AI Approaches in Computer-Aided Diagnosis and Recognition of Neoplastic Changes in MRI Brain Images
Advanced diagnosis systems provide doctors with an abundance of high-quality data, which allows for diagnosing dangerous diseases, such as brain cancers. Unfortunately, humans flooded with such plentiful information might overlook tumor symptoms. Hence, diagnostical devices are becoming more commonl...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-11-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/12/23/11880 |
| _version_ | 1797463676429533184 |
|---|---|
| author | Jakub Kluk Marek R. Ogiela |
| author_facet | Jakub Kluk Marek R. Ogiela |
| author_sort | Jakub Kluk |
| collection | DOAJ |
| description | Advanced diagnosis systems provide doctors with an abundance of high-quality data, which allows for diagnosing dangerous diseases, such as brain cancers. Unfortunately, humans flooded with such plentiful information might overlook tumor symptoms. Hence, diagnostical devices are becoming more commonly combined with software systems, enhancing the decisioning process. This work picks up the subject of designing a neural network based system that allows for automatic brain tumor diagnosis from MRI images and points out important areas. The application intends to speed up the diagnosis and lower the risk of slipping up on a neoplastic lesion. The study based on two types of neural networks, Convolutional Neural Networks and Vision Transformers, aimed to assess the capabilities of the innovative ViT and its possible future evolution compared with well-known CNNs. The research reveals a tumor recognition rate as high as 90% with both architectures, while the Vision Transformer turned out to be easier to train and provided more detailed decision reasoning. The results show that computer-aided diagnosis and ViTs might be a significant part of modern medicine development in IoT and healthcare systems. |
| first_indexed | 2024-03-09T17:54:07Z |
| format | Article |
| id | doaj.art-4ecb8c853afb4da6a4e99071b9f3212b |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-09T17:54:07Z |
| publishDate | 2022-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-4ecb8c853afb4da6a4e99071b9f3212b2023-11-24T10:27:22ZengMDPI AGApplied Sciences2076-34172022-11-0112231188010.3390/app122311880AI Approaches in Computer-Aided Diagnosis and Recognition of Neoplastic Changes in MRI Brain ImagesJakub Kluk0Marek R. Ogiela1Cryptography and Cognitive Informatics Laboratory, AGH University of Science and Technology, 30 Mickiewicza Ave., 30-059 Krakow, PolandCryptography and Cognitive Informatics Laboratory, AGH University of Science and Technology, 30 Mickiewicza Ave., 30-059 Krakow, PolandAdvanced diagnosis systems provide doctors with an abundance of high-quality data, which allows for diagnosing dangerous diseases, such as brain cancers. Unfortunately, humans flooded with such plentiful information might overlook tumor symptoms. Hence, diagnostical devices are becoming more commonly combined with software systems, enhancing the decisioning process. This work picks up the subject of designing a neural network based system that allows for automatic brain tumor diagnosis from MRI images and points out important areas. The application intends to speed up the diagnosis and lower the risk of slipping up on a neoplastic lesion. The study based on two types of neural networks, Convolutional Neural Networks and Vision Transformers, aimed to assess the capabilities of the innovative ViT and its possible future evolution compared with well-known CNNs. The research reveals a tumor recognition rate as high as 90% with both architectures, while the Vision Transformer turned out to be easier to train and provided more detailed decision reasoning. The results show that computer-aided diagnosis and ViTs might be a significant part of modern medicine development in IoT and healthcare systems.https://www.mdpi.com/2076-3417/12/23/11880artificial neural networkscomputer visionmagnetic resonance imagingconvolutional networksvision transformercancerous diseases |
| spellingShingle | Jakub Kluk Marek R. Ogiela AI Approaches in Computer-Aided Diagnosis and Recognition of Neoplastic Changes in MRI Brain Images Applied Sciences artificial neural networks computer vision magnetic resonance imaging convolutional networks vision transformer cancerous diseases |
| title | AI Approaches in Computer-Aided Diagnosis and Recognition of Neoplastic Changes in MRI Brain Images |
| title_full | AI Approaches in Computer-Aided Diagnosis and Recognition of Neoplastic Changes in MRI Brain Images |
| title_fullStr | AI Approaches in Computer-Aided Diagnosis and Recognition of Neoplastic Changes in MRI Brain Images |
| title_full_unstemmed | AI Approaches in Computer-Aided Diagnosis and Recognition of Neoplastic Changes in MRI Brain Images |
| title_short | AI Approaches in Computer-Aided Diagnosis and Recognition of Neoplastic Changes in MRI Brain Images |
| title_sort | ai approaches in computer aided diagnosis and recognition of neoplastic changes in mri brain images |
| topic | artificial neural networks computer vision magnetic resonance imaging convolutional networks vision transformer cancerous diseases |
| url | https://www.mdpi.com/2076-3417/12/23/11880 |
| work_keys_str_mv | AT jakubkluk aiapproachesincomputeraideddiagnosisandrecognitionofneoplasticchangesinmribrainimages AT marekrogiela aiapproachesincomputeraideddiagnosisandrecognitionofneoplasticchangesinmribrainimages |