Applying a new computer-aided detection scheme generated imaging marker to predict short-term breast cancer risk

Abstract

This study aims to investigate the feasibility of identifying a new quantitative imaging marker based on false-positives generated by a computer-aided detection (CAD) scheme to help predict short-term breast cancer risk. An image dataset including four view mammograms acquired from 1044 women was retrospectively assembled. All mammograms were originally interpreted as negative by radiologists. In the next subsequent mammography screening, 402 women were diagnosed with breast cancer and 642 remained negative. An existing CAD scheme was applied 'as is' to process each image. From CAD-generated results, four detection features including the total number of (1) initial detection seeds and (2) the final detected false-positive regions, (3) average and (4) sum of detection scores, were computed from each image. Then, by combining the features computed from two bilateral images of left and right breasts from either craniocaudal or mediolateral oblique view, two logistic regression models were trained and tested using a leave-one-case-out cross-validation method to predict the likelihood of each testing case being positive in the next subsequent screening. The new prediction model yielded the maximum prediction accuracy with an area under a ROC curve of AUC = 0.65 ± 0.017 and the maximum adjusted odds ratio of 4.49 with a 95% confidence interval of (2.95, 6.83). The results also showed an increasing trend in the adjusted odds ratio and risk prediction scores (p < 0.01). Thus, this study demonstrated that CAD-generated false-positives might include valuable information, which needs to be further explored for identifying and/or developing more effective imaging markers for predicting short-term breast cancer risk.

New Frontiers: An Update on Computer-Aided Diagnosis for Breast Imaging in the Age of Artificial Intelligence

“…Perhaps most important, work is being done to develop DL-based CAD systems that can directly comment on clinical outcomes. A recently proposed CAD scheme suggests that novel quantitative image markers based on false-positive mammograms could predict short-term breast cancer risk [63]. Other DL-based CAD systems focus on risk assessment.…”

Developing global image feature analysis models to predict cancer risk and prognosis

“…1) Short-term breast cancer risk using the bilateral mammographic density asymmetrical features computed from the "prior" negative screening mammograms [34,47,48]; 2) Likelihood of the case being abnormal using the global image features computed from the "current" screening mammograms (case-based CAD scheme) [16,49]; 3) Response of breast tumors to neoadjuvant chemotherapies using the global kinetic image features computed from the breast MRI performed before chemotherapy [40]; 4) Response of ovarian cancer patients to chemotherapy using the global adiposity-related image features computed from abdominal CT images performed before chemotherapy [30,42].…”

“…Most CAD schemes include three steps: (1) detect suspicious regions that may depict tumors, (2) segment the targeted regions, and (3) train a machine learning model that fuses multiple image features computed from the segmented regions [15]. Despite great research enthusiasm and effort, false-positive detection rates of CAD schemes remain high [16], and whether using CAD can add values in clinical practice to help improve radiologists' performance in reading and interpreting mammograms remains controversial [17]. The technical challenges and limitations in developing CAD schemes may include but not limited to (1) difficulty in accurate segmentation of the targeted tumors from the images due to tissue overlap, connection, and fuzzy boundary, which reduce the accuracy and reproducibility of the computed image features to build robust machine learning models [18]; (2) high false-positive cues in the detection schemes, which can mislead radiologists and reduce their performance [19]; (3) use of small or biased training datasets, which causes overfitting and reduces robustness of CAD schemes when applied to new testing cases [20]; (4) higher correlation of the detection results between CAD and radiologists, which reduces the clinical utility of CAD as "the second reader" [21]; and (5) difficulty in developing multi-image-based CAD schemes [22] to fuse and compare variation of the image features in the longitudinal images [23] or different views of images [24].…”

Is bigger always better? A controversial journey to the center of machine learning design, with uses and misuses of big data for predicting water meter failures

“…To solve this latter problem, we could then move the decision threshold towards the direction of minimizing the number of false negatives [23]; for example, setting the decision threshold at the value of 0.3, as per the confusion matrix of Fig. 10.…”