Deep learning-based 3D in vivo dose reconstruction with EPID for MR-LINACs: A proof of concept study

2021 ◽  
Author(s):  
Yongbao Li ◽  
Fan Xiao ◽  
Biaoshui Liu ◽  
Mengke Qi ◽  
Xingyu Lu ◽  
...  
Keyword(s):  
Deep Learning ◽  
Intensity Modulated Radiotherapy ◽  
Ground Truth ◽  
Training Dataset ◽  
Average Dose ◽  
Dose Reconstruction ◽  
Imaging Device ◽  
Dose Volume Histograms ◽  
Return Effect

Abstract Objective: To develop a novel deep learning-based 3D in vivo dose reconstruction framework with electronic portal imaging device (EPID) for magnetic resonance-linear accelerators (MR-LINACs). Approach: The proposed method directly back-projected 2D portal dose into 3D patient coarse dose, which bypassed the complicated patient-to-EPID scatter estimation step used in conventional methods. A pre-trained convolutional neural network (CNN) was then employed to map the coarse dose to the final accurate dose. The electron return effect caused by the magnetic field was captured with the CNN model. Patient dose and portal dose datasets were synchronously generated with Monte Carlo simulation for 96 patients (78 cases for training and validation and 18 cases for testing) treated with fixed-beam intensity-modulated radiotherapy in four different tumor sites, including the brain, nasopharynx, lung, and rectum. Beam angles from the training dataset were further rotated 2–3 times, and doses were recalculated to augment the datasets. Results: The comparison between reconstructed doses and MC ground truth doses showed mean absolute errors < 0.88% for all tumor sites. The averaged 3D γ-passing rates (3%, 2 mm) were 97.42%±2.66% (brain), 98.53%±0.95% (nasopharynx), 99.41%±0.46% (lung), and 98.63%±1.01% (rectum). The dose volume histograms and indices also showed good consistency. The average dose reconstruction time, including back projection and CNN dose mapping, was less than 3 s for each individual beam. Significance: The proposed method can be potentially used for accurate and fast 3D dosimetric verification for online adaptive radiotherapy using MR-LINACs.

Abstract 20: Deep Learning-Based Automated Intracranial Hemorrhage Detection and Notification

Stroke ◽  
2020 ◽  
Vol 51 (Suppl_1) ◽  
Author(s):  
Benjamin Zahneisen ◽  
Matus Straka ◽  
Shalini Bammer ◽  
Greg Albers ◽  
Roland Bammer
Keyword(s):  
Deep Learning ◽  
Sensitivity And Specificity ◽  
Intracranial Hemorrhage ◽  
Health Care Professionals ◽  
Gold Standard ◽  
Ground Truth ◽  
Mobile App ◽  
Training Dataset ◽  
Convolutional Network ◽  
Ground Truth Data

Introduction: Ruling out hemorrhage (stroke or traumatic) prior to administration of thrombolytics is critical for Code Strokes. A triage software that identifies hemorrhages on head CTs and alerts radiologists would help to streamline patient care and increase diagnostic confidence and patient safety. ML approach: We trained a deep convolutional network with a hybrid 3D/2D architecture on unenhanced head CTs of 805 patients. Our training dataset comprised 348 positive hemorrhage cases (IPH=245, SAH=67, Sub/Epi-dural=70, IVH=83) (128 female) and 457 normal controls (217 female). Lesion outlines were drawn by experts and stored as binary masks that were used as ground truth data during the training phase (random 80/20 train/test split). Diagnostic sensitivity and specificity were defined on a per patient study level, i.e. a single, binary decision for presence/absence of a hemorrhage on a patient’s CT scan. Final validation was performed in 380 patients (167 positive). Tool: The hemorrhage detection module was prototyped in Python/Keras. It runs on a local LINUX server (4 CPUs, no GPUs) and is embedded in a larger image processing platform dedicated to stroke. Results: Processing time for a standard whole brain CT study (3-5mm slices) was around 2min. Upon completion, an instant notification (by email and/or mobile app) was sent to users to alert them about the suspected presence of a hemorrhage. Relative to neuroradiologist gold standard reads the algorithm’s sensitivity and specificity is 90.4% and 92.5% (95% CI: 85%-94% for both). Detection of acute intracranial hemorrhage can be automatized by deploying deep learning. It yielded very high sensitivity/specificity when compared to gold standard reads by a neuroradiologist. Volumes as small as 0.5mL could be detected reliably in the test dataset. The software can be deployed in busy practices to prioritize worklists and alert health care professionals to speed up therapeutic decision processes and interventions.

scholarly journals EVICAN—a balanced dataset for algorithm development in cell and nucleus segmentation

2020 ◽  
Vol 36 (12) ◽  
pp. 3863-3870
Author(s):  
Mischa Schwendy ◽  
Ronald E Unger ◽  
Sapun H Parekh
Keyword(s):  
Deep Learning ◽  
Cell Biology ◽  
Ground Truth ◽  
Training Data ◽  
Training Dataset ◽  
Visual Cell ◽  
Application Development ◽  
Ground Truth Data ◽  
Quantitative Image ◽  
Nucleus Segmentation

Abstract Motivation Deep learning use for quantitative image analysis is exponentially increasing. However, training accurate, widely deployable deep learning algorithms requires a plethora of annotated (ground truth) data. Image collections must contain not only thousands of images to provide sufficient example objects (i.e. cells), but also contain an adequate degree of image heterogeneity. Results We present a new dataset, EVICAN—Expert visual cell annotation, comprising partially annotated grayscale images of 30 different cell lines from multiple microscopes, contrast mechanisms and magnifications that is readily usable as training data for computer vision applications. With 4600 images and ∼26 000 segmented cells, our collection offers an unparalleled heterogeneous training dataset for cell biology deep learning application development. Availability and implementation The dataset is freely available (https://edmond.mpdl.mpg.de/imeji/collection/l45s16atmi6Aa4sI?q=). Using a Mask R-CNN implementation, we demonstrate automated segmentation of cells and nuclei from brightfield images with a mean average precision of 61.6 % at a Jaccard Index above 0.5.

scholarly journals Development of U-Net Breast Density Segmentation Method for Fat-Sat MR Images Using Transfer Learning Based on Non-Fat-Sat Model

2021 ◽  
Author(s):  
Yang Zhang ◽  
Siwa Chan ◽  
Jeon-Hor Chen ◽  
Kai-Ting Chang ◽  
Chin-Yao Lin ◽  
...  
Keyword(s):  
Deep Learning ◽  
Transfer Learning ◽  
Breast Tissue ◽  
Ground Truth ◽  
Training Dataset ◽  
Breast Volume ◽  
Segmentation Method ◽  
Tissue Segmentation ◽  
Testing Dataset ◽  
The Mean

AbstractTo develop a U-net deep learning method for breast tissue segmentation on fat-sat T1-weighted (T1W) MRI using transfer learning (TL) from a model developed for non-fat-sat images. The training dataset (N = 126) was imaged on a 1.5 T MR scanner, and the independent testing dataset (N = 40) was imaged on a 3 T scanner, both using fat-sat T1W pulse sequence. Pre-contrast images acquired in the dynamic-contrast-enhanced (DCE) MRI sequence were used for analysis. All patients had unilateral cancer, and the segmentation was performed using the contralateral normal breast. The ground truth of breast and fibroglandular tissue (FGT) segmentation was generated using a template-based segmentation method with a clustering algorithm. The deep learning segmentation was performed using U-net models trained with and without TL, by using initial values of trainable parameters taken from the previous model for non-fat-sat images. The ground truth of each case was used to evaluate the segmentation performance of the U-net models by calculating the dice similarity coefficient (DSC) and the overall accuracy based on all pixels. Pearson’s correlation was used to evaluate the correlation of breast volume and FGT volume between the U-net prediction output and the ground truth. In the training dataset, the evaluation was performed using tenfold cross-validation, and the mean DSC with and without TL was 0.97 vs. 0.95 for breast and 0.86 vs. 0.80 for FGT. When the final model developed with and without TL from the training dataset was applied to the testing dataset, the mean DSC was 0.89 vs. 0.83 for breast and 0.81 vs. 0.81 for FGT, respectively. Application of TL not only improved the DSC, but also decreased the required training case number. Lastly, there was a high correlation (R2 > 0.90) for both the training and testing datasets between the U-net prediction output and ground truth for breast volume and FGT volume. U-net can be applied to perform breast tissue segmentation on fat-sat images, and TL is an efficient strategy to develop a specific model for each different dataset.

scholarly journals Facilitating MR-Guided Adaptive Proton Therapy in Children Using Deep Learning-Based Synthetic CT

2021 ◽  
Author(s):  
Chuang Wang ◽  
Jinsoo Uh ◽  
Thomas E. Merchant ◽  
Chia-ho Hua ◽  
Sahaja Acharya
Keyword(s):  
Deep Learning ◽  
Brain Tumors ◽  
Proton Therapy ◽  
Hounsfield Unit ◽  
Ground Truth ◽  
Absolute Error ◽  
Training Dataset ◽  
Generative Adversarial Networks ◽  
Adaptive Planning ◽  
Patients Âgés

Abstract Purpose To determine whether self-attention cycle-generative adversarial networks (cycle-GANs), a novel deep-learning method, can generate accurate synthetic computed tomography (sCT) to facilitate adaptive proton therapy in children with brain tumors. Materials and Methods Both CT and T1-weighted magnetic resonance imaging (MRI) of 125 children (ages 1-20 years) with brain tumors were included in the training dataset. A model introducing a self-attention mechanism into the conventional cycle-GAN was created to enhance tissue interfaces and reduce noise. The test dataset consisted of 7 patients (ages 2-14 years) who underwent adaptive planning because of changes in anatomy discovered on MRI during proton therapy. The MRI during proton therapy-based sCT was compared with replanning CT (ground truth). Results The Hounsfield unit-mean absolute error was significantly reduced with self-attention cycle-GAN, as compared with conventional cycle-GAN (65.3 ± 13.9 versus 88.9 ± 19.3, P &lt; .01). The average 3-dimensional gamma passing rates (2%/2 mm criteria) for the original plan on the anatomy of the day and for the adapted plan were high (97.6% ± 1.2% and 98.9 ± 0.9%, respectively) when using sCT generated by self-attention cycle-GAN. The mean absolute differences in clinical target volume (CTV) receiving 95% of the prescription dose and 80% distal falloff along the beam axis were 1.1% ± 0.8% and 1.1 ± 0.9 mm, respectively. Areas of greatest dose difference were distal to the CTV and corresponded to shifts in distal falloff. Plan adaptation was appropriately triggered in all test patients when using sCT. Conclusion The novel cycle-GAN model with self-attention outperforms conventional cycle-GAN for children with brain tumors. Encouraging dosimetric results suggest that sCT generation can be used to identify patients who would benefit from adaptive replanning.

scholarly journals A NEW STEREO DENSE MATCHING BENCHMARK DATASET FOR DEEP LEARNING

2021 ◽  
Vol XLIII-B2-2021 ◽  
pp. 405-412
Author(s):  
T. Wu ◽  
B. Vallet ◽  
M. Pierrot-Deseilligny ◽  
E. Rupnik
Keyword(s):  
Deep Learning ◽  
Ground Truth ◽  
Fine Tuning ◽  
Training Dataset ◽  
Ground Truth Data ◽  
Aerial Photogrammetry ◽  
Dense Matching ◽  
Benchmark Datasets ◽  
3D Scene ◽  
The Impact

Abstract. Stereo dense matching is a fundamental task for 3D scene reconstruction. Recently, deep learning based methods have proven effective on some benchmark datasets, for example Middlebury and KITTI stereo. However, it is not easy to find a training dataset for aerial photogrammetry. Generating ground truth data for real scenes is a challenging task. In the photogrammetry community, many evaluation methods use digital surface models (DSM) to generate the ground truth disparity for the stereo pairs, but in this case interpolation may bring errors in the estimated disparity. In this paper, we publish a stereo dense matching dataset based on ISPRS Vaihingen dataset, and use it to evaluate some traditional and deep learning based methods. The evaluation shows that learning-based methods outperform traditional methods significantly when the fine tuning is done on a similar landscape. The benchmark also investigates the impact of the base to height ratio on the performance of the evaluated methods. The dataset can be found in https://github.com/whuwuteng/benchmark_ISPRS2021.

scholarly journals Direct pixel to pixel principal strain mapping from tagging MRI using end to end deep convolutional neural network (DeepStrain)

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Khaled Z. Abd-Elmoniem ◽  
Inas A. Yassine ◽  
Nader S. Metwalli ◽  
Ahmed Hamimi ◽  
Ronald Ouwerkerk ◽  
...  
Keyword(s):  
Neural Network ◽  
Monte Carlo ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Ground Truth ◽  
Image Resolution ◽  
Principal Strain ◽  
Strain Mapping ◽  
End To End

AbstractRegional soft tissue mechanical strain offers crucial insights into tissue's mechanical function and vital indicators for different related disorders. Tagging magnetic resonance imaging (tMRI) has been the standard method for assessing the mechanical characteristics of organs such as the heart, the liver, and the brain. However, constructing accurate artifact-free pixelwise strain maps at the native resolution of the tagged images has for decades been a challenging unsolved task. In this work, we developed an end-to-end deep-learning framework for pixel-to-pixel mapping of the two-dimensional Eulerian principal strains $$\varvec{{\varepsilon }}_{\boldsymbol{p1}}$$ ε p 1 and $$\varvec{{\varepsilon }}_{\boldsymbol{p2}}$$ ε p 2 directly from 1-1 spatial modulation of magnetization (SPAMM) tMRI at native image resolution using convolutional neural network (CNN). Four different deep learning conditional generative adversarial network (cGAN) approaches were examined. Validations were performed using Monte Carlo computational model simulations, and in-vivo datasets, and compared to the harmonic phase (HARP) method, a conventional and validated method for tMRI analysis, with six different filter settings. Principal strain maps of Monte Carlo tMRI simulations with various anatomical, functional, and imaging parameters demonstrate artifact-free solid agreements with the corresponding ground-truth maps. Correlations with the ground-truth strain maps were R = 0.90 and 0.92 for the best-proposed cGAN approach compared to R = 0.12 and 0.73 for the best HARP method for $$\varvec{{\varepsilon }}_{\boldsymbol{p1}}$$ ε p 1 and $$\varvec{{\varepsilon }}_{\boldsymbol{p2}}$$ ε p 2 , respectively. The proposed cGAN approach's error was substantially lower than the error in the best HARP method at all strain ranges. In-vivo results are presented for both healthy subjects and patients with cardiac conditions (Pulmonary Hypertension). Strain maps, obtained directly from their corresponding tagged MR images, depict for the first time anatomical, functional, and temporal details at pixelwise native high resolution with unprecedented clarity. This work demonstrates the feasibility of using the deep learning cGAN for direct myocardial and liver Eulerian strain mapping from tMRI at native image resolution with minimal artifacts.

Four-dimensional dose reconstruction through in vivo phase matching of cine images of electronic portal imaging device

2016 ◽  
Vol 43 (7) ◽  
pp. 4420-4430 ◽  
Author(s):  
Jihyung Yoon ◽  
Jae Won Jung ◽  
Jong Oh Kim ◽  
Byong Yong Yi ◽  
Inhwan Yeo
Keyword(s):  
Phase Matching ◽  
Electronic Portal Imaging Device ◽  
Dose Reconstruction ◽  
Imaging Device ◽  
Portal Imaging ◽  
Electronic Portal Imaging

scholarly journals Biopsy-free in vivo virtual histology of skin using deep learning

2021 ◽  
Vol 10 (1) ◽  
Author(s):  
Jingxi Li ◽  
Jason Garfinkel ◽  
Xiaoran Zhang ◽  
Di Wu ◽  
Yijie Zhang ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Ex Vivo ◽  
Normal Skin ◽  
Ground Truth ◽  
Skin Neoplasms ◽  
Cellular Level ◽  
Label Free ◽  
Virtual Histology

AbstractAn invasive biopsy followed by histological staining is the benchmark for pathological diagnosis of skin tumors. The process is cumbersome and time-consuming, often leading to unnecessary biopsies and scars. Emerging noninvasive optical technologies such as reflectance confocal microscopy (RCM) can provide label-free, cellular-level resolution, in vivo images of skin without performing a biopsy. Although RCM is a useful diagnostic tool, it requires specialized training because the acquired images are grayscale, lack nuclear features, and are difficult to correlate with tissue pathology. Here, we present a deep learning-based framework that uses a convolutional neural network to rapidly transform in vivo RCM images of unstained skin into virtually-stained hematoxylin and eosin-like images with microscopic resolution, enabling visualization of the epidermis, dermal-epidermal junction, and superficial dermis layers. The network was trained under an adversarial learning scheme, which takes ex vivo RCM images of excised unstained/label-free tissue as inputs and uses the microscopic images of the same tissue labeled with acetic acid nuclear contrast staining as the ground truth. We show that this trained neural network can be used to rapidly perform virtual histology of in vivo, label-free RCM images of normal skin structure, basal cell carcinoma, and melanocytic nevi with pigmented melanocytes, demonstrating similar histological features to traditional histology from the same excised tissue. This application of deep learning-based virtual staining to noninvasive imaging technologies may permit more rapid diagnoses of malignant skin neoplasms and reduce invasive skin biopsies.

scholarly journals Deep Learning-Based Analysis of Macaque Corneal Sub-Basal Nerve Fibers in Confocal Microscopy Images

2019 ◽  
Author(s):  
Jonathan D. Oakley ◽  
Daniel B. Russakoff ◽  
Megan E. McCarron ◽  
Rachel L. Weinberg ◽  
Jessica M. Izzi ◽  
...  
Keyword(s):  
Deep Learning ◽  
Confocal Microscopy ◽  
Nerve Fiber ◽  
Ground Truth ◽  
Automated Analysis ◽  
Nerve Fibers ◽  
Fiber Analysis ◽  
Quantitative Measurements ◽  
Nerve Fiber Analysis

AbstractPurposeTo describe and assess different deep learning-based methods for automated measurement of macaque corneal sub-basal nerves using in vivo confocal microscopy (IVCM).MethodsThe automated assessment of corneal nerve fiber length (CNFL) in IVCM images is of increasing clinical interest. These measurements are important biomarkers in a number of diseases including diabetes mellitus, human immunodeficiency virus, Parkinson’s disease and multiple sclerosis. Animal models of these and other diseases play an important role in understanding the disease processes as efforts toward developing new and effective therapeutics are made. And while automated methods exist for nerve fiber analysis in clinical data, differences in anatomy and image quality make the macaque data more challenging and has motivated the work reported here.Toward this goal, nerves in macaque corneal IVCM images were manually labelled using an ImageJ plugin (NeuronJ). Different deep convolutional neural network (CNN) architectures were evaluated for accuracy relative to the ground truth manual tracings. The best performing model was used on separately acquired macaque ICVM images to additionally compare inter-reader variability.ConclusionsDeep learning-based segmentation of sub-basal nerves in IVCM images shows excellent correlation to manual segmentations in macaque data. The technique is indistinguishable across readers and paves the way for more widespread adoption of objective automated analysis of sub-basal nerves in IVCM.Translational RelevanceQuantitative measurements of corneal sub-basal nerves are important biomarkers for disease screening and management. This work reports on different approaches that, in using deep learning-based techniques, leverage state of the art analysis methods to demonstrate performance akin to human graders. In application, the approach is robust, rapid and objective, offering utility to a variety of clinical studies using IVCM.

scholarly journals Using Deep Learning to Count Albatrosses from Space: Assessing Results in Light of Ground Truth Uncertainty

2020 ◽  
Vol 12 (12) ◽  
pp. 2026 ◽  
Author(s):  
Ellen Bowler ◽  
Peter T. Fretwell ◽  
Geoffrey French ◽  
Michal Mackiewicz
Keyword(s):  
Deep Learning ◽  
Network Performance ◽  
Cross Validation ◽  
State Of The Art ◽  
Ground Truth ◽  
Training Dataset ◽  
Learning Approaches ◽  
Wildlife Species ◽  
Training Schemes ◽  
Very High

Many wildlife species inhabit inaccessible environments, limiting researchers ability to conduct essential population surveys. Recently, very high resolution (sub-metre) satellite imagery has enabled remote monitoring of certain species directly from space; however, manual analysis of the imagery is time-consuming, expensive and subjective. State-of-the-art deep learning approaches can automate this process; however, often image datasets are small, and uncertainty in ground truth labels can affect supervised training schemes and the interpretation of errors. In this paper, we investigate these challenges by conducting both manual and automated counts of nesting Wandering Albatrosses on four separate islands, captured by the 31 cm resolution WorldView-3 sensor. We collect counts from six observers, and train a convolutional neural network (U-Net) using leave-one-island-out cross-validation and different combinations of ground truth labels. We show that (1) interobserver variation in manual counts is significant and differs between the four islands, (2) the small dataset can limit the networks ability to generalise to unseen imagery and (3) the choice of ground truth labels can have a significant impact on our assessment of network performance. Our final results show the network detects albatrosses as accurately as human observers for two of the islands, while in the other two misclassifications are largely caused by the presence of noise, cloud cover and habitat, which was not present in the training dataset. While the results show promise, we stress the importance of considering these factors for any study where data is limited and observer confidence is variable.

Sign in / Sign up

Export Citation Format

Share Document