Medical diagnosis in the case of cancer requires a careful approach to patient data. It is important to consider, for example, risk factors, tumor characteristics, or tumor phenotype. This information determines the diagnosis and the most effective treatment in each individual case.

Researchers at the Mayo Clinic decided to determine whether this data could still be used in some way. Cancer is known to differ in its genome and phenotype, but research in this area, as well as diagnosis and treatment, is based on average risk factors.

Data from 90 patients with breast cancer were studied to indicate the future use of this information. The presence of mutations triggered by the disease was also noted in all of the subjects. After reviewing the information, the researchers concluded that about half of the patients’ biological data was missing, which represents a huge potential to study the disease on a case-by-case basis using the most accurate treatment.

Using a statistical method, the scientists projected the impact of random factors on the onset and course of the disease. The generated gene map was then matched to each patient’s disease. In this way, the researchers were able to determine the role of genes in the course of the disease. The same gene can affect patients in completely different ways. Moreover, this individual genetic map makes it possible to study the combination of drugs. Patients with identical diagnoses differed in their unique genetic sets and, as a result, in their treatment methods. This approach makes it possible to use individualized medicine successfully in further, larger-scale, studies.

In the case of cancer, the study of genes or proteins alone is not sufficient to determine its characteristics. The synergistic functioning of the different body systems requires a comprehensive analysis in order to make more use of multi-domain data. It is artificial intelligence that satisfies the need to analyze such large-scale data with predictive capabilities.

The analysis of genetic data in this study was carried out using a personalized mutation analyzer, which is the basis for a future artificial intelligence platform. The analyzer provides the input data that encodes the patient’s medical history, thereby creating robust artificial intelligence algorithms.

Future work on the project includes refining the analyzer and testing its effectiveness in other diseases.