Using AI for autoimmune disease management

Main applications of AI in autoimmune diseases:

AI is a popular tool used in the diagnosis of autoimmune diseases worldwide. It can improve efficiency and enhance the decision–making process by evaluating large data volumes. Computer algorithms, speech recognition, biometrics, and motion detection are some of the existing applications of AI. AI systems generally operate by taking in large quantities of labeled training data, analyzing the data on the association and trends, and making forecasts of future actions utilizing these patterns. The development of AI is based on three cognitions i.e., learning, reasoning, and self–correction (1). The primary purpose of AI is to collect data and develop protocols to further convert it into usable information. Algorithms are step–by–step commands given to the computer equipment to achieve a specific operation (3).

Key applications of AI in autoimmune diseases:

1. Recognition of patients suffering from autoimmune diseases:

In recent years, autoimmune disease studies have utilized the ML approach and speech recognition techniques to identify individuals with autoimmune illnesses from electronic medical data. These methods are demonstrated to accommodate the International Classification of Diseases billing codes. With the utilization of language modeling, electronic health records were able to recognize comorbidities associated with both alopecia and vitiligo. Thus, it was revealed that autoimmune disorders influenced both diseases (1).

2. Autoimmune disease diagnosis

AI technologies are most commonly used for the diagnosis of patients suffering from autoimmune illnesses. The purpose of the investigations is to differentiate patients from control subjects. The studies utilize a targeted ML approach for the early diagnosis of degenerative diseases such as Multiple Sclerosis (MS) and Rheumatoid Arthritis (RA). Immune system dysregulation including the innate and adaptive immune responses, genetic predisposition, and environmental variables— are the three factors that play a primary role in autoimmune disease diagnosis (2). The most common AI algorithms used are random forests and support vector machines. The AI-based methods are also employed for diagnostic classification involving subjects with various other autoimmune disorders such as controls, to differentiate between the overlapped diseases or similar symptoms or phenotypes, like Irritational Bowel Syndrome (IBS) and celiac stratification, or classification of several autoimmune disorders (1).

3. Categorization of autoimmune disease subgroups

One of the major applications of AI is the classification of autoimmune disorders subtypes. The three types of uncontrolled clustering used by the researchers are hierarchic clustering to detect new subtypes of autoimmune disorders; consensus clustering to categorize high, low, and mixed RA inflammatory response; and agglomerate hierarchical clustering to genetic signature cluster MS. The main types of AI-based algorithms used are support vector machines and random forests. Different types of data can be utilized such as genetic information and magnetic resonance imaging (MRI) scans (1).

4. To assess the risk of autoimmune disorders

Autoimmune diseases are linked with the predictions of the disease risk and the exploration of new risk variables through feature selection. In the majority of the studies, GWAS and exome data are used. In some research, single SNPs inside HLA areas or pre-selected genes and biomarkers were used to evaluate the risk. Clinical and genetic data were also studied along with patient medical data. The AI-based models utilized for data analysis were random forests, logistic regression, and support vector machines (1).

5. Monitoring of autoimmune diseases

Several studies of autoimmune illnesses like type 1 diabetes (T1D) apply the ML approach to track and manage predicted blood glucose levels, recognize hypoglycaemic incidents, and assist in decision-making via the utilization of case-based rationale or predictive analytics. Movement-tracking techniques for MS and RA were developed using activity measures. Support vector regression was the most frequently used AI-based algorithm (1).

6. To evaluate the progression and outcome of autoimmune diseases

The studies of immunodeficiency diseases majorly focused on the course and outcome of the illness. The main factors considered for this research are disease progression, treatment response, and survival prognosis. AI-based models are utilized to improve picture segmentation to aid in the prediction of prognostic factors. Clinical images were employed in these studies. Few researchers used omics information as the data source. Thus, the AI-based algorithms used in these studies were support vector machines, neural networks, and random forests (1).

7. Identification of relevant therapeutic targets

The model for autoinflammatory disorders (AIIDs) is based on the clustering of thousands of patient molecular profiling data as gene enrichment techniques are then applied to identify the dysregulated molecular pathways in comparison to healthy controls. Furthermore, network analysis and research are then conducted to represent the biological processes-based pathophysiology of the disease and to provide guidelines for novel treatment approaches. Using AIIDS models, precision medicine modalities that are most suited for certain patient groups may be implemented by developing novel drugs that interact with target candidates, finding pertinent combination treatments, or repurposing already-existing compounds (4).

Conclusion

AI is recently gaining popularity to solve health issues caused by autoimmune disorders. Around eighty types of autoimmune diseases impact a wide range of body parts. Sometimes, diagnosis of the illnesses can be challenging due to overlapping symptoms. Generally, autoimmune diseases are treated by reducing the immune system activation. The subsets of AI like ML and DL can assist in accurate clinical decision-making. The seven major applications of AI in autoimmune diseases are the identification of autoimmune disease patients, diagnosis of autoimmune diseases, classification of disease subtypes, assessment of involved risk factors, monitoring of disease, identification of therapeutic targets, and evaluation of the progress and outcome of disease(1,4). The most common AI-based algorithms used are support vector machines, neural networks, logistic regression, and random forests. The most frequently used patient information sources are clinical, medical, and genetic data. Thus, AI can help in making healthcare avenues more predictive and patient-friendly by analyzing relevant patient data by discovering newer, improved preventive care strategies for patients.

References

1. Mohammed IA. A systematic review of the applications of artificial intelligence in autoimmune diseases. International Journal of Creative Research Thoughts 2021; 9(7): g658 – g661.

2. C. Cosgriff, D. Stone, G. Weissman, R. Pirracchio and L. Celi, "The clinical artificial intelligence department: a prerequisite for success", BMJ Health & Care Informatics, vol. 27, no. 1, p. e100183, 2020.

3. S. Secinaro, D. Calandra, A. Secinaro, V. Muthurangu and P. Biancone, "The role of artificial intelligence in healthcare: a structured literature review", BMC Medical Informatics and Decision Making, vol. 21, no. 1, 2021.

4. Moingeon P. Artificial intelligence-driven drug development against autoimmune diseases. Trends in Pharmacological Sciences 2023;44(7): 411-424.

Tags: