Applying Data-driven Imaging Biomarker in Mammography for Breast Cancer Screening: Preliminary Study

Abstract

We assessed the feasibility of a data-driven imaging biomarker based on weakly supervised learning (DIB; an imaging biomarker derived from large-scale medical image data with deep learning technology) in mammography (DIB-MG). A total of 29,107 digital mammograms from five institutions (4,339 cancer cases and 24,768 normal cases) were included. After matching patients’ age, breast density, and equipment, 1,238 and 1,238 cases were chosen as validation and test sets, respectively, and the remainder were used for training. The core algorithm of DIB-MG is a deep convolutional neural network; a deep learning algorithm specialized for images. Each sample (case) is an exam composed of 4-view images (RCC, RMLO, LCC, and LMLO). For each case in a training set, the cancer probability inferred from DIB-MG is compared with the per-case ground-truth label. Then the model parameters in DIB-MG are updated based on the error between the prediction and the ground-truth. At the operating point (threshold) of 0.5, sensitivity was 75.6% and 76.1% when specificity was 90.2% and 88.5%, and AUC was 0.903 and 0.906 for the validation and test sets, respectively. This research showed the potential of DIB-MG as a screening tool for breast cancer.

Detection of Breast Cancer with Mammography: Effect of an Artificial Intelligence Support System

“…In radiology.rsna.org n Radiology: Volume 00: Number 0-2018 similar to that of the radiologist with the lowest performance in enriched and selected data sets, but only in very limited scenarios (eg, only soft-tissue lesions). A study by Kim et al (33) found that in-house-developed AI algorithms achieved a sensitivity of 76% and a specificity of 89% in a screening data set. Despite the differences in data sets, our results support the observed trend that AI algorithms are reaching a performance similar to that of radiologists for breast cancer detection in mammography.…”

Artificial intelligence and convolution neural networks assessing mammographic images: a narrative literature review

“…In 2018, a study using a CNN system was designed to differentiate between normal and abnormal images via biomarkers such as breast density, presence of mass and microcalcification . Their results indicated that abnormal images were more accurately identified when microcalcifications and benign cases were excluded.…”

“…The authors suggested that due to the training data having an uneven number of cases of normal and cancer cases, this caused overfitting and database dependency. This creates difficulty distinguishing between benign and malignant cases due to overreliance on heuristics and an inability to learn and adapt from new cases …”

“…Studies which used imaging modalities other than mammography were excluded. A total of 33 relevant peer‐reviewed articles were found . To compare the results among studies, relevant features were extracted: (1) author and publication year; (2) study aim; (3) utilised database; (4) number of images used; and (5) performance measures including sensitivity, specificity and area under the curve (AUC).…”

The Role of Deep Learning in Breast Screening

“…However, the inability to identify malignant lesions limits its ability to perform as an independent reader. Finally, Kim et al developed a system which made a malignancy prediction on an entire mammography case (all 4 views) [25]. The model was trained on malignant (biopsy proven) and normal (with at least 2 years of negative followup) cases from multiple hardware vendors.…”