In this work, we study convolutional neural network encoder-decoder architectures with pre-trained encoder weights for breast mass segmentation from digital screening mammograms. To automatically detect breast cancer, one fundamental task to achieve is the segmentation of the potential abnormal regions. Our objective was to find out whether encoder weights trained for breast cancer evaluation in comparison to those learned from natural images can yield a better model initialization, and furthermore improved segmentation results. We applied transfer learning and initialized the encoder, namely ResNet34 and ResNet22, with ImageNet weights and weights learned from breast cancer classification, respectively. A large clinically-realistic Finnish mammography screening dataset was utilized in model training and evaluation. Furthermore, an independent Portuguese INbreast dataset was utilized for further evaluation of the models. 5-fold cross-validation was applied for training. Soft Focal Tversky loss was used to calculate the model training time error. Dice score and Intersection over Union were used in quantifying the degree of similarity between the annotated and automatically produced segmentation masks. The best performing encoder-decoder with ResNet34 encoder tailed with U-Net decoder yielded Dice scores (mean±SD) of 0.7677±0.2134 for the Finnish dataset, and ResNet22 encoder tailed with U-Net decoder 0.8430±0.1091 for the INbreast dataset. No large differences in segmentation accuracy were found between the encoders initialized with weights pre-trained from breast cancer evaluation, and of those from natural image classification.
Purpose: Coronary artery calcium (CAC) scoring with computed tomography (CT) has been proposed as a screening tool for coronary artery disease, but concerns remain regarding the radiation dose of CT CAC scoring. Photon counting detectors and iterative reconstruction (IR) are promising approaches for patient dose reduction, yet the preservation of CAC scores with IR has been questioned. The purpose of this study was to investigate the applicability of IR for quantification of CAC using a photon counting flat-detector.
Approach: We imaged a cardiac rod phantom with calcium hydroxyapatite (CaHA) inserts with different noise levels using an experimental photon counting flat-detector CT setup to simulate the clinical CAC scoring protocol. We applied filtered back projection (FBP) and two IR algorithms with different regularization strengths. We compared the air kerma values, image quality parameters [noise magnitude, noise power spectrum, modulation transfer function (MTF), and contrast-to-noise ratio], and CaHA quantification accuracy between FBP and IR.
Results: IR regularization strength influenced CAC scores significantly (p < 0.05). The CAC volumes and scores between FBP and IRs were the most similar when the IR regularization strength was chosen to match the MTF of the FBP reconstruction.
Conclusion: When the regularization strength is selected to produce comparable spatial resolution with FBP, IR can yield comparable CAC scores and volumes with FBP. Nonetheless, at the lowest radiation dose setting, FBP produced more accurate CAC volumes and scores compared to IR, and no improved CAC scoring accuracy at low dose was demonstrated with the utilized IR methods.
X-ray computed tomography (CT) is widely used in diagnostic imaging. Due to the growing number of CT scans worldwide, the consequent increase in populational dose is of concern. Therefore, strategies for dose reduction are investigated. One strategy is to perform interior computed tomography (iCT), where X-ray attenuation data are collected only from an internal region-of-interest. The resulting incomplete measurement is called a truncated sinogram (TS). Missing data from the surrounding structures results in reconstruction artifacts with traditional methods. In this work, a deep learning framework for iCT is presented. TS is extended with a U-net convolutional neural network, and the extended sinogram is reconstructed with filtered backprojection (FBP). U-net was trained for 300 epochs with L1 loss. Truncated and full sinograms were simulated from CT angiography slice images for training data.1097/193/152 sinograms from 500 patients were used in the training, validation, and test sets, respectively. Our method was compared with FBP applied to TS (TS-FBP), adaptive sinogram detruncation followed by FBP (ADT-FBP), total variation regularization applied to TS, and FBPConvNet using TS-FBP as input. The best root-mean-square error (0.04±0.01, mean±SD) and peak signal-to-noise-ratio (29.5±2.9) dB in the test set were observed with the proposed method. However, slightly higher structural similarity indices were observed with FBPConvNet (0.97±0.01) and ADT-FBP (0.97±0.01) than with our method (0.96 ± 0.01). This work suggests that extension of truncated sinogram data with U-Net is a feasible way to reconstruct iCT data without artifacts that render image quality undesirable for medical diagnostics.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.