Machine learning techniques for prediction of multiple sclerosis progression

Abstract

Patients afflicted by multiple sclerosis experience a relapsing-remitting course in about 85% of the cases. Furthermore, after a 10/15-year period their situation tends to worse, resulting in what is considered the second phase of multiple sclerosis. While treatments are now available to reduce the symptoms and slow down the progression of the disease, the administration of drugs must be adapted to the course of the disease, and predicting relapsing periods and the worsening of the symptoms can greatly improve the outcome of the treatment. For this reason, indicators such as the patient-reported outcome measures (PROMs) have been largely used to support early diagnosis and prediction of future relapsing periods in patients affected by multiple sclerosis. However, such indicators are insufficient, as the prediction they provide is often not accurate enough. In this paper, machine learning techniques have been applied to data obtained from clinical trial, in order to improve the prediction capabilities and provide doctors with an additional instrument to evaluate the clinical situation of patients. After the application of correlation indicators and the use of principal component analysis for the reduction of the dimensionality of the feature space, classification algorithms have been applied and compared, in order to identify the best suiting one for our purposes. After the application of re-balance algorithms, the accuracy of the machine learning-based prediction system reaches 79%, demonstrating the capability of the framework to correctly predict future progression of disability.

Predicting future multiple sclerosis disease progression from MR scans

“…To further improve DL models, it might be beneficial to consider including additional patient and clinical data such as age at onset and relapse, and where available, self-reported quality of life data, as demonstrated by the work of Pinto et al [6] and Branco et al [5].…”

“…In recent years, multiple studies applied machine learning (ML) algorithms to predict future disease progression measured by EDSS. Branco et al [5] aimed to predict EDSS progression after a 3-year period based on multiple questionnaires administered at the time of baseline visit. They trained multiple traditional ML models, such as Logistic regression, Random Forest and Linear Support Vector Classification for three different upsampling techniques and achieved the values of sensitivity and specificity over 0.85 for RF trained with random upsampling.…”

Multiple sclerosis and computational biology (Review)

“…Clinical data may additionally be used in machine learning implementations, mainly for the diagnosis and prediction of the progression of the disease (82). Linear classifiers, such as the one proposed by Branco et al, predict the type of MS progression by using information from a clinical questionnaire (87). Another type of linear classifier, a multi-layer perceptron, made use of MRI metrics generated through graph theory analysis in order to perform classifications of MS types (88).…”

Artificial intelligence will change MS care within the next 10 years: Commentary

“…Moreover, the bias of unseen data propagation in AI could be mitigated with proper study design, quality control, transparency, regulation, and focus on patients’ related outcomes. 1,2 Post hoc studies and quality control checks are necessary to improve the interpretability of AI output. Altogether we believe that the answer to the key question “Artificial Intelligence will change MS care within the next ten years?” should be fluid considering the great potential offered by AI but also the need to further develop procedures to correctly identify data to analyze, to standardize data collection across different MS centers, and to validate the algorithm performance against the neurologist’s one.…”

“…Moreover, the bias of unseen data propagation in AI could be mitigated with proper study design, quality control, transparency, regulation, and focus on patients' related outcomes. 1,2 Post hoc studies and quality control checks are necessary to improve the interpretability of AI output. Altogether we believe that the answer to the key question "Artificial Intelligence will change MS care within the next ten years?"…”