Computer-aided Detection versus Independent Double Reading of Masses on Mammograms

Karssemeijer, Nico; Otten, J.D.M.; Verbeek, A.L.M.; Groenewoud, Johanna H.; Koning, Harry J. de; Hendriks, Jan H. C. L.; Holland, Roland

doi:10.1148/radiol.2271011962

Cited by 118 publications

(81 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The system uses artificial intelligence and pattern recognition technology to highlight suspicious regions on a mammogram with size-varying marks [10]. Possible masses on the mammogram are marked by ellipses; possible microcalcifications are indicated by squares.…”

Section: Cad Systemmentioning

confidence: 99%

“…Many studies have validated CAD techniques in screening populations and found a sensitivity ranging from 73 [2] to 96% [3]. Moreover, the impact of a CAD system on the performance of mammogram observers was evaluated in several studies [2,[4][5][6][7][8][9][10][11][12][13][14], showing that there is insufficient evidence to claim that CAD improves cancer detection rates but that it does increase recall rates in screening programs for breast cancer [15].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Computer-aided detection in full-field digital mammography in a clinical population: performance of radiologist and technologists

Biggelaar

Kessels

Engelshoven

et al. 2009

Breast Cancer Res Treat

View full text Add to dashboard Cite

The purpose of the study was to evaluate the impact of a computer-aided detection (CAD) system on the performance of mammogram readers in interpreting digital mammograms in a clinical population. Furthermore, the ability of a CAD system to detect breast cancer in digital mammography was studied in comparison to the performance of radiologists and technologists as mammogram readers. Digital mammograms of 1,048 consecutive patients were evaluated by a radiologist and three technologists. Abnormalities were recorded and an imaging conclusion was given as a BI-RADS score before and after CAD analysis. Pathology results during 12 months follow up were used as a reference standard for breast cancer. Fifty-one malignancies were found in 50 patients. Sensitivity and specificity were computed before and after CAD analysis and provided with 95% CIs. In order to assess the detection rate of malignancies by CAD and the observers, the pathological locations of these 51 breast cancers were matched with the locations of the CAD marks and the mammographic locations that were considered to be suspicious by the observers. For all observers, the sensitivity rates did not change after application of CAD. A mean sensitivity of 92% was found for all technologists and 84% for the radiologist. For two technologists, the specificity decreased (from 84 to 83% and from 77 to 75%). For the radiologist and one technologist, the application of CAD did not have any impact on the specificity rates (95 and 83%, respectively). CAD detected 78% of all malignancies. Five malignancies were indicated by CAD without being noticed as suspicious by the observers. In conclusion, the results show that systematic application of CAD in a clinical patient population failed to improve the overall sensitivity of mammogram interpretation by the readers and was associated with an increase in false-positive results. However, CAD marked five malignancies that were missed by the different readers.

show abstract

Section: Cad Systemmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Computer-aided detection in full-field digital mammography in a clinical population: performance of radiologist and technologists

Biggelaar

Kessels

Engelshoven

et al. 2009

Breast Cancer Res Treat

View full text Add to dashboard Cite

show abstract

“…There is also a strong likelihood that digital mammography will replace film in the future, and some evidence this will result in improved accuracy (Fischer et al, 2002;Nawano, 1995). Further, double reading of mammograms is a generally recognized method for improving accuracy (Beam, Conant, and Sickles, 2003;Karssemeijer et al, 2003;Kopans, 2000;Kwek et al, 2003;Liston and Dall, 2003). Kleit and Ruiz (2003) found that availability of previous mammogram reduced false positives by 50 percent.…”

Section: The Practice O F Mammography: Data From One Mammographermentioning

confidence: 99%

Detection and selection decisions in the practice of screening mammography

Stewart

Mumpower

2004

J Policy Anal Manage

View full text Add to dashboard Cite

Body weight and length, chest girth, and seven postcranial limb segment lengths are compared between two guenon species, Chlorocebus (Cercopithecus) aethiops (vervets) and Cercopithecus mitis (blue monkeys), exhibiting different habitual locomotor preferences. The subjects, all adults, were wild caught for a non‐related research project (Turner et al. [1986] Genetic and morphological studies on two species of Kenyan monkeys, C. aethiops and C. mitis. In: Else JG, Lee PC, editors. Primate evolution, proceedings of the Xth International Congress of Primatology, Cambridge. London). The morphological results are interpreted within the context of previously published observations of primate locomotion and social organization. The sample is unique in that the body weight of each individual is known, allowing the effects of body‐size scaling to be assessed in interspecific and intersexual comparisons. C. mitis has a significantly (P < 0.05) greater body weight and trunk length than C. aethiops. A shorter trunk may function to reduce spinal flexibility for ground‐running in the latter. Proximal limb segments (arm and thigh) are significantly greater in C. mitis, reflecting known adaptations to committed arboreal quadrupedal locomotion. By contrast, relative distal limb segments (forearm, crus, and foot) are significantly longer in C. aethiops, concordant with a locomotor repertoire that includes substantial terrestrial quadrupedalism, in addition to arboreal agility, and also the requisite transition between ground and canopy. Although normally associated with arboreal monkeys, greater relative tail length occurs in the more terrestrial vervets. However, because vervets exploit both arboreal and terrestrial habitats, a longer tail may compensate for diminished balance during arboreal quadrupedalism resulting from the greater “brachial” and “crural” indices that enhance their ground quadrupedalism. Most interspecific differences in body proportions are explicable by differences in locomotor modalities. Some results, however, contradict commonly held “tenets” that relate body size and morphology exclusively to locomotion. Generally associated with terrestriality, sexual dimorphism (male/female) is greater in the more arboreal blue monkeys. A more intense, seasonal mating competition may account for this incongruity. Am J Phys Anthropol, 2004. © 2004 Wiley‐Liss, Inc.

show abstract

“…Although there are several CAD studies [6][7][8] that looked at the performance of individual radiologists before and after using CAD, most CAD systems consider consensus as the reference standard for development and evaluation when interpretations of multiple radiologists are available. Within the consensus approach, either only the consensus opinion is known [9][10][11] or individual interpretations are known but a consensus opinion is formed [12][13][14][15][16][17][18][19][20].…”

Section: Introductionmentioning

confidence: 99%

“…The combination was performed in such a manner that a case was considered positive if at least one of the observers marked it as positive. In another study for comparing computer-aided detection versus independent double reading of masses in mammograms, Karssemeijer et al [13] evaluated three reading conditions: a single radiologist's interpretation, emulated double reading (emulation was performed by combining radiologist interpretations pair-wise and then averaging the results of produced pairs), and emulated CAD as a second reader by combining CAD results with single radiologist interpretations. Muratmatsu et al [14] used the mean or mode to form a consensus when investigating a psychophysical measure for evaluation of similar images for mammographic masses.…”

Section: Introductionmentioning

confidence: 99%

Consensus Versus Disagreement in Imaging Research: a Case Study Using the LIDC Database

et al. 2011

View full text Add to dashboard Cite

Traditionally, image studies evaluating the effectiveness of computer-aided diagnosis (CAD) use a single label from a medical expert compared with a single label produced by CAD. The purpose of this research is to present a CAD system based on Belief Decision Tree classification algorithm, capable of learning from probabilistic input (based on intra-reader variability) and providing probabilistic output. We compared our approach against a traditional decision tree approach with respect to a traditional performance metric (accuracy) and a probabilistic one (area under the distance-threshold curve-AuC dt ). The probabilistic classification technique showed notable performance improvement in comparison with the traditional one with respect to both evaluation metrics. Specifically, when applying crossvalidation technique on the training subset of instances, boosts of 28.26% and 30.28% were noted for the probabilistic approach with respect to accuracy and AuC dt , respectively. Furthermore, on the validation subset of instances, boosts of 20.64% and 23.21% were noted again for the probabilistic approach with respect to the same two metrics. In addition, we compared our CAD system results with diagnostic data available for a small subset of the Lung Image Database Consortium database. We discovered that when our CAD system errs, it generally does so with low confidence. Predictions produced by the system also agree with diagnoses of truly benign nodules more often than radiologists, offering the possibility of reducing the false positives.

show abstract

Computer-aided Detection versus Independent Double Reading of Masses on Mammograms

Abstract: Although independent double reading yields the best detection performance, the presence and probability of CAD mass markers can improve mammogram interpretation.

Cited by 118 publications

References 21 publications

Computer-aided detection in full-field digital mammography in a clinical population: performance of radiologist and technologists

Computer-aided detection in full-field digital mammography in a clinical population: performance of radiologist and technologists

Detection and selection decisions in the practice of screening mammography

Consensus Versus Disagreement in Imaging Research: a Case Study Using the LIDC Database

Contact Info

Product

Resources

About