An Explainable Machine Learning Pipeline for Stroke Prediction on Imbalanced Data

Kokkotis, Christos; Giarmatzis, Georgios; Γιαννακού, Ερασμία; Moustakidis, Serafeim; Tsatalas, Themistoklis; Tsiptsios, Dimitrios; Vadikolias, Κonstantinos; Aggelousis, Nikolaos

doi:10.3390/diagnostics12102392

Cited by 49 publications

(25 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given the age-related nature of the disease, since more than half of all patients are aged over 65 [ 3 ], in combination with the increasing global population and the constant improvement of life expectancy [ 4 ], the number of stroke survivors will rise in an unprecedented manner. Consequently, the early and precise recognition of patients with unfavorable prognoses is essential in order to personalize treatment and tailor rehabilitation to each individual’s needs [ 5 ].…”

Section: Introductionmentioning

confidence: 99%

Exploring the Impact of Cerebral Microbleeds on Stroke Management

Sousanidou

Tsiptsios

Christidi

et al. 2023

Neurology International

Self Cite

View full text Add to dashboard Cite

Stroke constitutes a major cause of functional disability and mortality, with increasing prevalence. Thus, the timely and accurate prognosis of stroke outcomes based on clinical or radiological markers is vital for both physicians and stroke survivors. Among radiological markers, cerebral microbleeds (CMBs) constitute markers of blood leakage from pathologically fragile small vessels. In the present review, we evaluated whether CMBs affect ischemic and hemorrhagic stroke outcomes and explored the fundamental question of whether CMBs may shift the risk–benefit balance away from reperfusion therapy or antithrombotic use in acute ischemic stroke patients. A literature review of two databases (MEDLINE and Scopus) was conducted to identify all the relevant studies published between 1 January 2012 and 9 November 2022. Only full-text articles published in the English language were included. Forty-one articles were traced and included in the present review. Our findings highlight the utility of CMB assessments, not only in the prognostication of hemorrhagic complications of reperfusion therapy, but also in forecasting hemorrhagic and ischemic stroke patients’ functional outcomes, thus indicating that a biomarker-based approach may aid in the provision of counseling for patients and families, improve the selection of more appropriate medical therapies, and contribute to a more accurate choice of patients for reperfusion therapy.

show abstract

Section: Introductionmentioning

confidence: 99%

Exploring the Impact of Cerebral Microbleeds on Stroke Management

Sousanidou

Tsiptsios

Christidi

et al. 2023

Neurology International

Self Cite

View full text Add to dashboard Cite

show abstract

“…From Table 2, it is observed that the proposed Rough-Fuzzy based Synthetic Data Generation (RFSDG) method provides the best result when the RF classifier model is applied to BS (obtained using both oversampling and undersampling technique) dataset. So, we have considered RFSDG + RF as the final proposed model and compared it with some state-of-the-art methods Chawla et al [11], Lee and Kim [59], Schölkopf et al [60], Vuttipittayamongkol and Elyan [61], Zhenchuan Li [20], Kokkotis et al [62], described below.…”

Section: Comparison With Other Methodsmentioning

confidence: 99%

Rough-Fuzzy Based Synthetic Data Generation Exploring Boundary Region of Rough Sets to Handle Class Imbalance Problem

Naushin

Das

Nayak

et al. 2023

Axioms

View full text Add to dashboard Cite

Class imbalance is a prevalent problem that not only reduces the performance of the machine learning techniques but also causes the lacking of the inherent complex characteristics of data. Though the researchers have proposed various ways to deal with the problem, they have yet to consider how to select a proper treatment, especially when uncertainty levels are high. Applying rough-fuzzy theory to the imbalanced data learning problem could be a promising research direction that generates the synthetic data and removes the outliers. The proposed work identifies the positive, boundary, and negative regions of the target set using the rough set theory and removes the objects in the negative region as outliers. It also explores the positive and boundary regions of the rough set by applying the fuzzy theory to generate the samples of the minority class and remove the samples of the majority class. Thus the proposed rough-fuzzy approach performs both oversampling and undersampling to handle the imbalanced class problem. The experimental results demonstrate that the novel technique allows qualitative and quantitative data handling.

show abstract

“…Thus, a crucial need for prompt and accurate identification of patients with an unfavorable prognosis has emerged. Additionally, such an approach may allow the development of personalized rehabilitation programs based on the propensity for recovery of each individual [ 4 ].…”

Section: Introductionmentioning

confidence: 99%

Leukoaraiosis as a Predictor of Depression and Cognitive Impairment among Stroke Survivors: A Systematic Review

Tziaka

Christidi

Tsiptsios

et al. 2023

Neurology International

Self Cite

View full text Add to dashboard Cite

Stroke survivors are at increased risk of developing depression and cognitive decline. Thus, it is crucial for both clinicians and stroke survivors to be provided with timely and accurate prognostication of post-stroke depression (PSD) and post-stroke dementia (PSDem). Several biomarkers regarding stroke patients’ propensity to develop PSD and PSDem have been implemented so far, leukoaraiosis (LA) being among them. The purpose of the present study was to review all available work published within the last decade dealing with pre-existing LA as a predictor of depression (PSD) and cognitive dysfunction (cognitive impairment or PSDem) in stroke patients. A literature search of two databases (MEDLINE and Scopus) was conducted to identify all relevant studies published between 1 January 2012 and 25 June 2022 that dealt with the clinical utility of preexisting LA as a prognostic indicator of PSD and PSDem/cognitive impairment. Only full-text articles published in the English language were included. Thirty-four articles were traced and are included in the present review. LA burden, serving as a surrogate marker of “brain frailty” among stroke patients, appears to be able to offer significant information about the possibility of developing PSD or cognitive dysfunction. Determining the extent of pre-existing white matter abnormalities can properly guide decision making in acute stroke settings, as a greater degree of such lesioning is usually coupled with neuropsychiatric aftermaths, such as PSD and PSDem.

show abstract

An Explainable Machine Learning Pipeline for Stroke Prediction on Imbalanced Data

Cited by 49 publications

References 47 publications

Exploring the Impact of Cerebral Microbleeds on Stroke Management

Exploring the Impact of Cerebral Microbleeds on Stroke Management

Rough-Fuzzy Based Synthetic Data Generation Exploring Boundary Region of Rough Sets to Handle Class Imbalance Problem

Leukoaraiosis as a Predictor of Depression and Cognitive Impairment among Stroke Survivors: A Systematic Review

Contact Info

Product

Resources

About