The image analysis of the brain with machine learning continues to be a relevant work for the detection of different characteristics of this complex organ. Recent research has observed that there are differences in the structure of the brain, specifically in white matter, when learning and using a second language. This work focuses on knowing the brain from the classification of Magnetic Resonance Images (MRIs) of bilingual and monolingual people who have English as their common language. Different artificial neural networks of a hidden layer were tested until reaching two neurons in that layer. The number of entries used was nine hundred and the classifier registered a high percentage of effectiveness. The training was supervised which could be improved in a future investigation. This task is usually carried out by an expert human with Tract-Based Spatial Statistics analysis and fractional anisotropy expressed in different colors on a screen. So, this proposal presents another option to quantitatively analyse this type of phenomena which allows to contribute to neuroscience by automatically detecting bilingual people of monolinguals by using machine learning from MRIs. This reinforces what is reported in manual detections and the way that a machine can do it.

中文翻译：

---

机器学习对两种语言的人的脑图像分类。

通过机器学习对大脑进行图像分析仍然是检测这种复杂器官不同特征的一项重要工作。最近的研究发现，学习和使用第二语言时，大脑的结构存在差异，特别是在白质方面。这项工作的重点是从以英语为通用语言的双语和单语人士的磁共振图像（MRI）分类中了解大脑。测试了隐藏层的不同人工神经网络，直到到达该层中的两个神经元为止。使用的条目数为900，分类器的有效性百分比很高。对培训进行了监督，可以在以后的调查中加以改进。此任务通常由具有基于行迹的空间统计分析和以不同颜色在屏幕上表示的分数各向异性的专家人员完成。因此，该建议提出了另一种方法来定量分析这种现象，该现象通过使用MRI机器学习自动检测单语种的双语人士，从而有助于神经科学。这加强了手动检测中报告的内容以及机器执行此操作的方式。