How machine learning, AI can aid treatment
Machine learning and artificial intelligence (AI) can be deployed to identify patients likely to suffer from autoimmune diseases and thus deploy timely diagnosis, monitor disease prognosis and prevent autoimmune comorbidities.
Three elements contribute to autoimmune disease development: Genetic predisposition, environmental factors and an impaired immune system. Autoimmune disease prevalence has been reported in 15-29 per cent of the population.
Autoimmune conditions are risk factors for other infections with data indicating that patients suffering from rheumatoid arthritis and asthmatic disease are vulnerable to Covid-19 infection.
Traditional methods of autoimmune disease diagnosis are time-consuming and have a high failure rate, resulting in delayed treatment and, at times, wrong diagnosis leading to mistreatment.
In cases where there is delay in diagnosis, the patient is vulnerable to other infections, some of which could be highly transmissible, resulting in high morbidity and mortality rates.
Machine learning and artificial intelligence (AI) can be deployed to identify patients likely to suffer from autoimmune diseases and thus deploy timely diagnosis, monitor disease prognosis and prevent autoimmune comorbidities.
Diagnosis
In a review, machine learning algorithms such as the unsupervised hierarchical clustering and self-organising maps have been used to predict patients for autoimmune diseases from large electronic medical records involving clinical data, genetic, exomic and other types of huge datasets.
Machine learning has been used for diagnosis, classification, predication risk and monitoring of multiple sclerosis, type 1 diabetes, rheumatoid arthritis, inflammatory bowel disease and systemic erythematosus lupus. In all cases, autoimmune diseases were treated in a timely manner and allowed for personalised care.
Kenya, like many other African countries, is experiencing a surge in autoimmune disease and related comorbidities such as Covid-19, cancer and viral infections. There is a need to adopt AI and machine learning in getting patterns from the electronic medical records.
These can then be used for classification and regression of autoimmune disorders, make timely diagnosis and targeted deployment of therapies to avert a serious emerging public health crisis.
Dr Mutua is the director, ImmunoBiologic Research/Consultancy Centre, Makueni. [email protected].
