Phyllis J. Kornguth

Prediction of breast cancer malignancy using an artificial neural network

Background. An artificial neural network (ANN) was developed to predict breast cancer from mammographic findings. This network was evaluated in a retrospective study.

The use of acoustic streaming in breast lesion diagnosis : A clinical study

Results from a clinical study are presented, in which ultrasonically-induced acoustic streaming was successfully used to differentiate fluid-filled lesions (cysts) from solid lesions in the breast. In this study, high-intensity ultrasound pulses from a modified commercial scanner were used to induce acoustic streaming in cyst fluid, and this motion was detected using Doppler methods. Acoustic streaming was generated and detected in 14 of 15 simple cysts, and 4 of 14 sonographically indeterminate breast lesions. This lesion differentiation method appears to be particularly suited for diagnosis of small, possibly newer, cysts that appear indeterminate on conventional sonography due to their size. The results indicate that this method would be a useful adjunct to conventional sonography for the purpose of breast lesion classification.

Pain during mammography : characteristics and relationship to demographic and medical variables

&NA; Reports of pain during mammography show that there is great variability in both the incidence of reported pain (0.2–62%) and the intensity of that pain. Much of that variability may be due to the measures used to rate mammography pain. This is the first study that has examined the incidence, quality and intensity of mammography pain using a variety of pain measures. A sample of 119 women undergoing screening mammography was studied using four pain scales, three well‐validated measures frequently used in the pain research literature as well as a pain/discomfort measure frequently reported in the radiology literature. A large proportion (up to 91%) of women report having some degree of pain during mammography. The intensity of that pain was typically in the low to moderate range, but a small proportion of women (< 15%) reported intense pain. The incidence of reported pain was related to the pain measure used. Pain measures that provided a woman with many options for reporting pain were associated with a higher incidence of pain than a scale that provided only one or two options. Thus, some of the variability in reported incidence of pain during mammography can be explained by the pain scale used in the study. Demographic and medical variables could explain 18–20% of the variance in mammography pain. Two of the variables that were shown to consistently predict a painful mammographic experience were (1) average pain at the last mammogram and (2) breast density. This study demonstrated that the pain measure selected for use in a particular study may depend on the population being studied. A college education was found to be an important predictor of pain scores on the McGill Pain Questionnaire. Thus, this pain measure may be of limited usefulness in studying a population of women with little formal education.

A novel ultrasonic technique for differentiating cysts from solid lesions: preliminary results in the breast.

The feasibility of a new ultrasonic technique to distinguish cysts from solid lesions is explored. High intensity pulses are used to induce acoustic streaming in cyst fluid, and this motion is detected using Doppler techniques. Acoustic streaming cannot be generated in solid lesions, therefore, its detection would indicate a cyst. In six of seven breast cysts motion was clearly generated and detected in vivo. Ultrasonic pulses with intensities up to 4.4 W cm-2 (I(spta) in water) were focused on the cysts for 10 s. Lesion diameters ranged from 0.6 to 2.5 cm; induced flow velocities were less than 4.0 cm s-1.

Effect of patient histoy data on the prediction of breast cancer from mammographic findings with artificial neural networks

Rationale and Objectives. The authors evaluated the contribution of medical history data to the prediction of breast cancer with artificial neural network (ANN) models based on mammographic findings. Materials and Methods. Three ANNs were developed: The first used 10 Breast Imaging Reporting and Data System (BI-RADS) variables; the second, the BI-RADS variables plus patient age; the third, the BI-RADS variables, patient age, and seven other history variables, for a total of 18 inputs. Performance of the ANNs and the original radiologists impression were evaluated with five metrics: receiver operating characteristic area index (Az); specificity at given sensitivities of 100%, 98%, and 95%; and positive predictive value. Results. All three ANNs consistently outperformed the radiologists impression over all five performance metrics. The patient-age variable was particularly valuable. Adding the age variable to the basic ANN model, which used only the BI-RADS findings, significantly improved Az (P = .028). In fact, replacing all history data with just the age variable resulted in virtually no changes for Az or specificity at 98% sensitivity (P = .324 and P = .410, respectively). Conclusion. Patient age was an important variable for the prediction of breast cancer from mammographic findings with the ANNs. For this data set, all history data could be replaced with age alone.

Pain coping and the pain experience during mammography : a preliminary study

&NA; This study examined how pain coping efficacy and pain coping strategies were related to reports of pain during mammography. Subjects were 125 women over the age of 50 undergoing screening mammograms. Prior to their mammogram, all subjects completed the Coping Strategies Questionnaire (CSQ) to assess how they cope with day‐to‐day pain experiences. Ratings of pain during the mammogram were collected using a 6‐point pain/discomfort scale, a 100‐mm Visual Analog Scale, the adjective checklist of the McGill Pain Questionnaire, and the Brief Pain Inventory. Up to 93% of the women reported the mammogram examination was painful. On average, women rated the mammography pain in the low to moderate range. Considerable variability in pain ratings was found, however, with some women reporting severe pain and others reporting little or no pain. Correlational analyses were conducted to examine how coping efficacy (CSQ ratings of ability to decrease pain and ability to control pain) and coping strategies (CSQ pain coping strategy subscales) related to variations in pain report. There was a pattern for ratings of ability to decrease pain to be related to lower ratings of current mammography pain. Women who rated their ability to decrease pain as high reported lower average levels of mammography pain, lower ratings on the mammography pain/discomfort scale, and were much more likely to report having had lower levels of pain during their last mammogram. These findings suggest that women who rate their coping efficacy in decreasing day‐to‐day pain as low may be at higher risk for having a painful mammogram. Individual pain coping strategies were not generally correlated with pain ratings. Behavioral interventions (e.g., patient controlled breast compression) and cognitive therapy interventions (e.g., training in the use of calming self‐statements or distraction techniques) designed to increase coping efficacy potentially could be useful in reducing pain in women who are at risk for pain during mammography.

Mammography pain and discomfort: a cognitive-behavioral perspective

Although some women report having little pain or discomfort during mammography, other women find mammography to be a painful and uncomfortable experience. Cognitive and behavioral factors may influence the perception of pain and discomfort during mammography. This review critically evaluates the research on mammography pain from a cognitive-behavioral perspective. The review is in three sections. The first reviews studies measuring pain and discomfort in women who have recently had mammograms and studies investigating the relative importance of pain and discomfort in influencing womens decisions to have a mammogram. The second section presents a cognitive-behavioral model of mammography pain that is based on theories of behavior and self-regulation developed by Kanfer and Hagerman (1987). The review concludes with a discussion of the implications of the cognitive-behavioral perspective for clinical management and research on mammography pain and discomfort.

Reminder Letter, Tailored Stepped-Care, and Self-Choice Comparison for Repeat Mammography

BACKGROUND The main benefits of mammography come from regular on-schedule screening. However, few studies have examined interventions to achieve repeat screening. SETTING AND PARTICIPANTS Participants were women aged 50 to 74, recruited through one setting in Rhode Island and another in North Carolina. Participants had a mammogram already scheduled at recruitment, and had to keep that appointment in order to be eligible for the repeat mammography intervention. A total of 1614 women were in the intervention sample. DESIGN A four-group randomized design was used: Group 1, a simple reminder letter; Group 2, a 2-month, tailored, stepped intervention delivered 2 months after the completed mammogram; Group 3, a 10-month, tailored, stepped intervention delivered 2 months before the repeat mammogram was due; and Group 4, self-choice of one of the above three strategies. INTERVENTION The intervention took place between June 1996 and May 1997. The reminder letter and two levels of the stepped intervention were delivered by mail. The third level of the stepped strategy was a counselor telephone call. Groups 2 and 3 were identical, except for timing. OUTCOME MEASURE Obtaining the next due mammogram within 15 months, based on clinic records. RESULTS There were no statistically significant differences among the four groups, both in the total sample and at the two sites separately. CONCLUSIONS On average, a simple reminder may be as effective as more complex strategies for women with a prior on-schedule exam. However, attention is still needed to identify women at risk of lapsing from screening. Some women may require more-intensive interventions.

Postmenopausal Estrogen Use and Invasive versus in situ Breast Cancer Risk

To examine the effect of cancer histopathology on the relationship between estrogen-replacement therapy (ERT) use and breast cancer risk, we performed a case-control study of 109 postmenopausal women 45 years or older with in situ or invasive breast cancer matched to 545 controls. When in situ and invasive tumors were combined, the overall odds ratio (OR) describing the association between ERT use and breast cancer risk was not statistically significantly elevated (adjusted OR = 1.48, 95% confidence interval [CI] = 0.89-2.47). When the analyses were confined to women with invasive disease, risk estimates were uniformly higher (adjusted OR = 1.85, 95% CI = 1.00-3.45). In contrast, the overall estimate for the relationship between ERT use and in situ breast cancer was close to 1 (adjusted OR = 1.08, 95% CI = 0.42-2.77). The positive association between ERT use and invasive breast cancer we observed, and the lack of association in women with in situ disease, may represent a distinct biological difference or may be related to the small sample size of our study.

Utility of Targeted Sonography in the Evaluation of Focal Breast Pain

Objective. To determine the utility of targeted sonography in the evaluation of patients with focal breast pain. Methods. From January 1995 through December 1999, 110 targeted sonographic examinations were performed in 99 patients for evaluation of focal breast pain in the absence of an associated palpable mass. The sonographic, mammographic, and clinical findings were reviewed. The hospital pathology database was searched to identify any interval cancers and false‐negative interpretations. Results. No cancer was identified in any of the 110 examinations. Eighty‐five (77.3%) of the examinations had negative findings. Cysts were identified in 15 cases (13.6%), and 3 solid masses (2.7%) were identified. Two of these 3 solid masses had biopsies and were shown to be benign, whereas the third mass was followed for 29 months without change. Most patients were premenopausal, had no family or personal history of breast cancer, and were not taking exogenous hormones. Eighty‐five patients (77%) were referred by primary care physicians. Conclusions. In patients with focal breast pain without an associated palpable mass, sonography may be more useful for patient reassurance than for cancer detection.