Gross feature recognition of Anatomical Images based on Atlas grid (GAIA): Incorporating the local discrepancy between an atlas and a target image to capture the features of anatomic brain MRI

Abstract

We aimed to develop a new method to convert T1-weighted brain MRIs to feature vectors, which could be used for content-based image retrieval (CBIR). To overcome the wide range of anatomical variability in clinical cases and the inconsistency of imaging protocols, we introduced the Gross feature recognition of Anatomical Images based on Atlas grid (GAIA), in which the local intensity alteration, caused by pathological (e.g., ischemia) or physiological (development and aging) intensity changes, as well as by atlas–image misregistration, is used to capture the anatomical features of target images.As a proof-of-concept, the GAIA was applied for pattern recognition of the neuroanatomical features of multiple stages of Alzheimer's disease, Huntington's disease, spinocerebellar ataxia type 6, and four subtypes of primary progressive aphasia. For each of these diseases, feature vectors based on a training dataset were applied to a test dataset to evaluate the accuracy of pattern recognition. The feature vectors extracted from the training dataset agreed well with the known pathological hallmarks of the selected neurodegenerative diseases. Overall, discriminant scores of the test images accurately categorized these test images to the correct disease categories. Images without typical disease-related anatomical features were misclassified. The proposed method is a promising method for image feature extraction based on disease-related anatomical features, which should enable users to submit a patient image and search past clinical cases with similar anatomical phenotypes.

High-throughput neuro-imaging informatics

“…To appropriately integrate diagnostic information to characterize the anatomical features related to each disease category, PCA and LDA were applied sequentially to a dataset consisting of 102 T1-weighted images from AD, primary progressive aphasia, Huntington's disease, hereditary spinocerebellar ataxia and normal control participants. These were parcellated based on the JHU-atlas [the images used for this analysis are a portion of a dataset published with the methodological detail (Qin et al, 2013 )]. The weighted feature vectors efficiently captured known disease-specific anatomical alterations.…”

OpenMAP-T1: A Rapid Deep Learning Approach to Parcellate 280 Anatomical Regions to Cover the Whole Brain

“…Brain image parcellation constitutes a pivotal aspect of neuroscientific and clinical investigations, delineating a repertoire of parcels that correspond to biologically or functionally pertinent cerebral units. These defined parcels facilitate quantitative analyses of neuroimaging data for each individual region [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19]. While numerous criteria exist for the parcellation of cerebral territories, the designation of regional labels predominantly relies on established anatomical or neurofunctional insights.…”

Efficient feature embedding of 3D brain MRI images for content-based image retrieval with deep metric learning

“…Hence, these technologies are not suitable for handling the images stored in a PACS, which may contain a wide variety of diseases with different pathological features. An atlas-based brain MRI parcellation approach, in which the anatomical and pathological features of the brain are extracted from local brain volumes or intensities obtained from approximately 250 anatomical structures, has demonstrated excellent performance in terms of retrieval when applied to neurodegenerative diseases such as primary progressive aphasia [6], [7], AD, Huntington's disease, and spinocerebellar ataxia [7]. The major advantage of the atlas-based approach is the anatomically meaningful and highly effective dimension reduction, which makes the biological and pathological interpretation of the CBIR results straightforward.…”