THE ROLE OF ARTIFICIAL INTELLIGENCE (AI) ON MRI BRAIN EXAMINATION WITH CLINICAL ISCHEMIC STROKE

Abstract

Background: Application of Artificial Intelligence (AI) in radiology is named automatic image interpretation of neuroimaging stroke. It takes a short time to minimize the patient's brain damage. Purpose: Determine the role of AI in ischemic brain stroke MRI examination and find out the advantages and disadvantages of applying AI to ischemic brain stroke MRI examination. Review: It was a descriptive and qualitative study with a literature review approach. The selection of articles used the ScienceDirect, Scopus, ProQuest, PubMed, and Publish or Perish databases. The inclusion criteria included full articles, with the topic of AI on ischemic brain stroke MRI examinations published in the 2017 – 2022 range, articles published by English-language international journals with a classification of Q1 – Q3, and having DOI. Seven relevant pieces of article were obtained, then descriptive analysis was carried out by comparing and presenting the articles descriptively in tabular form. Result: The role of AI in MRI brain examination with clinical ischemic stroke, namely its role in automatic lesion segmentation, Time Since Stroke (TSS) classification, and infarct volume prediction. The advantages of AI included short image processing times and accurate results. The disadvantages of AI tended to decrease performance in small lesions, a large number of patients, limited data, and false positive results. The value of the Dice Score Coefficient (DSC) (0.53 – 0.86) was already high even though it had not reached 1 because it depended on the strength of the data used. Conclusion: The role of AI in MRI imaging of ischemic brain stroke helps in the diagnosis and prognosis of ischemic stroke patients. AI in stroke neuroimaging has advantages in time effectiveness and disadvantages in data limitations.

Keywords

• artificial intelligence
• mri
• brain ischemic stroke