基於多特徵磁振造影與人工智慧之自殺意念的評估方法及評估系統

Translated title of the contribution: Evaluation method and evaluation system for suicidal ideation based on multi-feature magnetic resonance imaging and artificial intelligence

Jun-Cheng Weng (Inventor), Vincent Chin-Hung Chen (Inventor)

Research output: Patent

Abstract

An evaluation method adapted to an evaluation system comprises the following steps of: scanning by a magnetic resonance imaging system to obtain multiple whole brain images at multiple b values; using an arithmetic unit to analyze each whole brain image to obtain a corresponding generalized q-sampling imaging (GQI); using the arithmetic unit to calculate an index value of each voxel for each GQI; using an autoencoder to perform the feature extraction on each index value corresponding to the GQI to obtain a corresponding compressed image; using the arithmetic unit to classify each compressed image according to a machine learning model which is correspondent to the compressed image that has been feature-extracted corresponding to the index values to evaluate whether a corresponding user has a suicidal ideation. Accordingly, it can assist to evaluate and determine whether the user is a depression patient with suicidal ideation.
Translated title of the contributionEvaluation method and evaluation system for suicidal ideation based on multi-feature magnetic resonance imaging and artificial intelligence
Original languageChinese (Traditional)
IPCA61B 5/055(2006.01); G01R 33/48(2006.01)
StatePublished - 01 08 2021

Bibliographical note

公開公告號: 2.0212808E8
Announcement ID: 2.0212808E8

Fingerprint

Dive into the research topics of 'Evaluation method and evaluation system for suicidal ideation based on multi-feature magnetic resonance imaging and artificial intelligence'. Together they form a unique fingerprint.

Cite this