Previous studies have shown that there is a strong correlation between radiologists' diagnoses and their gaze when reading medical images. The extent to which gaze is attracted by content in a visual scene can be characterised as visual saliency. There is a potential for the use of visual saliency in computer-aided diagnosis in radiology. However, little is known about what methods are effective for diagnostic images, and how these methods could be adapted to address specific applications in diagnostic imaging. In this study, we investigate 20 state-of-the-art saliency models including 10 traditional models and 10 deep learning-based models in predicting radiologists' visual attention while reading 196 mammograms. We found that deep learning-based models represent the most effective type of methods for predicting radiologists' gaze in mammogram reading; and that the performance of these saliency models can be significantly improved by transfer learning. In particular, an enhanced model can be achieved by pre-training the model on a large-scale natural image saliency dataset and then finetuning it on the target medical image dataset. In addition, based on a systematic selection of backbone networks and network architectures, we proposed a parallel multi-stream encoded model which outperforms the state-of-the-art approaches for predicting saliency of mammograms.
Previous chapters in this book have concentrated on the instrumentation and measurement techniques for dosimetry of the therapeutic beams. In this chapter, a variety of measurement and calculation techniques will be reviewed for characterizing the radiation dose from x-ray imaging systems used in radiation therapy (RT). X-ray imaging systems are now used extensively throughout a patient's treatment for all complex RTs, and in many cases for simple palliative RTs as well. Nearly, all patients will undergo a multislice computed tomography (CT) examination for localizing the target volume and nearby organs at risk. In addition, a variety of x-ray imaging systems are available to image the patient at the point of treatment, either immediately prior to beam delivery (e.g., Korreman et al., 2010; Moore et al., 2014) to ensure accurate patient alignment, or during beam delivery to monitor intrafraction motion 22.1 Introduction 22.2 Dose Measurement for CT 22.3 Dose Measurement for CBCT for Radiotherapy Applications 22.4 Measurement (and Calculation) of Dose for Planar kV Imaging 22.5 Dose Measurement for MV Portal Imaging 22.6 Dose Measurement for MVCT and MV-CBCT 22.7 Dosimeters for All Modalities 22.8 Dose Calculation Methods 22.9 Estimating Effective Dose and Risk 22.10 Combining Dose from RT and Imaging 22.11 Clinical Consequences and Benefits 22.12 Closing Remarks 9781482252217_C022.indd 561 21/07/17 5:53 PM * MV-CBCT was developed and commercialized by Siemens Medical Systems, but is no longer commercially available since Siemens stopped producing radiation therapy treatment machines.
230 Background: SCALOP, a randomized, multicenter, phase II trial, tested efficacy/toxicity of gemcitabine (Gem) versus capecitabine (Cap) based CRT following 4 cycles of induction chemotherapy. 114 patients from 28 centers were registered and 75 patients had CRT. The RT Quality Assurance included a test case. GTVs outlined by investigators were evaluated by the chief investigator (CI). We report analysis of the test case. Methods: The test case was LANPC of head of pancreas. The gold standard GTV (gsGTV) was outlined by the CI and trial radiologist. Protocol specified margins (GTV + 2cm sup-inf, 1.5 cm radial) were applied to generate the gold standard PTV (gsPTV). CT-DICOM dataset of the planning scan was sent to participating centers for tumor volume delineation by investigators and individualized feedback was provided by CI. CERR software was used to compare the investigator volumes (iGTV and iPTV) against the gold standard. Results: Volumes from 25 investigators were analyzed. The results are summarized below (Table). Slice-by-slice conformity indices and qualitative review showed that most variation occurred at the superior and inferior extremes of the tumor and a peri-tumoral node was frequently missed by investigators. Conclusions: SCALOP is the first trial in LANPC to test investigator-delineated volumes using conformity indices, and has demonstrated considerable variation in iGTVs. Trial workshops and real-time central review of tumor outlines are proposed. [Table: see text]
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.