Clinically Missed Cancer: How Effectively Can Radiologists Use Computer-Aided Detection?

Nishikawa, Robert M.; Schmidt, Robert A.; Linver, Michael N.; Edwards, Alexandra; Papaioannou, John; Stull, Margaret A.

doi:10.2214/ajr.11.6423

Cited by 44 publications

(29 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Cases were first grouped into BI-RADS categories [1–5], then randomly selected from within each category to mirror the distribution of BI-RADS assessments seen in the DMIST study. The cases were not consecutively chosen.…”

Section: Methodsmentioning

confidence: 99%

Impact of Computer-Aided Detection Systems on Radiologist Accuracy With Digital Mammography

Cole

Zhang²,

Marques³

et al. 2014

American Journal of Roentgenology

View full text Add to dashboard Cite

OBJECTIVE The purpose of this study was to assess the impact of computer-aided detection (CAD) systems on the performance of radiologists with digital mammograms acquired during the Digital Mammographic Imaging Screening Trial (DMIST). MATERIALS AND METHODS Only those DMIST cases with proven cancer status by biopsy or 1-year follow-up that had available digital images were included in this multireader, multicase ROC study. Two commercially available CAD systems for digital mammography were used: iCAD SecondLook, version 1.4; and R2 ImageChecker Cenova, version 1.0. Fourteen radiologists interpreted, without and with CAD, a set of 300 cases (150 cancer, 150 benign or normal) on the iCAD SecondLook system, and 15 radiologists interpreted a different set of 300 cases (150 cancer, 150 benign or normal) on the R2 ImageChecker Cenova system. RESULTS The average AUC was 0.71 (95% CI, 0.66–0.76) without and 0.72 (95% CI, 0.67–0.77) with the iCAD system (p = 0.07). Similarly, the average AUC was 0.71 (95% CI, 0.66–0.76) without and 0.72 (95% CI 0.67–0.77) with the R2 system (p = 0.08). Sensitivity and specificity differences without and with CAD for both systems also were not significant. CONCLUSION Radiologists in our studies rarely changed their diagnostic decisions after the addition of CAD. The application of CAD had no statistically significant effect on radiologist AUC, sensitivity, or specificity performance with digital mammograms from DMIST.

show abstract

Section: Methodsmentioning

confidence: 99%

Impact of Computer-Aided Detection Systems on Radiologist Accuracy With Digital Mammography

Cole

Zhang²,

Marques³

et al. 2014

American Journal of Roentgenology

View full text Add to dashboard Cite

show abstract

“…An error prediction algorithm based on eye-tracking data and image features analysis improved radiologists' decisionmaking process by 25% [21]. This novel computer aided detection (CAD) concept was an inspiration for further investigation [22] and might provide a solution to current CAD system, which can increase radiology sensitivity 10%, but with a comparable deterioration in specicity [23]. Pietrzyk et al (2012) proposed a method to tackle the problem of specicity reduction with CAD system [21].…”

Section: Eye-trackingmentioning

confidence: 99%

Visual Perception in Nuclear Medicine and Volumetric Data

Pietrzyk

2015

Acta Phys. Pol. A

View full text Add to dashboard Cite

Any diagnostic decision that involves imaging techniques depends not entirely on the images themselves but also on the perception and interpretation by a radiologist. In fact, such human factors appear to be a bottleneck to imaging accuracy. A better understanding of visual perception and decision-making processes seem to be necessary for the advancement and optimization of medical imaging. Therefore, research on the accuracy of medical imaging draws on techniques from a wide range of disciplines in an attempt to minimize diagnostic errors and improve the healthcare system. This short review of the psychophysical studies in the radiological domain focuses on the recent work conducted in the visual perception of nuclear medicine images and volumetric data.

show abstract

“…Finally, computer-aided diagnostic techniques, yet used in some screening programs [151,152] , could be added to standard EUS images for the differentiation of pancreatic carcinoma from chronic pancreatitis [151,153] . With digital image processing and computer-aided EUS image differentiation technologies, physicians could use the computer output as a ''second opinion'' and make the final decisions as reported by the high diagnostic accuracy (98%) of a recent study [154] .…”

Section: Future Perspectivesmentioning

confidence: 99%

Endoscopic ultrasonography for surveillance of individuals at high risk for pancreatic cancer

Lami

Biagini

Galli

2014

WJGE

View full text Add to dashboard Cite

Pancreatic cancer is a highly lethal disease with a genetic susceptibility and familial aggregation found in 3%-16% of patients. Early diagnosis remains the only hope for curative treatment and improvement of prognosis. This can be reached by the implementation of an intensive screening program, actually recommended for individuals at high-risk for pancreatic cancer development. The aim of this strategy is to identify premalignant precursors or asymptomatic pancreatic cancer lesions, curable by surgery. Endoscopic ultrasound (EUS) with or without fine needle aspiration (FNA) seems to be the most promising technique for early detection of pancreatic cancer. It has been described as a highly sensitive and accurate tool, especially for small and cystic lesions. Pancreatic intraepithelial neoplasia, a precursor lesion which is highly represented in highrisk individuals, seems to have characteristics chronic pancreatitis-like changes well detected by EUS. Many screening protocols have demonstrated high diagnostic yields for pancreatic pre-malignant lesions, allowing prophylactic pancreatectomies. However, it shows a high interobserver variety even among experienced endosonographers and a low sensitivity in case of chronic pancreatitis. Some new techniques such as contrast-enhanced harmonic EUS, computer-aided diagnostic techniques, confocal laser endomicroscopy miniprobe and the detection of DNA abnormalities or protein markers by FNA, promise improvement of the diagnostic yield of EUS. As the resolution of imaging improves and as our knowledge of precursor lesions grows, we believe that EUS could become the most suitable method to detect curable pancreatic neoplasms in correctly identified asymptomatic at-risk patients.© 2014 Baishideng Publishing Group Inc. All rights reserved. Key words: Endoscopic ultrasonography; Pancreatic cancer; SurveillanceCore tip: In the era of early diagnosis and screening programs, endoscopic ultrasound (EUS) represents the most promising tool able to identify pancreatic precursor neoplasms in high risk individuals. If compared to other imaging techniques, it is highly accurate to diagnose small pancreatic cancer and pre-malignant lesions, with very low rate of complications and limitations. Here are reported the current role of EUS in various international screening programs and its future possible developments.

show abstract

Clinically Missed Cancer: How Effectively Can Radiologists Use Computer-Aided Detection?

Abstract: Use of CADe can increase radiologist sensitivity 10% with a comparable increase in recall rate. There is potential for CADe to have a bigger clinical impact because radiologists failed to recognize a correct computer prompt in 71% of missed cancer cases [corrected].

Cited by 44 publications

References 47 publications

Impact of Computer-Aided Detection Systems on Radiologist Accuracy With Digital Mammography

Impact of Computer-Aided Detection Systems on Radiologist Accuracy With Digital Mammography

Visual Perception in Nuclear Medicine and Volumetric Data

Endoscopic ultrasonography for surveillance of individuals at high risk for pancreatic cancer

Contact Info

Product

Resources

About