The detection and extraction of complex anatomical structures usually involves a trade-off between the complexity of local feature extraction and classification, and the complexity and performance of the subsequent structural inference from the viewpoint of combinatorial optimization. Concerning the latter, computationally efficient methods are of particular interest that return the globally-optimal structure. We present an efficient method for part-based localization of anatomical structures which embeds contextual shape knowledge in a probabilistic graphical model. It allows for robust detection even when some of the part detections are missing. The application scenario for our statistical evaluation is spine detection and labeling in magnetic resonance images.

Download full-text PDF

Source
http://dx.doi.org/10.1007/978-3-540-73273-0_11DOI Listing

Publication Analysis

Top Keywords

spine detection
8
detection labeling
8
graphical model
8
anatomical structures
8
labeling parts-based
4
parts-based graphical
4
model detection
4
detection extraction
4
extraction complex
4
complex anatomical
4

Similar Publications

Study Design: A prospective web-based survey.

Purpose: Although intraoperative neurophysiological monitoring (IONM) is critical in spine surgery, its usage is largely based on the surgeon's discretion, and studies on its usage trends in Asia-Pacific countries are lacking. This study aimed to examine current trends in IONM usage in Asia-Pacific countries.

View Article and Find Full Text PDF

Background: Spinal epidural arachnoid cysts (SEACs) are rare, non-neoplastic pathologies that can cause compressive myelopathy. Preoperative identification of the exact fistula location is crucial for minimally invasive management.

Methods: This single-center retrospective study included 27 patients with SEACs who underwent "double-needle puncture myelography" to precisely localize the fistula before minimally invasive surgery.

View Article and Find Full Text PDF

Association between resolution of MRI-detected inflammation and improved clinical outcomes in axial spondyloarthritis under long-term anti-TNF therapy.

RMD Open

January 2025

Corporate Member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Department of Gastroenterology, Infectious Diseases and Rheumatology (including Nutrition Medicine), Charite Universitatsmedizin Berlin, Berlin, Germany.

Objectives: In this post-hoc analysis of ESTHER trial, we aimed to investigate the longitudinal relationship between inflammation on MRI and the achievement of inactive disease/low disease activity in patients with axial spondyloarthritis (axSpA) treated with long-term tumor necrosis factor (TNF) inhibitor etanercept.

Methods: Of the 76 patients with active axSpA in the ESTHER trial, we included all patients treated with etanercept for at least 6 months for main analysis. All clinical and MRI data from 4.

View Article and Find Full Text PDF

This study aimed to delineate trends in intervertebral disc degeneration among Mexican patients, specifically focusing on the distribution and correlation between BMI and Pfirrmann classification results within the Mexican population. Conducted using the public health database of Mexico City. The study involved 51 patients sampled via convenience sampling, with exclusive utilization of internal MRI data from L4-L5 and L5-S1 discs.

View Article and Find Full Text PDF

Background: The diagnosis of intracranial extraosseous Ewing's sarcoma (EES) poses challenges due to the absence of specific clinical and imaging features prior to surgery. It is crucial to differentiate the tumor from other small round cell malignancies postoperatively.

Observations: A 7-year-old patient was admitted to the authors' hospital due to the in situ recurrence of a posterior fossa tumor more than 1 month after the initial surgery for headache.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!