AI predicts how cancer evolve


Tumors have an inconsistent nature that is constantly changing and altering its course, and this is primarily why treating cancer is such a huge challenge for doctors and researchers. However, if medical science could manage to predict the course of evolution of a particular cancer tumor, doctors can easily change the course of the treatment before the tumor gets the opportunity to adapt to the treatment and become resistant, which will also boost the survival rates of patients suffering from cancer.

A team of researchers at The Institute of Cancer Research (ICR) and University of Edinburgh undertook a research with the aim of understanding the evolution of cancer tumors. The scientists employed the use of artificial intelligence to make predictions about the progression and evolution of cancers, and this can allow doctors to create a personalized cancer treatment from the earliest stage, which was simply not possible earlier.

AI predict cancer tumour evolution

The scientists have introduced a new technique termed as the REVOLVER, which basically stands from the repeated evolution of cancer. This technique identifies the patterns in DNA mutation within the cancers, and then it utilizes this information to predict genetic changes that will occur in the future.

The researchers have invented a mighty artificial intelligence technique that is capable of forecasting the future progression of the evolution of cancer tumors on the basis of several patterns of DNA mutation, which were previously shrouded inside complicated sets of data. This artificial intelligence tool will empower doctors to eliminate the biggest challenge in treating cancer: its ability to generate an unpredictable course of evolution that leaves experts in the dark about what might happen to the patient next.

The researchers have also succeeded in establishing associations between certain patterns of repetitive tumor mutations and the rates of patient survival. For instance, breast tumors that had a pattern of errors within the genetic materials that also codes for P53, the protein that suppresses cancerous tumors, was followed by the occurrence of mutations in chromosome 8 had a lesser survival time as compared to the tumors that had similar patterns of genetic changes. This indicates that repetitive sequins and patterns of DNA mutations can be considered as an indication of the prognosis, which will facilitate doctors in designing beneficial treatments in the future.

Discovering Patterns

The team of researchers has also introduced a machine learning technique that is capable of transmitting information regarding tumors amongst similar patients. This technique will allow the easy identification of sequences and patterns within the order in which genetic mutations tend to occur, amongst both patterns that are repeated within and between the tumors of the patient. This will allow doctors to apply or compare on tumor pattern of mutations to predict the course of evolution for another tumor.

Artificial Intelligence Cancer Prediction
Library RF

The researchers analyzed 768 samples of datasets and information retrieved from 178 patients suffering from colorectal, lung, renal and breast cancer. Each type of cancer was analyzed to identify and compare the genomic changes that occurred in every single tumor. The scientists identified the repetitive patterns, and then, this information was amalgamated with existing knowledge of cancer biological growth and evolution to assist the researchers in predicting the future course of tumor growth, development and evolution.

If tumors with similar patterns are discovered to be developing a resistance to any given cancer treatment, this innovative technique will prove highly useful in predicting whether or not the patients are likely to develop resistance to the treatment in the future.

This ground-breaking study has revealed how scientific research can gain value and insight through collaborations between scientists and researchers of different fields and disciplines. By providing the solution to a statistical machine learning problem, the researchers were capable of providing extremely valuable insight on the evolution of cancerous tumors.

This is a remarkable example of how the infinite scope and valuable power of artificial intelligence is capable of identified complex and complicated data patterns to further scientific understanding and bring about marked improvements in the field of medical science and human healthcare.


Please enter your comment!
Please enter your name here