Performance Parameters for Screening and Diagnostic Mammography: Specialist and General Radiologists

Sickles, Edward A.; Wolverton, Dulcy; Dee, Katherine E.

doi:10.1148/radiol.2243011482

Cited by 225 publications

(132 citation statements)

References 20 publications

Supporting

Mentioning

121

Contrasting

Unclassified

Order By: Relevance

“…Neuroradiologists, oncologic radiologists, and mammography-trained specialists were better at interpreting images within their areas of expertise than community radiologists with many years of experience. [24][25][26] The converse argument also seems to be true. A study by Branstetter et al 27 documents that subspecialty radiologists outside of their field of expertise do worse than senior radiology residents when interpreting basic body films, further highlighting the importance of ongoing experience.…”

Section: Discussionmentioning

confidence: 94%

Quality Control in Neuroradiology: Discrepancies in Image Interpretation among Academic Neuroradiologists

Babiarz¹,

Yousem²

2011

AJNR Am J Neuroradiol

View full text Add to dashboard Cite

SUMMARY:Prior studies have found a 3%-6% clinically significant error rate in radiology practice. We set out to assess discrepancy rates between subspecialty-trained university-based neuroradiologists. Over 17 months, university neuroradiologists randomly reviewed 1000 studies and reports of previously read examinations of patients in whom follow-up studies were read. The discrepancies between the original and "second opinion" reports were scored according to a 5-point scale: 1, no change; 2, clinically insignificant detection discrepancy; 3, clinically insignificant interpretation discrepancy; 4, clinically significant detection discrepancy; and 5, clinically significant interpretation discrepancy. Of the 1000 studies, 876 (87.6%) showed agreements with the original report. The neuroradiology division had a 2.0% (20/1000; 95% CI, 1.1%-2.9%) rate of clinically significant discrepancies involving 8 CTs and 12 MR images. Discrepancies were classified as vascular (n ϭ 7), neoplastic (n ϭ 9), congenital (n ϭ 2), and artifacts (n ϭ 2). Individual neuroradiologist's scores ranged from 0% to 7.7% Ϯ 2.3% (n ϭ 18). Both CT and MR imaging studies had a discrepancy rate of 2.0%. Our quality assessment study could serve as initial data before intervention as part of a PQI project.ABBREVIATIONS: ABR ϭ American Board of Radiology; ACGME ϭ Accreditation Council for Graduate Medical Education; CI ϭ confidence interval; PQI ϭ practice quality improvement R adiologic detection and interpretation errors will not be fully eliminated until the advent of "perfect diagnostic tests" and "perfect observers." 1,2 In the meantime, radiologists, similar to other physicians, struggle with assessing physician performance and reporting quality, to improve and deliver the best care possible.3 L. Henry Garland pioneered the work on radiologic errors more than 60 years ago. [4][5][6] He uncovered a 30% rate of missed radiologic findings in a series of radiographs with abnormal findings among expert reviewers. Subsequently Garland's results have been replicated by other researchers. [7][8][9] Most interesting, comparable rates of "mistakes" were discovered in other specialties. 10,11 In deriving the radiologic error rate, Garland used exclusively abnormal studies-that is, he tested radiologists in environments in which disease prevalence reached 100%. Because in the typical clinical setting, there are a substantial number of examinations with normal findings, Garland hypothesized that the expected radiologic error rate in everyday practice is closer to 5%.Subsequent studies confirmed the radiologic error rate in all-comers radiology practice to be in the 3%-6% range. 3,12-14Soffa et al 14 sampled approximately 7000 cases read by 26 radiologists and uncovered a 3% disagreement rate in general radiology, 3.6% in diagnostic mammography, 5.8% in screening mammography, and 4.1% in sonography, yielding the overall error rate of 3.5%. Robinson et al 13 compared reports for skeletal, chest, and abdominal radiographs completed by 3 radiologists and found ...

show abstract

Section: Discussionmentioning

confidence: 94%

Quality Control in Neuroradiology: Discrepancies in Image Interpretation among Academic Neuroradiologists

Babiarz¹,

Yousem²

2011

AJNR Am J Neuroradiol

View full text Add to dashboard Cite

show abstract

“…When using new KNN, the classification performance level of the interactive CAD system is significantly increased (p < 0.001) to AUC 00.897 with the corresponding 95% CI of [0.887, 0.907]. (1) to the bottom right ROI (12), the reference image set of CC view includes three ROIs (6, 9, and 10) depicting malignant masses, three ROIs (3, 5, and 11) depicting benign masses, and six ROIs (1, 2, 4, 7, 8, and 12) depicting CAD-cued falsepositive regions, while the reference image set of MLO view includes five malignant ROIs (3,5,6,9, and 12), one benign ROI (2), and six CAD-cued false-positive ROIs (1,4,7,8,10, and 11) Fig. 4 Comparison of two ROC-type performance curves of our interactive CAD scheme evaluated using a leave-one-out testing method involving 3,600 ROIs in the reference database.…”

Section: Resultsmentioning

confidence: 99%

“…Although a number of screening tools such as breast magnetic resonance imaging [3,4] have been approved and used as adjunct screening tools for the highrisk women identified by epidemiology-based risk models [5], periodic screening with mammography is the only clinically accepted imaging modality for screening the general population to date. However, interpreting mammograms by radiologists is difficult and time-consuming due to the low cancer prevalence in the screening environment (i.e., less than three to five cancers detected per every 1,000 screening examinations), as well as a large variability of depicted breast abnormalities and overlapped dense fibro-glandular tissue on the two-dimensional projected images [6,7]. As a result, both detection sensitivity and specificity of screening mammography are not satisfactory [8] particularly in younger women with dense breasts and in other high-risk groups.…”

Section: Introductionmentioning

confidence: 99%

An Interactive System for Computer-Aided Diagnosis of Breast Masses

Wang

Liu

et al. 2012

J Digit Imaging

View full text Add to dashboard Cite

Although mammography is the only clinically accepted imaging modality for screening the general population to detect breast cancer, interpreting mammograms is difficult with lower sensitivity and specificity. To provide radiologists "a visual aid" in interpreting mammograms, we developed and tested an interactive system for computeraided detection and diagnosis (CAD) of mass-like cancers. Using this system, an observer can view CAD-cued mass regions depicted on one image and then query any suspicious regions (either cued or not cued by CAD). CAD scheme automatically segments the suspicious region or accepts manually defined region and computes a set of image features. Using content-based image retrieval (CBIR) algorithm, CAD searches for a set of reference images depicting "abnormalities" similar to the queried region. Based on image retrieval results and a decision algorithm, a classification score is assigned to the queried region. In this study, a reference database with 1,800 malignant mass regions and 1,800 benign and CAD-generated false-positive regions was used. A modified CBIR algorithm with a new function of stretching the attributes in the multi-dimensional space and decision scheme was optimized using a genetic algorithm. Using a leave-one-out testing method to classify suspicious mass regions, we compared the classification performance using two CBIR algorithms with either equally weighted or optimally stretched attributes. Using the modified CBIR algorithm, the area under receiver operating characteristic curve was significantly increased from 0.865 ±0.006 to 0.897±0.005 (p<0.001). This study demonstrated the feasibility of developing an interactive CAD system with a large reference database and achieving improved performance.

show abstract

“…However, for screening mammography, radiologists' experience and annual volume have been studied. In general, radiologist training has been associated with increased performance, [12][13][14][15][16][17] but the literature is conflicting on the relationship between annual volume per radiologist and performance. Our study support findings from five studies showing volume to be of importance.…”

Section: Discussionmentioning

confidence: 99%

“…Radiologist training in general has been associated with increased performance. [12][13][14][15][16][17] Data on the relationship between volume read per radiologist and performance are, however, conflicting, 13,[18][19][20][21][22] whereas associations have been found between some other indicators of radiologists' experience and performance. [23][24][25] However, most studies focused on mammography screening, and no large community based study has evaluated the association between organisational factors and performance of diagnostic mammography.…”

mentioning

confidence: 99%

Performance of clinical mammography: A nationwide study from Denmark

Jensen

Vejborg

Severinsen

et al. 2006

Intl Journal of Cancer

View full text Add to dashboard Cite

Clinical mammography is the key tool for breast cancer diagnosis, but little is known about the impact of the organisational set-up on the performance. We evaluated whether organisatorial factors influence the performance of clinical mammography. Clinical mammography data from all clinics in Denmark in the year 2000 were collected and linked to cancer outcome. Use of the National Institute of Radiation Hygiene register for identification of radiology clinics ensured comprehensive nationwide registration. We used the final mammographic assessment at the end of the imaging work-up to determine sensitivity, specificity and accuracy, the latter using a receiver-operating characteristics (ROC) analysis. In 96,534 clinical mammography examinations, sensitivity was 75% and specificity 99%. The presence of at least one high volumereading radiologist in the clinic increased accuracy (AUC 5 0.91 for <1,000 examinations/year and 0.92 for >2,000 examinations/ year, p 5 0.017). The examination volume per clinic showed no clear effect on performance, as accuracy was significantly higher in clinics with a medium number of examinations (AUC 5 0.93 for 2,000-4,000 examinations/year and 0.90 for >6,000 examinations/year, p 5 0.003). Accuracy was significantly lower in regions with high annual utilisation rate of clinical mammography, which means the proportion of examined women in a region (AUC 5 0.90 for 3.0-5.0% annual utilisation rate and AUC 5 0.93 for 2.0-2.5% annual utilisation rate, and p 5 0.001), indicating that clinical mammography worked best in patient populations of purely symptomatic women. Our data indicate that to increase the accuracy of clinical mammography at the community level, the presence of an experienced radiologist should be prioritised ahead of raising the clinic size. ' 2006 Wiley-Liss, Inc.Key words: mammography; breast cancer; organization; sensitivity; specificity; accuracy Diagnostic mammography is the most commonly used tool for diagnosing breast cancers in women with signs or symptoms of this disease. It is most often undertaken in combination with a clinical examination and supplemented with ultrasound or other imaging techniques. 1 In contrast to mammography screening, performance of diagnostic mammography has been examined only in a few studies. In addition, the majority of these studies are singleclinic studies covering small populations, 1-5 and only two studies have evaluated the performance of diagnostic mammography at the community level. 6,7 Performance of mammography is known to be influenced by factors related to woman characteristics, such as age, breast density, history of previous breast lesions and presence of symptoms. 6,[8][9][10][11] However, performance of mammography may also be related to the organisational setup, such as the type of clinic, the number of examinations undertaken and the experience of the radiologists. Radiologist training in general has been associated with increased performance. 12-17 Data on the relationship between volume read per radiologist and performan...

show abstract

Performance Parameters for Screening and Diagnostic Mammography: Specialist and General Radiologists

Abstract: Specialist radiologists detect more cancers and more early-stage cancers, recommend more biopsies, and have lower recall rates than general radiologists.

Cited by 225 publications

References 20 publications

Quality Control in Neuroradiology: Discrepancies in Image Interpretation among Academic Neuroradiologists

Quality Control in Neuroradiology: Discrepancies in Image Interpretation among Academic Neuroradiologists

An Interactive System for Computer-Aided Diagnosis of Breast Masses

Performance of clinical mammography: A nationwide study from Denmark

Contact Info

Product

Resources

About