Dataset Information

Automatic Labeling of Special Diagnostic Mammography Views from Images and DICOM Headers.

ABSTRACT: Applying state-of-the-art machine learning techniques to medical images requires a thorough selection and normalization of input data. One of such steps in digital mammography screening for breast cancer is the labeling and removal of special diagnostic views, in which diagnostic tools or magnification are applied to assist in assessment of suspicious initial findings. As a common task in medical informatics is prediction of disease and its stage, these special diagnostic views, which are only enriched among the cohort of diseased cases, will bias machine learning disease predictions. In order to automate this process, here, we develop a machine learning pipeline that utilizes both DICOM headers and images to predict such views in an automatic manner, allowing for their removal and the generation of unbiased datasets. We achieve AUC of 99.72% in predicting special mammogram views when combining both types of models. Finally, we apply these models to clean up a dataset of about 772,000 images with expected sensitivity of 99.0%. The pipeline presented in this paper can be applied to other datasets to obtain high-quality image sets suitable to train algorithms for disease detection.

SUBMITTER: Lituiev DS

PROVIDER: S-EPMC6456464 | biostudies-literature | 2019 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Automatic Labeling of Special Diagnostic Mammography Views from Images and DICOM Headers.

Lituiev Dmytro S DS Trivedi Hari H Panahiazar Maryam M Norgeot Beau B Seo Youngho Y Franc Benjamin B Harnish Roy R Kawczynski Michael M Hadley Dexter D

Journal of digital imaging 20190401 2

Applying state-of-the-art machine learning techniques to medical images requires a thorough selection and normalization of input data. One of such steps in digital mammography screening for breast cancer is the labeling and removal of special diagnostic views, in which diagnostic tools or magnification are applied to assist in assessment of suspicious initial findings. As a common task in medical informatics is prediction of disease and its stage, these special diagnostic views, which are only e ...[more]

PMID: 30465142

Similar Datasets

Project description:BackgroundBreast cancer is the most commonly diagnosed cancer among women and a leading cause of death from cancer in women in Europe. Although breast cancer incidence is on the rise worldwide, breast cancer mortality over the past 25 years has been stable or decreasing in some countries and a fall in breast cancer mortality rates in most European countries in the 1990s was reported by several studies, in contrast, in Greece have not reported these favourable trends. In Greece, the age-standardised incidence and mortality rate for breast cancer per 100.000 in 2006 was 81,8 and 21,7 and although it is lower than most other countries in Europe, the fall in breast cancer mortality that observed has not been as great as in other European countries. There is no national strategy for screening in this country. This study reports on the use of mammography among middle-aged women in rural Crete and investigates barriers to mammography screening encountered by women and their primary care physicians.MethodsDesignSemi-structured individual interviews.Setting and participantsThirty women between 45-65 years of age, with a mean age of 54,6 years, and standard deviation 6,8 from rural areas of Crete and 28 qualified primary care physicians, with a mean age of 44,7 years and standard deviation 7,0 serving this rural population.Main outcome measureQualitative thematic analysis.ResultsMost women identified several reasons for not using mammography. These included poor knowledge of the benefits and indications for mammography screening, fear of pain during the procedure, fear of a serious diagnosis, embarrassment, stress while anticipating the results, cost and lack of physician recommendation. Physicians identified difficulties in scheduling an appointment as one reason women did not use mammography and both women and physicians identified distance from the screening site, transportation problems and the absence of symptoms as reasons for non-use.ConclusionWomen are inhibited from participating in mammography screening in rural Crete. The provision of more accessible screening services may improve this. However physician recommendation is important in overcoming women's inhibitions. Primary care physicians serving rural areas need to be aware of barriers preventing women from attending mammography screening and provide women with information and advice in a sensitive way so women can make informed decisions regarding breast cancer screening.

Dataset Information

Automatic Labeling of Special Diagnostic Mammography Views from Images and DICOM Headers.

Publications

Automatic Labeling of Special Diagnostic Mammography Views from Images and DICOM Headers.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets