错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Convolutional Neural Network and Recursive Feature Elimination Based Model for the Diagnosis of Mild Cognitive Impairments

  • Harsh Bhasin,
  • Abheer Mehrotra,
  • Ansh Ohri

摘要

Aging leads to reduced cognitive abilities. A person heading toward dementia will also show such signs but with more severity. This stage is referred to as Mild Cognitive Impairment. It has been observed that nearly one-fifth of the people suffering from MCI convert to dementia. The patients suffering from MCI who convert to dementia are called MCI-Converts and those who do not convert are called MCI Non-converts. This work proposes a model that extracts gray matter from the s-MRI brain volume and explores the applicability of combination of Convolutional Neural Network and Recursive Feature Elimination in the diagnosis of MCI. Features of the data obtained using the carefully crafted CNN followed by the selection of the appropriate features using Recursive Feature Elimination are then used to accomplish the task. Empirical analysis has been done to select the parameters of the CNN. The results are better than the state of the art and pave way for the exploitation of the Deep Learning models to classify the Converts and the Non-converts.