Learning better deep features for the prediction of occult invasive disease in ductal carcinoma in situ through transfer learning

2018 ◽  
Author(s):  
Yinhao Ren ◽  
Bibo Shi ◽  
Rui Hou ◽  
Lars J. Grimm ◽  
Maciej A. Mazurowski ◽  
...  
Keyword(s):  
Transfer Learning ◽  
Ductal Carcinoma In Situ ◽  
Carcinoma In Situ ◽  
Ductal Carcinoma ◽  
Invasive Disease

Genomic Differences Between Pure Ductal Carcinoma In Situ of the Breast and that Associated with Invasive Disease: a Calibrated aCGH Study

2008 ◽  
Vol 14 (14) ◽  
pp. 4446-4454 ◽  
Author(s):  
Vladimir V. Iakovlev ◽  
Nona C.R. Arneson ◽  
Vietty Wong ◽  
Chunjie Wang ◽  
Stephanie Leung ◽  
...  
Keyword(s):  
Ductal Carcinoma In Situ ◽  
Carcinoma In Situ ◽  
Ductal Carcinoma ◽  
Invasive Disease

scholarly journals The influence of clinicopathological features on the predictive accuracy of conventional breast imaging in determining the extent of screen-detected high-grade pure ductal carcinoma in situ

2011 ◽  
Vol 93 (5) ◽  
pp. 385-390 ◽  
Author(s):  
L Hayward ◽  
RS Oeppen ◽  
AV Grima ◽  
GT Royle ◽  
CM Rubin ◽  
...  
Keyword(s):  
Ductal Carcinoma In Situ ◽  
Carcinoma In Situ ◽  
Breast Imaging ◽  
Predictive Accuracy ◽  
Ductal Carcinoma ◽  
Preoperative Staging ◽  
Invasive Disease ◽  
High Grade ◽  
Excision Rate

INTRODUCTION The extent of calcified ductal carcinoma in situ (DCIS) detected by screening mammography is a determinant for treatment with breast conserving surgery (BCS). However, DCIS may be uncalcified and almost a quarter of patients with DCIS treated initially by BCS either require a second operation or are found to have unexpected invasive disease following surgery. Identification of these cases might guide selective implementation of additional diagnostic procedures. METHODS A retrospective review of patients with a preoperative diagnosis of pure high-grade DCIS at the Southampton and Salisbury Breast Screening Unit over a ten-year period was carried out. Mammograms were reviewed independently by a consultant radiologist and additional factors including the Breast Imaging Reporting and Data System (BI-RADS®) breast density score, DCIS extent and disease location within the breast recorded. RESULTS Unexpected invasive disease was found in 35 of 144 patients (24%). Within our unit the re-excision rate for all screen-detected DCIS is currently 23% but for patients included in this study with high-grade DCIS the re-excision rate was 39% (34/87). The extent of DCIS (p=0.008) and lack of expression of the oestrogen receptor (ER) predicted the requirement for re-excision in both univariate (p=0.004) and multivariate analysis (p=0.005). CONCLUSIONS High-grade DCIS may be focally uncalcified, leading to underestimation of disease extent, which might be related to ER status. Invasive foci associated with high-grade DCIS are often mammographically occult. Exploration of additional biomarkers and targeted use of further diagnostic techniques may improve the preoperative staging of DCIS.

Ability of Intraoperative Pathologic Analysis of Ductal Carcinoma In Situ to Guide Selective Use of Sentinel Lymph Node Surgery

2018 ◽  
Vol 84 (4) ◽  
pp. 537-542 ◽  
Author(s):  
Brittany L. Murphy ◽  
Alexandra B. Gonzalez ◽  
Michael G. Keeney ◽  
Beiyun Chen ◽  
Amy L. Conners ◽  
...  
Keyword(s):  
Lymph Node ◽  
Sentinel Lymph Node ◽  
Ductal Carcinoma In Situ ◽  
Carcinoma In Situ ◽  
Ductal Carcinoma ◽  
Invasive Disease ◽  
Final Pathology ◽  
Selective Approach ◽  
Pathologic Analysis

For patients with ductal carcinoma in situ (DCIS), sentinel lymph node (SLN) surgery is generally reserved for patients at high risk of being upstaged to invasive disease. The use of frozen section (FS) pathologic analysis of the primary tumor may allow for selective surgical nodal staging within one procedure. We sought to define the reliability of FS for detection of upstaging. Eight hundred and twenty-seven patients were identified with DCIS on core needle biopsy that underwent 834 operations at our institution between January 2004 and October 2014. We calculated the rate of upstage from DCIS to invasive cancer on both intraoperative FS and final pathology to determine the performance of FS. Upstage rate on final pathology was 118/834 (14.1%) 95 per cent confidence interval 11.8 to 16.7 per cent. FS identified 88/118 (74.6%) of the upstages. Specificity was 99.3 per cent (711/716). Overall accuracy was 95.8 per cent (799/834) and the positive predictive value was 96.0 per cent (711/741 patients). Mean size of invasive cancers identifiedon FS was 5.6 mm, versus 3.5 mm for those identified only on permanent section, P = 0.11. Intraoperative FS analysis of DCIS is useful for identification of upstage to invasive disease. This may facilitate a selective approach to SLN surgery that both decreases unnecessary SLN surgery and the need for a second operation.

Prediction of occult invasive disease in ductal carcinoma in situ using computer-extracted mammographic features

2017 ◽  
Author(s):  
Bibo Shi ◽  
Lars J. Grimm ◽  
Maciej A. Mazurowski ◽  
Jeffrey R. Marks ◽  
Lorraine M. King ◽  
...  
Keyword(s):  
Ductal Carcinoma In Situ ◽  
Carcinoma In Situ ◽  
Ductal Carcinoma ◽  
Invasive Disease
Sign in / Sign up

Export Citation Format

Share Document