scholarly journals Classification of Hematoxylin and Eosin-Stained Breast Cancer Histology Microscopy Images Using Transfer Learning with EfficientNets

2021 ◽  
Vol 2021 ◽  
pp. 1-17
Author(s):  
Chanaleä Munien ◽  
Serestina Viriri
Keyword(s):  
Breast Cancer ◽  
Human Error ◽  
Final Diagnosis ◽  
Normalization Method ◽  
Fine Tuning ◽  
In Situ Carcinoma ◽  
Hematoxylin And Eosin ◽  
The Cost ◽  
The Impact

Breast cancer is a fatal disease and is a leading cause of death in women worldwide. The process of diagnosis based on biopsy tissue is nontrivial, time-consuming, and prone to human error, and there may be conflict about the final diagnosis due to interobserver variability. Computer-aided diagnosis systems have been designed and implemented to combat these issues. These systems contribute significantly to increasing the efficiency and accuracy and reducing the cost of diagnosis. Moreover, these systems must perform better so that their determined diagnosis can be more reliable. This research investigates the application of the EfficientNet architecture for the classification of hematoxylin and eosin-stained breast cancer histology images provided by the ICIAR2018 dataset. Specifically, seven EfficientNets were fine-tuned and evaluated on their ability to classify images into four classes: normal, benign, in situ carcinoma, and invasive carcinoma. Moreover, two standard stain normalization techniques, Reinhard and Macenko, were observed to measure the impact of stain normalization on performance. The outcome of this approach reveals that the EfficientNet-B2 model yielded an accuracy and sensitivity of 98.33% using Reinhard stain normalization method on the training images and an accuracy and sensitivity of 96.67% using the Macenko stain normalization method. These satisfactory results indicate that transferring generic features from natural images to medical images through fine-tuning on EfficientNets can achieve satisfactory results.

scholarly journals Impact of 2013 ASCO/CAP guidelines on various HER2 reporting categories in breast cancer by fluorescent in-situ hybridization and Immunohistochemistry: A meta-analysis with systematic review

2020 ◽  
Vol 5 ◽  
pp. 14-26 ◽  
Author(s):  
Sunil Pasricha ◽  
Smita Asthana ◽  
Satyanarayana Labani ◽  
Uma Kailash ◽  
Abhinav Srivastav ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Systematic Review ◽  
Meta Analysis ◽  
Pooled Analysis ◽  
Paired Data ◽  
Treatment Benefit ◽  
Before And After ◽  
Primary Test ◽  
The Cost ◽  
The Impact

Objective: The ASCO/CAP guidelines for reporting HER2 in breast cancer, first released in 2007, aimed to standardize the reporting protocol, and were updated in 2013 and 2018, to ensure right treatment. Several studies have analyzed the changes attributed to 2013 updated guidelines, and majority of them found increase in positive and equivocal cases. However, the precise implication of these updated guidelines is still contentious, in spite of the latest update (2018 guidelines) addressing some of the issues. We conducted systematic review and meta- analysis to see the impact of 2013 guidelines on various HER2 reporting categories by both FISH and IHC. Materials and Methods: After extensively searching the pertinent literature, 16 studies were included for the systematic review. We divided our approach in three strategies: (1) Studies in which breast cancer cases were scored for HER2 by FISH or IHC as a primary test concurrently by both 2007 and 2013 guidelines, (2) Studies in which HER2 results were equivocal by IHC and were followed by reflex-FISH test by both 2007 and 2013 guidelines, and (3) Studies in which trends of HER2 reporting were compared in the two periods before and after implementation of updated 2013 guidelines. All the paired data in these respective categories was pooled and analyzed statistically to see the overall impact of the updated guidelines. Results: In the first category, by pooled analysis of primary FISH testing there has been a significant increase in the equivocal cases (P < 0.001) and positive cases (P = 0.037). We also found 8.3% and 0.8% of all the negative cases from 2007 guidelines shifted to equivocal and positive categories, respectively. Similarly by primary IHC testing there has been a significant increase in both equivocal cases (P < 0.001) and positive cases (P = 0.02). In the second category of reflex-FISH testing there was a substantial increase in the equivocal cases (P < 0.0001); however there is insignificant decrease (10% to 9.7%; P = 0.66) in the amplified cases. In the third approach for evaluating the trend, with the implementation of 2013 guidelines, there was increase in the equivocal category (P = 0.025) and positive category (P = 0.0088) by IHC. By FISH test also there was significant increase in the equivocal category (P < 0.001) while the increase in the positive category was non-significant (P = 0.159). Conclusions: The updated 2013 guidelines has significantly increased the positive and equivocal cases using primary FISH or IHC test and with further reflex testing, thereby increasing the double equivocal cases and increasing the cost and delaying the decision for definite management. However, whether the additional patients becoming eligible for HDT will derive treatment benefit needs to be answered by further large clinical trials.

Impact of Advancing Technology on Nuclear Facility Operation

2019 ◽  
Vol 6 (1) ◽  
Author(s):  
Jonathan K. Corrado ◽  
Ronald M. Sega
Keyword(s):  
Human Error ◽  
New Technologies ◽  
Nuclear Industry ◽  
Nuclear Facilities ◽  
Event Scale ◽  
Technological Advances ◽  
Significant Difference ◽  
The Cost ◽  
The Impact ◽  
The U.S

Abstract Many unfortunate and unintended adverse industrial incidents occur across the U.S. each year, and the nuclear industry is no exception. Depending on the severity, these incidents can be problematic for people, the facilities, and surrounding environments. These incidents occur for a number of varying reasons, but more often than not, human error is an accomplice. This article explores whether the complexity and changing technologies, which affect the way operators interact within the systems of the nuclear facilities, exacerbate the severity of incidents caused by human error. A review of nuclear incidents in the U.S. from 1955 to 2010 reaching level three or higher on the International Nuclear Event Scale (INES) scale was conducted. The cost of each incident at facilities that had recently undergone technological changes affecting plant operator's jobs was compared to those facilities which had not undergone changes. A t-test was applied and determined a statistically significant difference between the two groups. This affirmed that technological advances at nuclear facilities that affect how operators interact within the plant system increase the severity of resulting incidents. Next, a follow-on study was conducted to determine the impact from the incorporation of new technologies into nuclear facilities. The data indicated that spending more money on upgrades increased the capacity of the facility as well as the number of incidents reported, but the incident severity was minor.

Computing the cost of care per day of breast cancer survivor care.

2018 ◽  
Vol 36 (7_suppl) ◽  
pp. 10-10 ◽  
Author(s):  
Douglas W. Blayney ◽  
Craig Lindquist ◽  
Tina Seto ◽  
Nhat Minh Hoang ◽  
Allison W. Kurian
Keyword(s):  
Breast Cancer ◽  
Cancer Survivor ◽  
Breast Cancer Survivor ◽  
Radiation Treatment ◽  
Cost Of Care ◽  
Survivorship Care ◽  
Co Morbidity ◽  
Social Security Death Index ◽  
The Cost ◽  
The Impact

10 Background: Cost of breast cancer survivor surveillance, of survivorship care and of variation in care practices are unknown. Furthermore, it is not known whether intense surveillance care adds value. We developed a method to measure the cost of surveillance to account for varying follow-up duration [cost of care per day (CCPD)], and explored the impact of surveillance cost on survival. Methods: We queried the Oncoshare database {Kurian et al Cancer 2014}, which amalgamates data from Stanford Health Care’s (SHC) electronic health record (EHR) (imaging, infused drugs, inpatient and outpatient facility and professional services), from the California Cancer Registry, and the Social Security Death Index. We included breast cancer patients diagnosed 2000-2014, Stages 0-III who had surgery, chemotherapy or radiation treatment at SHC, and who had more than two visits at SHC within 3 years of their treatment completion. We tallied Common Procedural Terminology (CPT©) codes assigned to each service, and mapped each CPT code to the corresponding code and date in the CMS Medicare fee schedule. For patients with breast cancer relapse, we explored the post-relapse survival of the costliest 20% compared with the other patients. Results: CCPD was $2.45 for care delivered at SHC. Among the three breast cancer subtypes (luminal, Her-2 over-expressed and triple negative) there was no difference in cost. Among patients who relapsed, those in the most expensive 20% CCPD had significantly shorter survival than other patients. The high-cost patients had more co-morbidity [cerebrovascular disease (4% for low cost vs 7% for high), chronic pulmonary disease (5% vs 10%), CHF (2% vs 7%), diabetes (4% vs 7%), liver disease (4% vs 9%)]. Conclusions: Cost of care per day (CCPD) is a useful metric to assess value of surveillance and survivorship care, and is also applicable to initial treatment and post-relapse care, to identify “positive deviants” ( those who have developed best practices) in high value care delivery. We captured only costs for treatment at SHC, and merging our data with claims data from 3rd party carriers could increase the accuracy and validity of the CCPD. We identified a model for further testing to reduce total spending for high-quality oncology care.

scholarly journals Cost-Effectiveness of MR-Mammography in Breast Cancer Screening of Women With Extremely Dense Breasts After Two Rounds of Screening

2021 ◽  
Vol 11 ◽  
Author(s):  
Fabian Tollens ◽  
Pascal A. T. Baltzer ◽  
Matthias Dietzel ◽  
Moritz L. Schnitzer ◽  
Wolfgang G. Kunz ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Cost Effectiveness ◽  
Cancer Screening ◽  
Breast Cancer Screening ◽  
Sensitivity Analyses ◽  
Intermediate Risk ◽  
Mr Mammography ◽  
Markov Modelling ◽  
The Cost ◽  
The Impact

ObjectivesTo evaluate the cost-effectiveness of MR-mammography (MRM) vs. x-ray based mammography (XM) in two-yearly screening women of intermediate risk for breast cancer in the light of recent literature.MethodsDecision analysis and Markov modelling were used to compare cumulative costs (in US-$) and outcomes (in QALYs) of MRM vs. XM over the model runtime of 20 years. The perspective of the U.S. healthcare system was selected. Incremental cost-effectiveness ratios (ICER) were calculated and related to a willingness to pay-threshold of $ 100,000 per QALY in order to evaluate the cost-effectiveness. Deterministic and probabilistic sensitivity analyses were conducted to test the impact of variations of the input parameters. In particular, variations of the rate of false positive findings beyond the first screening round and their impact on cost-effectiveness were assessed.ResultsBreast cancer screening with MRM resulted in increased costs and superior effectiveness. Cumulative average costs of $ 6,081 per woman and cumulative effects of 15.12 QALYs were determined for MRM, whereas screening with XM resulted in costs of $ 5,810 and 15.10 QALYs, resulting in an ICER of $ 13,493 per QALY gained. When the specificity of MRM in the second and subsequent screening rounds was varied from 92% to 99%, the ICER resulted in a range from $ 38,849 to $ 5,062 per QALY.ConclusionsBased on most recent data on the diagnostic performance beyond the first screening round, MRM may remain the economically preferable alternative in screening women of intermediate risk for breast cancer due to their dense breast tissue.

scholarly journals Cost-effectiveness analysis of radiotherapy techniques for whole breast irradiation

PLoS ONE ◽  
2021 ◽  
Vol 16 (3) ◽  
pp. e0248220
Author(s):  
Yibo Xie ◽  
Beibei Guo ◽  
Rui Zhang
Keyword(s):  
Breast Cancer ◽  
Cost Effectiveness ◽  
Cost Effective ◽  
Advanced Technology ◽  
Sensitivity Analyses ◽  
Tumor Control ◽  
Life Years ◽  
Whole Breast Radiotherapy ◽  
The Cost ◽  
The Impact

Background The current standard of care (SOC) for whole breast radiotherapy (WBRT) in the US is conventional tangential photon fields. Advanced WBRT techniques may provide similar tumor control and better normal tissue sparing, but it is controversial whether the medical benefits of an advanced technology are significant enough to justify its higher cost. Objective To analyze the cost-effectiveness of six advanced WBRT techniques compared with SOC. Methods We developed a Markov model to simulate health states for one cohort of women (65-year-old) with early-stage breast cancer over 15 years after WBRT. The cost effectiveness analyses of field-in-field (FIF), hybrid intensity modulated radiotherapy (IMRT), full IMRT, standard volumetric modulated arc therapy (STD-VMAT), multiple arc VMAT (MA-VMAT), non-coplanar VMAT (NC-VMAT) compared with SOC were performed with both tumor control and radiogenic side effects considered. Transition probabilities and utilities for each health state were obtained from literature. Costs incurred by payers were adopted from literature and Medicare data. Quality-adjusted life years (QALYs) and incremental cost-effectiveness ratio (ICER) were calculated. One-way sensitivity analyses and probabilistic sensitivity analyses (PSA) were performed to evaluate the impact of uncertainties on the final results. Results FIF has the lowest ICER value of 1,511 $/QALY. The one-way analyses show that the cost-effectiveness of advanced WBRT techniques is most sensitive to the probability of developing contralateral breast cancer. PSAs show that SOC is more cost effective than almost all advanced WBRT techniques at a willingness-to-pay (WTP) threshold of 50,000 $/QALY, while FIF, hybrid IMRT and MA-VMAT are more cost-effective than SOC with a probability of 59.2%, 72.3% and 72.6% at a WTP threshold of 100,000 $/QALY, respectively. Conclusions FIF might be the most cost-effective option for WBRT patients at a WTP threshold of 50,000 $/QALY, while hybrid IMRT and MA-VMAT might be the most cost-effective options at a WTP threshold of 100,000 $/QALY.

Sign in / Sign up

Export Citation Format

Share Document