skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Genetic Algorithm for Analysis of Abdominal Aortic Aneurysms in Radiology Reports

Abstract

An abdominal aortic aneurysm is a problem in which the wall of the artery that supplies blood to the abdomen and lower extremities expands under pressure or balloons outward. Patients must undergo surgery to repair such aneurysm, and there is currently no known indicator of success or failure from this surgery. Our work uses a genetic algorithm to analyze radiology reports from these patients to look for common patterns in the language used as well as common features of both successful and unsuccessful surgieries. The results of the genetic algorithm show that patients with complications or unusual characteristics can be identified from a set of radiology reports without the use of search keywords, clustering, categorization, or ontology. This allows medical researchers to search and identify interesting patient records without the need for explicitly defining what interesting patient records are.

Authors:
 [1];  [1];  [1];  [1]
  1. ORNL
Publication Date:
Research Org.:
Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
Sponsoring Org.:
USDOE Laboratory Directed Research and Development (LDRD) Program
OSTI Identifier:
986825
DOE Contract Number:
AC05-00OR22725
Resource Type:
Conference
Resource Relation:
Conference: Genetic and Evoluationary Computation Conference 2010 (GECCO 2010), Portland, OR, USA, 20100707, 20100707
Country of Publication:
United States
Language:
English

Citation Formats

Patton, Robert M, Beckerman, Barbara G, Treadwell, Jim N, and Potok, Thomas E. Genetic Algorithm for Analysis of Abdominal Aortic Aneurysms in Radiology Reports. United States: N. p., 2010. Web. doi:10.1145/1830761.1830828.
Patton, Robert M, Beckerman, Barbara G, Treadwell, Jim N, & Potok, Thomas E. Genetic Algorithm for Analysis of Abdominal Aortic Aneurysms in Radiology Reports. United States. doi:10.1145/1830761.1830828.
Patton, Robert M, Beckerman, Barbara G, Treadwell, Jim N, and Potok, Thomas E. Fri . "Genetic Algorithm for Analysis of Abdominal Aortic Aneurysms in Radiology Reports". United States. doi:10.1145/1830761.1830828.
@article{osti_986825,
title = {Genetic Algorithm for Analysis of Abdominal Aortic Aneurysms in Radiology Reports},
author = {Patton, Robert M and Beckerman, Barbara G and Treadwell, Jim N and Potok, Thomas E},
abstractNote = {An abdominal aortic aneurysm is a problem in which the wall of the artery that supplies blood to the abdomen and lower extremities expands under pressure or balloons outward. Patients must undergo surgery to repair such aneurysm, and there is currently no known indicator of success or failure from this surgery. Our work uses a genetic algorithm to analyze radiology reports from these patients to look for common patterns in the language used as well as common features of both successful and unsuccessful surgieries. The results of the genetic algorithm show that patients with complications or unusual characteristics can be identified from a set of radiology reports without the use of search keywords, clustering, categorization, or ontology. This allows medical researchers to search and identify interesting patient records without the need for explicitly defining what interesting patient records are.},
doi = {10.1145/1830761.1830828},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Fri Jan 01 00:00:00 EST 2010},
month = {Fri Jan 01 00:00:00 EST 2010}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • The association between AAA and gender, smoking (SM), hypertension (HTN) and inguinal herniation (IH) was examined in 141 AAA probands and 139 of their 1st degree relatives with aortic exam (36 affected, 103 unaffected). There was no significant difference between age at diagnosis of affecteds and age at exam of unaffecteds. Of 181 males, 142 had AAA; of 99 females, 35 had AAA. Using log-linear modeling AAA was significantly associated at the 5% level with gender, SM and HTN but not IH. The association of AAA with SM and HTN held when males and females were analyzed separately. HTN wasmore » -1.5 times more common in both affected males and females, while SM was 1.5 and 2 times more common in affected males and females, respectively. Tests of association and linkage analyses were performed with relevant candidate genes: 3 COL3A1 polymorphisms (C/T, ALA/THR, AvaII), 2 ELN polymorphisms (SER/GLY, (CA)n), FBN1(TAAA)n, 2 APOB polymorphisms (Xbal,Ins/Del), CLB4B (CA)n, PI and markers D1S243 (CA)n, HPR (CA)n and MFD23(CA)n. The loci were genotyped in > 100 AAA probands and > 95 normal controls. No statistically significant evidence of association at the 5% level was obtained for any of the loci using chi-square test of association. 28 families with 2 or more affecteds were analyzed using the affected pedigree member method (APM) and lod-score analyses. There was no evidence for linkage with any loci using APM. Lod-score analysis under an autosomal recessive model resulted in excluding linkage (lod score < -2) of all loci to AAA at {theta}=0.0. Under an autosomal dominant model, linkage was excluded at {theta}=0.0 to ELN, APOB, CLG4B, D1S243, HPR and MFD23. The various genes previously proposed in AAA pathogenesis are neither associated nor casually related in our study population.« less
  • Radiologists disagree with each other over the characteristics and features of what constitutes a normal mammogram and the terminology to use in the associated radiology report. Recently, the focus has been on classifying abnormal or suspicious reports, but even this process needs further layers of clustering and gradation, so that individual lesions can be more effectively classified. Using a genetic algorithm, the approach described here successfully learns phrase patterns for two distinct classes of radiology reports (normal and abnormal). These patterns can then be used as a basis for automatically analyzing, categorizing, clustering, or retrieving relevant radiology reports for themore » user.« less
  • Various computer-assisted technologies have been developed to assist radiologists in detecting cancer; however, the algorithms still lack high degrees of sensitivity and specificity, and must undergo machine learning against a training set with known pathologies in order to further refine the algorithms with higher validity of truth. This work describes an approach to learning cue phrase patterns in radiology reports that utilizes a genetic algorithm (GA) as the learning method. The approach described here successfully learned cue phrase patterns for two distinct classes of radiology reports. These patterns can then be used as a basis for automatically categorizing, clustering, ormore » retrieving relevant data for the user.« less
  • To quantitatively characterize processes of platelet thrombus formation in vivo, the kinetics and incorporation into thrombus of autologous In-111-labeled platelets were compared in six patients with aortic aneurysms and in seven patients with prosthetic aortic grafts. Although platelet survival was comparably shortened in both patient groups (mean, 5.8 days), the maximum radioactivity as determined by gamma camera imaging was higher in the aneurysms than in the grafts (3.3% +/- 1.6% vs. 1.6% +/- 1.1%, p = 0.05). Maximum In-111 uptake was also attained more quickly in the aneurysm patients (2.3 +/- 0.8 days vs. 3.5 +/- 1.3 days; p =more » 0.07). The experimental platelet kinetic and imaging data were subsequently evaluated by compartmental analysis to estimate both normal and disease-related components of platelet destruction. This analysis indicated that deposited platelet radioactivity had a longer residence time on grafts (2.9 +/- 1.7 days vs. 1.4 +/- 0.9 days, p = 0.07) but accumulated at a faster rate in aneurysms (5.0% +/- 3.4% per day vs. 1.4% +/- 0.9% per day, p = 0.02). As determined by imaging, only a proportion of increased platelet destruction was specifically due to the aneurysms (55% +/- 38%) or grafts (17% +/- 11%, p = 0.03). This result indicates additional components of platelet destruction unrelated to graft and aneurysm thrombus formation which, in some graft patients, may reflect a greater severity of vascular disease or other mechanisms causing a preferential shortening of platelet survival. Thus, the analytical approach described may be a useful one for discriminating components of in vivo platelet utilization including platelet removal due to normal hemostatic and senescent mechanisms, localized thrombus formation, and more generalized vascular disease.« less
  • Magnetic resonance imaging (MRI) was performed in 20 patients with radiologically or surgically proven abdominal aortic aneurysms using a Siemens Magnetom scanner with a 0.35-T superconductive magnet. Of nine patients who underwent surgical repair, MRI correctly demonstrated the origin of the aortic aneurysm in nine and accurately determined the status of the iliac arteries in eight. Of 11 patients who did not have surgical repair, MRI findings correlated well with other radiologic studies. MRI was found to be more reliable than sonography in determining the relation between the aneurysm and the renal arteries as well as the status of themore » iliac arteries. Despite these advantages, the authors still advocate sonography as the screening procedure of choice in patients with suspected abdominal aortic aneurysms because of its lower cost and ease of performance. MRI should be reserved for patients who have had unsuccessful or equivocal sonographic examinations.« less