Computer-assisted analysis of mammographic clustered calcifications

Freundlich, Irwin M.; Hunter, Tim B.; Seeley, George W.; D’Orsi, Carl J.; Sadowsky, Norman L.

doi:10.1016/s0009-9260(89)80209-0

Cited by 17 publications

(9 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Also, according to Ref. 25, the average distance between calcifications within a cluster was the only other cluster feature next to the number of microcalcifications per cluster that yielded statistically significant results in the same study, with a 92% chance for a cluster to be benign if mcdϾ1 mm.…”

Section: Clustering and False-positive Cluster Reductionmentioning

confidence: 68%

“…͑a͒ Number of microcalcifications per cluster, numc. According to a clinical study by Freundlich et al, 25 this feature was the most important for discriminating between benign and malignant clusters. In that study, a value of numcϽ10 resulted in an 82% chance for a cluster to be benign while if numcу10 there was a chance of 56% for the cluster to be benign.…”

Section: Clustering and False-positive Cluster Reductionmentioning

confidence: 95%

“…The overall low false-positive rate of our method can be partly attributed to the focus on segmenting only suspicious clusters, as opposed to searching for all real microcalcification clusters including benign and malignant ones. That enabled the use of features numc and mcd, which were shown to enable discrimination between malignant and benign clusters, 25 to remove false clusters with benign characteristics and thus decrease the number of false positives. The two-step clustering scheme described in Sec.…”

Section: B Performance Evaluationmentioning

confidence: 99%

See 2 more Smart Citations

Segmentation of suspicious clustered microcalcifications in mammograms

Gavrielides

Vargas-Voracek

et al. 2000

Medical Physics

View full text Add to dashboard Cite

We have developed a multistage computer-aided diagnosis (CAD) scheme for the automated segmentation of suspicious microcalcification clusters in digital mammograms. The scheme consisted of three main processing steps. First, the breast region was segmented and its high-frequency content was enhanced using unsharp masking. In the second step, individual microcalcifications were segmented using local histogram analysis on overlapping subimages. For this step, eight histogram features were extracted for each subimage and were used as input to a fuzzy rule-based classifier that identified subimages containing microcalcifications and assigned the appropriate thresholds to segment any microcalcifications within them. The final step clustered the segmented microcalcifications and extracted the following features for each cluster: the number of microcalcifications, the average distance between microcalcifications, and the average number of times pixels in the cluster were segmented in the second step. Fuzzy logic rules incorporating the cluster features were designed to remove nonsuspicious clusters, defined as those with typically benign characteristics. A database of 98 images, with 48 images containing one or more microcalcification clusters, provided training and testing sets to optimize the parameters and evaluate the CAD scheme, respectively. The results showed a true positive rate of 93.2% and an average of 0.73 false positive clusters per image. A comparison of our results with other reported segmentation results on the same database showed comparable sensitivity and at the same time an improved false positive rate. The performance of the CAD scheme is encouraging for its use as an automatic tool for efficient and accurate diagnosis of breast cancer.

show abstract

Section: Clustering and False-positive Cluster Reductionmentioning

confidence: 68%

Section: Clustering and False-positive Cluster Reductionmentioning

confidence: 95%

Section: B Performance Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

Segmentation of suspicious clustered microcalcifications in mammograms

Gavrielides

Vargas-Voracek

et al. 2000

Medical Physics

View full text Add to dashboard Cite

show abstract

“…Alternative enhancement techniques using locally adaptive image filters can differentially affect only certain components of a lesion such as sharpness of border details. In addition, computers can be programmed to recognize patterns of microcalcifications and render an opinion re-garding histology [13]. Similarly, once a lesion is in digital form, the degree of blur of its border can be quantitated and thresholds separating benign and malignant lesions derived [ 141.…”

Section: Discussionmentioning

confidence: 99%

Selective digital enhancement of conventional film mammography

Hale

Cook

Baniqued

et al. 1994

Journal of Surgical Oncology

View full text Add to dashboard Cite

Screening mammography continues to play a key role in the early diagnosis of non-palpable breast cancer. Approximately 5% of patients will have lesions on their mammograms that appear neither clearly benign nor malignant. The proper management of these patients is not well defined. In this study, the mammograms of 64 women who underwent breast biopsy were reviewed. After the initial review, the mammograms were reevaluated using a personal computer driven digital conversion and enhancement system. The majority of indeterminate lesions were assigned a determinate diagnosis by enhancing the areas of abnormality through enlargement, contrast adjustment, and brightness adjustment. The radiologists were able to maintain baseline sensitivity and specificity throughout this process. In contrast, enhancement of lesions initially assigned a diagnosis of characteristically benign or malignant demonstrated no advantage over plain film mammography. These findings suggest a role for the selective use of digital mammography and enhancement techniques.

show abstract

“…For each cluster, three features were extracted: ͑a͒ number of microcalcifications in the cluster, numc, ͑b͒ average distance between microcalcifications within the cluster, mcd, and ͑c͒ the average number of times the segmented pixels within the cluster were selected during the segmentation of individual microcalcifications, mism. In a previous clinical study, 26 the first two features were found to be the most important for discriminating between benign and malignant clusters. The numc feature was categorized into four groups: ͑1͒ 3 microcalcifications, ͑2͒ 4 -5, ͑3͒ 6 -9, and ͑4͒ у10, similar to the approach used in Ref.…”

Section: Cluster Feature Extractionmentioning

confidence: 84%

Parameter optimization of a computer‐aided diagnosis scheme for the segmentation of microcalcification clusters in mammograms

Gavrielides¹,

Lo²,

Floyd³

2002

Medical Physics

View full text Add to dashboard Cite

Our purpose in this study is to develop a parameter optimization technique for the segmentation of suspicious microcalcification clusters in digitized mammograms. In previous work, a computer-aided diagnosis (CAD) scheme was developed that used local histogram analysis of overlapping subimages and a fuzzy rule-based classifier to segment individual microcalcifications, and clustering analysis for reducing the number of false positive clusters. The performance of this previous CAD scheme depended on a large number of parameters such as the intervals used to calculate fuzzy membership values and on the combination of membership values used by each decision rule. These parameters were optimized empirically based on the performance of the algorithm on the training set. In order to overcome the limitations of manual training and rule generation, the segmentation algorithm was modified in order to incorporate automatic parameter optimization. For the segmentation of individual microcalcifications, the new algorithm used a neural network with fuzzy-scaled inputs. The fuzzy-scaled inputs were created by processing the histogram features with a family of membership functions, the parameters of which were automatically extracted from the distribution of the feature values. The neural network was trained to classify feature vectors as either positive or negative. Individual microcalcifications were segmented from positive subimages. After clustering, another neural network was trained to eliminate false positive clusters. A database of 98 images provided training and testing sets to optimize the parameters and evaluate the CAD scheme, respectively. The performance of the algorithm was evaluated with a FROC analysis. At a sensitivity rate of 93.2%, there was an average of 0.8 false positive clusters per image. The results are very comparable with those taken using our previously published rule-based method. However, the new algorithm is more suited to generalize its performance on a larger population, depends on two monotonic outputs making its evaluation much easier and can be trained in an automatic way making practical its application on a large database.

show abstract

Computer-assisted analysis of mammographic clustered calcifications

Cited by 17 publications

References 7 publications

Segmentation of suspicious clustered microcalcifications in mammograms

Segmentation of suspicious clustered microcalcifications in mammograms

Selective digital enhancement of conventional film mammography

Parameter optimization of a computer‐aided diagnosis scheme for the segmentation of microcalcification clusters in mammograms

Contact Info

Product

Resources

About