and textures of drinks and attempt to design drinks catering to specific locales. Meanwhile, Plant Jammer is an app that leverages artificial intelligence to suggest recipes based on the ingredients you have in your house.

As reported by the Imperial College London news, a researcher at Imperial College London and a chef have teamed up together to employ artificial intelligence to identify foods with anti-cancer properties and then compile these foods into a recipe collection. Dr. Kirill Veselkov is a researcher from Imperial College London’s Department of Surgery and Cancer. Dr. Veselkov and his team of researchers analyzed the molecular data of over 8000 food items. A large number of the molecules analyzed by the AI were flavonoids, responsible for giving vegetables and fruits their color. The research made use of the distributed computing application DreamLab, enabling members of the public to participate in the research and help identify around 110 anti-cancer molecules.

According to Veselkov, conditions like cancer, heart disorder and neurological diseases are tied to poor diet. Some studies suggest that poor diets could contribute to around a fifth of all deaths around the globe every year. Veskelov states that around half of cancer cases could potentially be averted by better lifestyle choices and better diets.

Veselkov worked alongside chef Jozef Youssef to create a recipe book made out of food anti-cancer foods. Youssef believes that while we are still a long way from personalized diets, the research conducted through DreamLab is still a critical step in advancing our understanding of dietary health and assisting people in changing their diet to something healthier. Youssef explained that the recipes in the book were designed to instruct people in methods of using ingredients to create meals that could help avert cancer and other forms of disease.

Michael Bronstein, another Imperial College London researcher, worked alongside Veseklov to conduct the research. Bronstein stated that the Hyperfoods project is the first known attempt to use neural networks to examine the effects of food molecule’s on people’s health. As Bronstein was quoted by ICL news:

“By modelling the ‘network effects’ of the interactions between food-based molecules and biomolecules in our body, we can identify which foods contain ingredients that might work in a similar way to medical drugs and have the potential to prevent or beat diseases. Our ambition is to provide a quantum leap in how our food is ‘prescribed’, designed and prepared – and this is a great first step.”

Although the researchers are excited by the results of their studies and have released the cookbook to the public, they also caution that the book should not be used in place of medical advice from a medical professional. The researchers were careful to state that the relationship between health and food molecules still needs more research. There is some evidence that the right diet and exercise can prevent certain types of cancer. However, effect sizes are often modest and it’s unclear if other food types can elicit similar effects.

The DreamLab app will continue supporting the research of Imperial College London, helping researchers there investigate the possibilities for combinations of food and drugs to treat cases of COVID-19, and any findings from the research project will be made available to the medical community to undergo clinical trials if promising.