چاپ صفحه |  صفحه نخست سامانه |  نویسندگان |  اطلاعات تفضیلی |  دانلود مقاله | دانشگاه علوم پزشکی تبریز |
| نویسنده | نفر چندم مقاله |
|---|---|
| مرتضی قوجازاده | دوم |
| علیرضا لطفی | سوم |
| هادی صالح پور | اول |
| عنوان | متن |
|---|---|
| کلمات کلیدی | fMRI,Epilepsy,Artificial Intelligence |
| خلاصه مقاله | Background and Aim : Epilepsy is a neurological disease characterized by abnormal neurophysiological activity leading to epileptic seizures or abnormal behavior, accompanied by varying degrees of loss of sensation or consciousness. Epilepsy neuroimaging is important for detecting the seizure onset zone, disease progression, predicting and preventing deficits from surgery and illuminating mechanisms of epileptogenesis. Functional magnetic resonance imaging (fMRI) has been accepted widely for evaluate the functional organization of the human brain. Artificial intelligence (AI) has been helpful in the diagnosis of other conditions like breast cancer, it is expected that soon we will be able to see more reliable results in the field of Epilepsy using artificial intelligence. Methods: A systematic search of Medline (via PubMed), Scopus and Web of Science was conducted from inception to March 2023. The search was performed using MeSH and free keywords such as “Artificial intelligence”, “Machine learning”, “Epilepsy” and “fMRI”. Two authors independently screened located articles in multiple levels of title, abstract, and full-text. Disagreements were resolved by third author opinion. All studies using Artificial Intelligence (AI) models or machine learning algorithms to detect epilepsy or identify the hemisphere of seizure onset, disease progression or any network strength changes in the area of epilepsy using fMRI data were included. Also, we've manually searched references of included studies and gray literature. Microsoft Office Excel 2021 software is used to extract data such as ML algorithms, fMRI Image properties, and the accuracy of models. Results: In a systematic search of databases, 435 articles were identified. After removing duplicates, 233 articles remained, 151 articles were excluded after reviewing the title and abstract, and after reviewing the full-text, 53 articles were excluded. Finally, 29 studies were included in this study. Among them, 23 studies have provided results in detecting accurate area of the disease or seizure onset and progression, 4 of them about lateralization of temporal lobe epilepsy and finally 2 of them have provided results about brain network strength changes in relation with epileptic areas. Support vector machine (SVM) is the most popular ML algorithms. About 70 % of studies classified healthy control hemispheres from seizure onset zones with more than 85% accuracy. Conclusion: The use of AI with fMRI data is offering new opportunities for the diagnosis and localization of epilepsy. By analyzing large datasets of fMRI scans, AI technology can enable clinicians to identify patterns in brain activity that are indicative of epilepsy and personalize surgery and treatment approaches for patients with greater accuracy. Additionally, AI models can help identify biomarkers that could potentially predict seizures and monitor disease progression over time. The integration of AI with fMRI data holds great promise to improve epilepsy care and deepen our comprehension of this complex condition. |
| نام فایل | تاریخ درج فایل | اندازه فایل | دانلود |
|---|---|---|---|
| photo_2024-03-20_00-01-29.jpg | 1403/01/01 | 52904 | دانلود |
| Slide1.JPG | 1403/01/01 | 2319444 | دانلود |
| IMG_20240402_164517_944.jpg | 1403/01/14 | 346342 | دانلود |
| IMG_20240402_164518_025.jpg | 1403/01/14 | 382123 | دانلود |
| Screenshot (1215).png | 1403/01/14 | 376991 | دانلود |
