Feasibility of training machine learning algorithms to predict clinical outcomes, identify diagnostic markers, and discover targets for treatment from multi-modal biomedical data.
The aims of our research are:
- To make it easier for doctors to predict whether a patient will have a full recovery from their illness;
- To be able to tell, in advance, whether a medicine will be helpful or might lead to harm without benefit;
- To make new medical tests to tell whether or not a person has a disease;
- To find new medicines, especially medicines that work in new ways.
The human mind is very powerful, and it is amazing that it is able to run on about as much energy as a light bulb. However, despite the incredible capacities of the human brain, there are some tasks that are difficult for the human mind, such as keeping track of large amounts of information. In contrast, as computers have become more powerful over time, they have gained the capacity to organize and store more and more information. Computer tasks that used to take hours or days can now be done in just seconds.
These new, powerful types of computers are now able to work with such large amounts of information that they can perform new types of tasks that were not possible until now. One of the new advanced types of calculations that can now be done with computers is the ability to create models of medical illnesses. By comparing new information to information that the computer has already analyzed, the computer is able to make predictions about whether or not a medicine or a test will be effective.
We plan to use the information from the UK BioBank is to make models of human health, and then check to see how accurate those models are. We need information, like the information that is stored in the UK BioBank, to be able to make our models. We believe that we might be able to find new tests and new medicines if we can manage to make a model that is accurate enough.
Public Health Impact
We believe that we can help to make more accurate tests, and more effective medicines, using our research. The impact of this work could be very important, and could even lead to being able to catch diseases early, before they spread. We especially hope to discover new, effective ways to treat disease.