Breast cancer risk prediction model: a nomogram based on common mammographic screening findings

A nomogram prognostic model for early hepatocellular carcinoma with diabetes mellitus after primary liver resection based on the admission characteristics

Background: Patients diagnosed with early-stage hepatocellular carcinoma (HCC) and diabetes mellitus (DM) are at a higher risk of experiencing complications and facing increased mortality rates. Hence, it is crucial to develop personalized clinical strategies for this particular subgroup upon their admission. The objective of this study is to determine the key prognostic factors in early HCC patients who received liver resection combined with DM and develop a practical personalized model for precise prediction of overall survival in these individuals. Method: A total of 1496 patients diagnosed hepatitis B virus (HBV) - related liver cancer from Beijing You'an Hospital were retrospectively enrolled, spanning from 1 January 2014, to 31 December 2019, and ultimately, 622 eligible patients of hepatocellular carcinoma (HCC) patients with diabetes were included in this present investigation. A multivariate COX regression analysis was conducted to identify prognostic factors that are independent of each other and develop a nomogram. The performance of the nomogram was evaluated using various statistical measures such as the C-index, receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) in both the training and validation groups. Survival rates were estimated using the Kaplan-Meier method. Results: The study included a total of 622 early HCC patients who underwent liver resection combined with DM. Random Forrest model and Multivariate Cox regression analysis revealed that drinking, tumor number, monocyte-to-lymphocyte ratio, white blood cell count and international normalized ratio at admission were identified as independent prognostic factors for early HCC patients who underwent liver resection combined with DM. The nomogram demonstrated good predictive performance in the training and validation cohorts based on the C-index values of 0 .756 and 0 .739 respectively, as well as the area under the curve values for 3-, 5-, and 8-year overall survival (0.797, 0.807, 0.840, and 0.725, 0.791, 0.855). Calibration curves and decision curve analysis indicated high accuracy and net clinical benefit rates. Furthermore, the nomogram successfully stratified enrolled patients into low-risk and high-risk groups based on their risk of overall survival. The difference in overall survival between these two groups was statistically significant in both the training and validation cohorts (p < 0.0001 and p = 0.0064). Conclusion: Our results indicate that the admission characteristics demonstrate a highly effective ability to predict the overall survival of early HCC patients who have undergone liver resection in combination with DM. The developed model has the potential to support healthcare professionals in making more informed initial clinical judgments for this particular subgroup of patients.

The dilemma of recalling well-circumscribed masses in a screening population: A narrative literature review and exploration of Dutch screening practice

BACKGROUND

In Dutch breast cancer screening, solitary, new or growing well-circumscribed masses should be recalled for further assessment. This results in cancers detected but also in false positive recalls, especially at initial screening. The aim of this study was to determine characteristics of well-circumscribed masses at mammography and identify potential methods to improve the recall strategy.

METHODS

A systematic literature search was performed using PubMed. In addition, follow-up data were retrieved on all 8860 recalled women in a Dutch screening region from 2014 to 2019.

RESULTS

Based on 15 articles identified in the literature search, we found that probably benign well-circumscribed masses that were kept under surveillance had a positive predictive value (PPV) of 0-2%. New or enlarging solitary well-circumscribed masses had a PPV of 10-12%. In general the detected carcinomas had a favorable prognosis. In our exploration of screening practice, 25% of recalls (2133/8860) were triggered by a well-circumscribed mass. Those recalls had a PPV of 2.0% for initial and 10.6% for subsequent screening. Most detected carcinomas had a favorable prognosis as well.

CONCLUSION

To recognize malignancies presenting as well-circumscribed masses, identifying solitary, new or growing lesions is key. This information is missing at initial screening since prior examinations are not available, leading to a low PPV. Access to prior clinical examinations may therefore improve this PPV. In addition, given the generally favorable prognosis of screen-detected malignant well-circumscribed masses, one may opt to recall these lesions at subsequent screening, if grown, rather than at initial screening.

From past to future: Bibliometric analysis of global research productivity on nomogram (2000-2021)

Background

Nomogram, a visual clinical predictive model, provides a scientific basis for clinical decision making. Herein, we investigated 20 years of nomogram research responses, focusing on current and future trends and analytical challenges.

Methods

We mined data of scientific literature from the Core Collection of Web of Science, searching for the original articles with title "Nomogram^*/Parton Table^*/Parton Nomogram^*", published within January 1st, 2000 to December 30th, 2021. Data records were validated using HistCite Version and analyzed with a transformable statistical method, the Bibliometrix 3.0 package of R Studio.

Results

In total, 4,176 original articles written by 19,158 authors were included from 915 sources. Annually, Nomogram publications are continually produced, which have rapidly grown since 2018. China published the most articles; however, its total citations ranked second after the United States. Both total citations and average article citations in the United States rank first globally, and a high degree of cooperation exists between countries. Frontiers in Oncology published the most papers (238); this number has grown rapidly since 2019. Journal of Urology had the highest H-index, with an average increase in publications over the past 20 years. Most research topics were tumor-related, among which tumor risk prediction and prognostic evaluation were the main contents. Research on prognostic assessment is more published and advanced, while risk prediction and diagnosis have good developmental prospects. Furthermore, nomogram of the urinary system has been highly developed. Following advancements in nomogram modeling, it has recently been applied to non-oncological subjects.

Conclusion

This bibliometric analysis provides a comprehensive overview of the current nomogram status, which could enable better understanding of its development over the years, and provide global researchers a comprehensive analysis and structured information to help identify hot spots and gaps in future research.

Nomogram for prediction of long-term survival with hepatocellular carcinoma based on NK cell counts

INTRODUCTION

Among all immune cells, natural killer (NK) cells play an important role as the first line of defense against tumor. The purpose of our study is to observe whether the NK cell counts can predict the overall survival of patients with hepatocellular carcinoma (HCC).

METHODS

To develop a novel model, from January 2010 to June 2015, HCC patients enrolled in Beijing Ditan hospital were divided into training and validation cohort. Cox multiple regression analysis was used to analyze the independent risk factors for 1-year, 3-year and 5-year overall survival (OS) of patients with HCC, and the nomogram was used to establish the prediction model. In addition, the decision tree was established to verify the contribution of NK cell counts to the survival of patients with HCC.

RESULTS

The model used in predicting overall survival of HCC included six variables (namely, NK cell counts, albumin (ALB) level, alpha-fetoprotein (AFP) level, portal vein tumor thrombus (PVTT), tumor number and treatment). The C-index of nomogram model in HCC patients predicting 1-year, 3-year and 5-year overall survival was 0.858, 0.788 and 0.782 respectively, which was higher than tumor-lymph node-metastasis (TNM) staging system, Okuda, model for end-stage liver disease (MELD), MELD-Na, the Chinese University Prognostic Index (CUPI) and Japan Integrated Staging (JIS) scores (p < 0.001). The decision tree showed the specific 5-year OS probability of HCC patients under different risk factors, and found that NK cell counts were the third in the column contribution.

CONCLUSIONS

Our study emphasizes the utility of NK cell counts for exploring interactions between long-term survival of HCC patients and predictor variables.

Downgrade BI-RADS 4A Patients Using Nomogram Based on Breast Magnetic Resonance Imaging, Ultrasound, and Mammography

Objectives

To downgrade BI-RADS 4A patients by constructing a nomogram using R software.

Materials and Methods

A total of 1,717 patients were retrospectively analyzed who underwent preoperative ultrasound, mammography, and magnetic resonance examinations in our hospital from August 2019 to September 2020, and a total of 458 patients of category BI-RADS 4A (mean age, 47 years; range 18–84 years; all women) were included. Multivariable logistic regression was used to screen out the independent influencing parameters that affect the benign and malignant tumors, and the nomogram was constructed by R language to downgrade BI-RADS 4A patients to eligible category.

Results

Of 458 BI-RADS 4A patients, 273 (59.6%) were degraded to category 3. The malignancy rate of these 273 lesions is 1.5% (4/273) (<2%), and the sensitivity reduced to 99.6%, the specificity increased from 4.41% to 45.3%, and the accuracy increased from 63.4% to 78.8%.

Conclusion

By constructing a nomogram, some patients can be downgraded to avoid unnecessary biopsy.

Establishment and validation of a predictive model for mortality within 30 days in patients with sepsis-induced blood pressure drop: A retrospective analysis

OBJECTIVES

To establish and validate an individualized nomogram to predict the probability of death within 30 days in patients with sepsis-induced blood pressure drop would help clinical physicians to pay attention to those with higher risk of death after admission to wards.

METHODS

A total of 1023 patients who were admitted to the Dongyang People's Hospital, China, enrolled in this study. They were divided into model group (717 patients) and validation group (306 patients). The study included 13 variables. The independent risk factors leading to death within 30 days were screened by univariate analyses and multivariate logistic regression analyses and used for Nomogram. The discrimination and correction of the prediction model were assessed by the area under the Receiver Operating Characteristic (ROC) curve and the calibration chart. The clinical effectiveness of the prediction model was assessed by the Decision Curve Analysis (DCA).

RESULTS

Seven variables were independent risk factors, included peritonitis, respiratory failure, cardiac insufficiency, consciousness disturbance, tumor history, albumin level, and creatinine level at the time of admission. The area under the ROC curve of the model group and validation group was 0.834 and 0.836. The P value of the two sets of calibration charts was 0.702 and 0.866. The DCA curves of the model group and validation group were above the two extreme (insignificant) curves.

CONCLUSIONS

The model described in this study could effectively predict the death of patients with sepsis-induced blood pressure drop.

Malignancy Risk Stratification Prediction of Amorphous Calcifications Based on Clinical and Mammographic Features

Purpose

To explore the potential factors influencing the malignancy risk of amorphous calcifications and establish a predictive nomogram for malignancy risk stratification.

Patients and Methods

Consecutive mammograms from January 2013 to December 2018 were retrospectively reviewed. Traditional clinical features were recorded, and mammographic features were estimated according to the 5th BI-RADS. Included calcifications were randomly divided into the training and validation cohorts. A nomogram was developed to graphically predict the risk of malignancy (risk) based on stepwise multivariate logistic regression analysis. The discrimination and calibration performance of the model were assessed in both the training and validation cohorts.

Results

Finally, 1018 amorphous calcifications with final pathological results in 907 women were identified with a malignancy rate of 28.4% (95% CI: 25.7%, 31.3%). The malignancy rates of subgroups divided by the distribution of calcifications, quantity of calcifications, age, menopausal status and family history of cancer were significantly different. There were 712 cases and 306 cases in the training and validation cohorts. The prediction nomogram was finally developed based on four risk factors, including age and distribution, maximum diameter and quantity of calcifications. The AUC of the nomogram was 0.799 (95% CI: 0.761, 0.836) in the training cohort and 0.795 (95% CI: 0.738, 0.852) in the validation cohort.

Conclusion

On mammography, the distribution, maximum diameter and quantity of calcifications are independent predictors of malignant amorphous calcifications and can be easily obtained in the clinic. The nomogram developed in this study for individualized malignancy risk stratification of amorphous calcifications shows good discrimination performance.

Does establishing a preoperative nomogram including ultrasonographic findings help predict the likelihood of malignancy in patients with microcalcifications?

BACKGROUND

Mammography (MG) is highly sensitive for detecting microcalcifications, but has low specificity. This study investigates whether establishing a preoperative nomogram including ultrasonographic findings can help predict the likelihood of malignancy in patients with mammographic microcalcification.

METHODS

Between May 2012 and January 2017, 475 patients with suspicious microcalcifications detected on MG underwent ultrasonography (US). The χ² test was used to screen risk factors among the variables. Then, a multivariate logistic regression analysis was performed to identify independent predictors of malignant microcalcifications. A mammographic nomogram (M nomogram) and mammographic-ultrasonographic nomogram (M-U nomogram) were established based on multivariate logistic regression models. The discriminatory ability and clinical utility of both nomograms were compared by the receiver operating characteristics curve and decision curve analysis. The calibration ability was evaluated using a calibration curve.

RESULTS

Among the cases, 68.2% (324/475) were pathologically diagnosed as breast cancer and 31.8% (151/475) were benign lesions. Based on multivariate logistic regression analysis, age, clinical manifestation, morphology and distribution of microcalcifications on MG and lesions associated with microcalcifications on US were confirmed as independent predictors of malignant microcalcifications. In terms of discrimination ability, the C-index of the M-U nomogram was significantly higher than that of the M nomogram (0.917 vs 0.897, p = 0.006). The bias-corrected curve was close to the ideal line in the calibration curve. Decision curve analysis suggested that the M-U nomogram was superior to M nomogram.

CONCLUSIONS

Combining mammographic parameters with ultrasonographic findings in a nomogram provided better performance than an M nomogram alone, especially for dense breasts, which suggests the value of ultrasonographic finding for individualized prediction of malignancy in patients with microcalcifications.

Preoperative Prediction of Axillary Lymph Node Metastasis in Breast Cancer Using Mammography-Based Radiomics Method

It is difficult to accurately assess axillary lymph nodes metastasis and the diagnosis of axillary lymph nodes in patients with breast cancer is invasive and has low-sensitivity preoperatively. This study aims to develop a mammography-based radiomics nomogram for the preoperative prediction of ALN metastasis in patients with breast cancer. This study enrolled 147 patients with clinicopathologically confirmed breast cancer and preoperative mammography. Features were extracted from each patient's mammography images. The least absolute shrinkage and selection operator regression method was used to select features and build a signature in the primary cohort. The performance of the signature was assessed using support vector machines. We developed a nomogram by incorporating the signature with the clinicopathologic risk factors. The nomogram performance was estimated by its calibration ability in the primary and validation cohorts. The signature was consisted of 10 selected ALN-status-related features. The AUC of the signature from the primary cohort was 0.895 (95% CI, 0.887-0.909) and 0.875 (95% CI, 0.698-0.891) for the validation cohort. The C-Index of the nomogram from the primary cohort was 0.779 (95% CI, 0.752-0.793) and 0.809 (95% CI, 0.794-0.833) for the validation cohort. Our nomogram is a reliable and non-invasive tool for preoperative prediction of ALN status and can be used to optimize current treatment strategy for breast cancer patients.

Review of non-clinical risk models to aid prevention of breast cancer

A disease risk model is a statistical method which assesses the probability that an individual will develop one or more diseases within a stated period of time. Such models take into account the presence or absence of specific epidemiological risk factors associated with the disease and thereby potentially identify individuals at higher risk. Such models are currently used clinically to identify people at higher risk, including identifying women who are at increased risk of developing breast cancer. Many genetic and non-genetic breast cancer risk models have been developed previously. We have evaluated existing non-genetic/non-clinical models for breast cancer that incorporate modifiable risk factors. This review focuses on risk models that can be used by women themselves in the community in the absence of clinical risk factors characterization. The inclusion of modifiable factors in these models means that they can be used to improve primary prevention and health education pertinent for breast cancer. Literature searches were conducted using PubMed, ScienceDirect and the Cochrane Database of Systematic Reviews. Fourteen studies were eligible for review with sample sizes ranging from 654 to 248,407 participants. All models reviewed had acceptable calibration measures, with expected/observed (E/O) ratios ranging from 0.79 to 1.17. However, discrimination measures were variable across studies with concordance statistics (C-statistics) ranging from 0.56 to 0.89. We conclude that breast cancer risk models that include modifiable risk factors have been well calibrated but have less ability to discriminate. The latter may be a consequence of the omission of some significant risk factors in the models or from applying models to studies with limited sample sizes. More importantly, external validation is missing for most of the models. Generalization across models is also problematic as some variables may not be considered applicable to some populations and each model performance is conditioned by particular population characteristics. In conclusion, it is clear that there is still a need to develop a more reliable model for estimating breast cancer risk which has a good calibration, ability to accurately discriminate high risk and with better generalizability across populations.

Association Between Imaging Characteristics and Different Molecular Subtypes of Breast Cancer

RATIONALE AND OBJECTIVE

Breast cancer can be divided into four major molecular subtypes based on the expression of hormone receptor (estrogen receptor and progesterone receptor), human epidermal growth factor receptor 2, HER2 status, and molecular proliferation rate (Ki67). In this study, we sought to investigate the association between breast cancer subtype and radiological findings in the Chinese population.

MATERIALS AND METHODS

Medical records of 300 consecutive invasive breast cancer patients were reviewed from the database: the Breast Imaging Reporting and Data System. The imaging characteristics of the lesions were evaluated. The molecular subtypes of breast cancer were classified into four types: luminal A, luminal B, HER2 overexpressed (HER2), and basal-like breast cancer (BLBC). Univariate and multivariate logistic regression analyses were performed to assess the association between the subtype (dependent variable) and mammography or 15 magnetic resonance imaging (MRI) indicators (independent variables).

RESULTS

Luminal A and B subtypes were commonly associated with "clustered calcification distribution," "nipple invasion," or "skin invasion" (P <0.05). The BLBC subtype was more commonly associated with "rim enhancement" and persistent inflow type enhancement in delayed phase (P <0.05). HER2 overexpressed cancers showed association with persistent enhancement in the delayed phase on MRI and "clustered calcification distribution" on mammography (P <0.05).

CONCLUSION

Certain radiological features are strongly associated with the molecular subtype and hormone receptor status of breast tumor, which are potentially useful tools in the diagnosis and subtyping of breast cancer.

Comparing Mammography Abnormality Features to Genetic Variants in the Prediction of Breast Cancer in Women Recommended for Breast Biopsy

RATIONALE AND OBJECTIVES

The discovery of germline genetic variants associated with breast cancer has engendered interest in risk stratification for improved, targeted detection and diagnosis. However, there has yet to be a comparison of the predictive ability of these genetic variants with mammography abnormality descriptors.

MATERIALS AND METHODS

Our institutional review board-approved, Health Insurance Portability and Accountability Act-compliant study utilized a personalized medicine registry in which participants consented to provide a DNA sample and to participate in longitudinal follow-up. In our retrospective, age-matched, case-controlled study of 373 cases and 395 controls who underwent breast biopsy, we collected risk factors selected a priori based on the literature, including demographic variables based on the Gail model, common germline genetic variants, and diagnostic mammography findings according to Breast Imaging Reporting and Data System (BI-RADS). We developed predictive models using logistic regression to determine the predictive ability of (1) demographic variables, (2) 10 selected genetic variants, or (3) mammography BI-RADS features. We evaluated each model in turn by calculating a risk score for each patient using 10-fold cross-validation, used this risk estimate to construct Receiver Operator Characteristic Curve (ROC) curves, and compared the area under the ROC curve (AUC) of each using the DeLong method.

RESULTS

The performance of the regression model using demographic risk factors was not statistically different from the model using genetic variants (P = 0.9). The model using mammography features (AUC = 0.689) was superior to both the demographic model (AUC = .598; P < 0.001) and the genetic model (AUC = .601; P < 0.001).

CONCLUSIONS

BI-RADS features exceeded the ability of demographic and 10 selected germline genetic variants to predict breast cancer in women recommended for biopsy.

External validation of a publicly available computer assisted diagnostic tool for mammographic mass lesions with two high prevalence research datasets

PURPOSE

Lesions detected at mammography are described with a highly standardized terminology: the breast imaging-reporting and data system (BI-RADS) lexicon. Up to now, no validated semantic computer assisted classification algorithm exists to interactively link combinations of morphological descriptors from the lexicon to a probabilistic risk estimate of malignancy. The authors therefore aim at the external validation of the mammographic mass diagnosis (MMassDx) algorithm. A classification algorithm like MMassDx must perform well in a variety of clinical circumstances and in datasets that were not used to generate the algorithm in order to ultimately become accepted in clinical routine.

METHODS

The MMassDx algorithm uses a naïve Bayes network and calculates post-test probabilities of malignancy based on two distinct sets of variables, (a) BI-RADS descriptors and age ("descriptor model") and (b) BI-RADS descriptors, age, and BI-RADS assessment categories ("inclusive model"). The authors evaluate both the MMassDx (descriptor) and MMassDx (inclusive) models using two large publicly available datasets of mammographic mass lesions: the digital database for screening mammography (DDSM) dataset, which contains two subsets from the same examinations-a medio-lateral oblique (MLO) view and cranio-caudal (CC) view dataset-and the mammographic mass (MM) dataset. The DDSM contains 1220 mass lesions and the MM dataset contains 961 mass lesions. The authors evaluate discriminative performance using area under the receiver-operating-characteristic curve (AUC) and compare this to the BI-RADS assessment categories alone (i.e., the clinical performance) using the DeLong method. The authors also evaluate whether assigned probabilistic risk estimates reflect the lesions' true risk of malignancy using calibration curves.

RESULTS

The authors demonstrate that the MMassDx algorithms show good discriminatory performance. AUC for the MMassDx (descriptor) model in the DDSM data is 0.876/0.895 (MLO/CC view) and AUC for the MMassDx (inclusive) model in the DDSM data is 0.891/0.900 (MLO/CC view). AUC for the MMassDx (descriptor) model in the MM data is 0.862 and AUC for the MMassDx (inclusive) model in the MM data is 0.900. In all scenarios, MMassDx performs significantly better than clinical performance, P < 0.05 each. The authors furthermore demonstrate that the MMassDx algorithm systematically underestimates the risk of malignancy in the DDSM and MM datasets, especially when low probabilities of malignancy are assigned.

CONCLUSIONS

The authors' results reveal that the MMassDx algorithms have good discriminatory performance but less accurate calibration when tested on two independent validation datasets. Improvement in calibration and testing in a prospective clinical population will be important steps in the pursuit of translation of these algorithms to the clinic.

Development of an online, publicly accessible naive Bayesian decision support tool for mammographic mass lesions based on the American College of Radiology (ACR) BI-RADS lexicon

PURPOSE

To develop and validate a decision support tool for mammographic mass lesions based on a standardized descriptor terminology (BI-RADS lexicon) to reduce variability of practice.

MATERIALS AND METHODS

We used separate training data (1,276 lesions, 138 malignant) and validation data (1,177 lesions, 175 malignant). We created naïve Bayes (NB) classifiers from the training data with tenfold cross-validation. Our "inclusive model" comprised BI-RADS categories, BI-RADS descriptors, and age as predictive variables; our "descriptor model" comprised BI-RADS descriptors and age. The resulting NB classifiers were applied to the validation data. We evaluated and compared classifier performance with ROC-analysis.

RESULTS

In the training data, the inclusive model yields an AUC of 0.959; the descriptor model yields an AUC of 0.910 (P < 0.001). The inclusive model is superior to the clinical performance (BI-RADS categories alone, P < 0.001); the descriptor model performs similarly. When applied to the validation data, the inclusive model yields an AUC of 0.935; the descriptor model yields an AUC of 0.876 (P < 0.001). Again, the inclusive model is superior to the clinical performance (P < 0.001); the descriptor model performs similarly.

CONCLUSION

We consider our classifier a step towards a more uniform interpretation of combinations of BI-RADS descriptors. We provide our classifier at www.ebm-radiology.com/nbmm/index.html .

KEY POINTS

• We provide a decision support tool for mammographic masses at www.ebm-radiology.com/nbmm/index.html . • Our tool may reduce variability of practice in BI-RADS category assignment. • A formal analysis of BI-RADS descriptors may enhance radiologists' diagnostic performance.

Evaluation of low-energy contrast-enhanced spectral mammography images by comparing them to full-field digital mammography using EUREF image quality criteria

Objective

Contrast-enhanced spectral mammography (CESM) examination results in a low-energy (LE) and contrast-enhanced image. The LE appears similar to a full-field digital mammogram (FFDM). Our aim was to evaluate LE CESM image quality by comparing it to FFDM using criteria defined by the European Reference Organization for Quality Assured Breast Screening and Diagnostic Services (EUREF).

Methods

A total of 147 cases with both FFDM and LE images were independently scored by two experienced radiologists using these (20) EUREF criteria. Contrast detail measurements were performed using a dedicated phantom. Differences in image quality scores, average glandular dose, and contrast detail measurements between LE and FFDM were tested for statistical significance.

Results

No significant differences in image quality scores were observed between LE and FFDM images for 17 out of 20 criteria. LE scored significantly lower on one criterion regarding the sharpness of the pectoral muscle (p < 0.001), and significantly better on two criteria on the visualization of micro-calcifications (p = 0.02 and p = 0.034). Dose and contrast detail measurements did not reveal any physical explanation for these observed differences.

Conclusions

Low-energy CESM images are non-inferior to FFDM images. From this perspective FFDM can be omitted in patients with an indication for CESM.

Key Points

• Low-energy CESM images are non-inferior to FFDM images.

• Micro-calcifications are significantly more visible on LE CESM than on FFDM.

• There is no physical explanation for this improved visibility of micro-calcifications.

• There is no need for an extra FFDM when CESM is indicated.

Using natural language processing to extract mammographic findings

OBJECTIVE

Structured data on mammographic findings are difficult to obtain without manual review. We developed and evaluated a rule-based natural language processing (NLP) system to extract mammographic findings from free-text mammography reports.

MATERIALS AND METHODS

The NLP system extracted four mammographic findings: mass, calcification, asymmetry, and architectural distortion, using a dictionary look-up method on 93,705 mammography reports from Group Health. Status annotations and anatomical location annotation were associated to each NLP detected finding through association rules. After excluding negated, uncertain, and historical findings, affirmative mentions of detected findings were summarized. Confidence flags were developed to denote reports with highly confident NLP results and reports with possible NLP errors. A random sample of 100 reports was manually abstracted to evaluate the accuracy of the system.

RESULTS

The NLP system correctly coded 96-99 out of our sample of 100 reports depending on findings. Measures of sensitivity, specificity and negative predictive values exceeded 0.92 for all findings. Positive predictive values were relatively low for some findings due to their low prevalence.

DISCUSSION

Our NLP system was implemented entirely in SAS Base, which makes it portable and easy to implement. It performed reasonably well with multiple applications, such as using confidence flags as a filter to improve the efficiency of manual review. Refinements of library and association rules, and testing on more diverse samples may further improve its performance.

CONCLUSION

Our NLP system successfully extracts clinically useful information from mammography reports. Moreover, SAS is a feasible platform for implementing NLP algorithms.

Update on breast cancer risk prediction and prevention

PURPOSE OF REVIEW

Breast cancer is the most common cancer in women worldwide. This review will focus on current prevention strategies for women at high risk.

RECENT FINDINGS

The identification of women who are at high risk of developing breast cancer is key to breast cancer prevention. Recent findings have shown that the inclusion of breast density and a panel of low-penetrance genetic polymorphisms can improve risk estimation compared with previous models. Preventive therapy with aromatase inhibitors has produced large reductions in breast cancer incidence in postmenopausal women. Tamoxifen confers long-term protection and is the only proven preventive treatment for premenopausal women. Several other agents, including metformin, bisphosphonates, aspirin and statins, have been found to be effective in nonrandomized settings.

SUMMARY

There are many options for the prevention of oestrogen-positive breast cancer, in postmenopausal women who can be given a selective oestrogen receptor modulator or an aromatase inhibitor. It still remains unclear how to prevent oestrogen-negative breast cancer, which occurs more often in premenopausal women. Identification of women at high risk of the disease is crucial, and the inclusion of breast density and a panel of genetic polymorphisms, which individually have low penetrance, can improve risk assessment.