Meeting Banner
Abstract #4767

Unsupervised AnomalyCEST: Unsupervised anomaly detection using CEST MRI of brain at 7T

Anshuman Swain1, Paul Jacobs1, Abeer Mathur1, and Ravinder Reddy1
1Center for Advanced Metabolic Imaging in Precision Medicine, Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

Synopsis

Keywords: CEST / APT / NOE, Diagnosis/Prediction, AI/ML, Anomaly detection, CEST/MT, Convolutional neural network

Motivation: Chemical exchange saturation transfer (CEST) MRI is a sensitive technique that provides metabolic information in vivo. However, conventional methods of deriving CEST metrics present limits of ill-posed conditions and computational complexity, while comparing healthy and pathological individuals requires sufficiently large sample sizes.

Goal(s): Use unsupervised anomaly detection for the assessment of anomalous voxels on an individual basis, circumventing the need for deriving traditional CEST metrics.

Approach: We implement a 1D convolutional autoencoder trained on healthy participants to detect anomalous regions in an MS subject by identifying voxels with high reconstruction error.

Results: The network effectively captures anomalous regions on an individual basis.

Impact: Traditional CEST MRI methods require large sample sizes and complex fitting techniques to compare healthy individuals with those having pathology. Unsupervised anomaly detection of Z-spectra will enable the detection of anomalous voxels linked to pathology at an individual level.

How to access this content:

For one year after publication, abstracts and videos are only open to registrants of this annual meeting. Registrants should use their existing login information. Non-registrant access can be purchased via the ISMRM E-Library.

After one year, current ISMRM & ISMRT members get free access to both the abstracts and videos. Non-members and non-registrants must purchase access via the ISMRM E-Library.

After two years, the meeting proceedings (abstracts) are opened to the public and require no login information. Videos remain behind password for access by members, registrants and E-Library customers.

Click here for more information on becoming a member.

Keywords