GTCreator: a flexible annotation tool for image-based datasets

Bernal, Jorge; Histace, Aymeric; Masana, Marc; Angermann, Quentin; Sánchez-Montes, Cristina; Miguel, Cristina Rodríguez de; Hammami, Maroua; García-Rodríguez, Ana; Córdova, Henry; Romain, Olivier; Fernández‐Esparrach, Glòria; Dray, Xavier; Sánchez, Francisco Javier Martín

doi:10.1007/s11548-018-1864-x

Cited by 33 publications

(30 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several tools have been proposed to assist in ground truth generation for specific image annotation tools and include Label Me (MIT, Computer Science and Artificial Intelligence Laboratory, Cambridge, Mass, USA), 34 VGG image annotator (Visual Geometry Group, Department of Engineering Science, University of Oxford, Oxford, UK), 35 and GTCreator. 36 However, none of the presently available labeling tools is purpose-built with a physician, let alone a GI endoscopist, in mind. The second challenge is that accurate labeling requires an expert, in this case a gastroenterologist, to review the images and provide the ground truth by hand-annotating and describing the key areas of interest, using standardized terminology and labels.…”

Section: Data Science Priorities Endoscopy Image and Video Library: Ementioning

confidence: 99%

Position statement on priorities for artificial intelligence in GI endoscopy: a report by the ASGE Task Force

Berzin

Parasa

Wallace

et al. 2020

Gastrointestinal Endoscopy

View full text Add to dashboard Cite

Artificial intelligence (AI) in GI endoscopy holds tremendous promise to augment clinical performance, establish better treatment plans, and improve patient outcomes. Although there are promising initial applications and preliminary clinical data for AI in gastroenterology, the field is still in a very early phase, with limited clinical use. The American Society for Gastrointestinal Endoscopy has convened an AI Task Force to develop guidance around clinical implementation, testing/validating algorithms, and building pathways for successful implementation of AI in GI endoscopy. This White Paper focuses on 3 areas: (1) priority use cases for development of AI algorithms in GI, both for specific clinical scenarios and for streamlining clinical workflows, quality reporting, and practice management; (2) data science priorities, including development of image libraries, and standardization of methods for storing, sharing, and annotating endoscopic images/video; and (3) research priorities, focusing on the importance of high-quality, prospective trials measuring clinically meaningful patient outcomes. (Gastrointest Endosc 2020;92:951-9.) Artificial intelligence (AI) is poised to play a crucial role throughout the practice of gastroenterology. Machine learning (ML) is a branch of AI focusing on computer algorithms that can learn from data and perform specific tasks and analyses, often without explicit human programming. Deep learning is an advanced subset of ML that relies on specific algorithms termed artificial neural networks. The enthusiasm for medical AI applications has centered around expectations that these technologies may assist physicians by assimilating, triaging, and interpreting clinical data (laboratory results, electrocardiographic tracings, CTs, etc) to support physician performance in a consistent manner in all phases of clinical care and automate certain burdensome tasks, such as clinical documentation and billing, to allow physicians to refocus on direct patient care and other meaningful professional activities.

show abstract

Section: Data Science Priorities Endoscopy Image and Video Library: Ementioning

confidence: 99%

Position statement on priorities for artificial intelligence in GI endoscopy: a report by the ASGE Task Force

Berzin

Parasa

Wallace

et al. 2020

Gastrointestinal Endoscopy

View full text Add to dashboard Cite

show abstract

“…Experiments were run on a NVIDIA GTX 1080 GPU with 8 GB memory. The network has been pretrained using CVC-VideoClinicDB [21,22], whose polyp masks are not precise but approximated to elliptical shapes. The datasets in "Transformations" section are then used to finetune this pretrained model with fixed parameters for all experiments:…”

Section: Transformationsmentioning

confidence: 99%

“…On the one side, and in order to increase the size of the training set, a first approach would be to increase the number of annotated samples by experts. In this regard, efforts are been focused on developing tools which facilitates the manual annotation of images, such as GTCreatorTool [22], which is a flexible annotation tool which minimizes annotation time and allows for sharing annotations among experts. Beyond the transformations analysed in this paper, other alternatives would be to add polyps in nonpolypoid samples [41] or more advances approaches such as emulating data augmentation during learning by the image generation through a hetero-encoder [42].…”

Section: Transformations Whose Effect On Performance Dependsmentioning

confidence: 99%

“…Computer-assisted diagnosis (CAD) systems for early detection of colorectal cancer have also benefited from the application of deep learning techniques [16][17][18]. Publicly available datasets range from hundreds of images with a manually segmented binary mask, such as CVC-EndoSceneStill [19] or Kvasir-SEG [20], to thousands of video frames with an approximated elliptical binary mask, such as CVC-VideoClinicDB [21,22]. For polyp segmentation, it is easy to find several works in which data augmentation has been used.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Unravelling the effect of data augmentation transformations in polyp segmentation

et al. 2020

View full text Add to dashboard Cite

Purpose Data augmentation is a common technique to overcome the lack of large annotated databases, a usual situation when applying deep learning to medical imaging problems. Nevertheless, there is no consensus on which transformations to apply for a particular field. This work aims at identifying the effect of different transformations on polyp segmentation using deep learning. Methods A set of transformations and ranges have been selected, considering image-based (width and height shift, rotation, shear, zooming, horizontal and vertical flip and elastic deformation), pixel-based (changes in brightness and contrast) and application-based (specular lights and blurry frames) transformations. A model has been trained under the same conditions without data augmentation transformations (baseline) and for each of the transformation and ranges, using CVC-EndoSceneStill and Kvasir-SEG, independently. Statistical analysis is performed to compare the baseline performance against results of each range of each transformation on the same test set for each dataset. Results This basic method identifies the most adequate transformations for each dataset. For CVC-EndoSceneStill, changes in brightness and contrast significantly improve the model performance. On the contrary, Kvasir-SEG benefits to a greater extent from the image-based transformations, especially rotation and shear. Augmentation with synthetic specular lights also improves the performance. Conclusion Despite being infrequently used, pixel-based transformations show a great potential to improve polyp segmentation in CVC-EndoSceneStill. On the other hand, image-based transformations are more suitable for Kvasir-SEG. Problem-based transformations behave similarly in both datasets. Polyp area, brightness and contrast of the dataset have an influence on these differences.

show abstract

“…Although there are some publicly available datasets (e.g. [17]- [21]), higher quality and a larger quantity of fully annotated datasets of polyp images and videos are highly desirable [14], [15]. Unlike a still frame dataset, a database of polyp videos can preserve temporal dependencies among frames.…”

Section: Introductionmentioning

confidence: 99%

A Framework With a Fully Convolutional Neural Network for Semi-Automatic Colon Polyp Annotation

et al. 2019

View full text Add to dashboard Cite

Deep learning has delivered promising results for automatic polyp detection and segmentation. However, deep learning is known for being data-hungry, and its performance is correlated with the amount of available training data. The lack of large labeled polyp training images is one of the major obstacles in performance improvement of automatic polyp detection and segmentation. Labeling is typically performed by an endoscopist, who performs pixel-level annotation of polyps. Manual polyp labeling of a video sequence is difficult and time-consuming. We propose a semi-automatic annotation framework powered by a convolutional neural network (CNN) to speed up polyp annotation in video-based datasets. Our CNN network requires only ground-truth (manually annotated masks) of a few frames in a video for training and annotating the rest of the frames in a semi-supervised manner. To generate masks similar to the ground-truth masks, we use some pre and post-processing steps such as different data augmentation strategies, morphological operations, Fourier descriptors, and a second stage fine-tuning. We use Fourier coefficients of the ground-truth masks to select similar generated output masks. The results show that it is possible to 1) produce ∼ 96% of Dice similarity score between the polyp masks provided by clinicians and the masks generated by our framework, and 2) save clinicians time as they need to manually annotate only a few frames instead of annotating the entire video, frame-by-frame.

show abstract

GTCreator: a flexible annotation tool for image-based datasets

Cited by 33 publications

References 9 publications

Position statement on priorities for artificial intelligence in GI endoscopy: a report by the ASGE Task Force

Position statement on priorities for artificial intelligence in GI endoscopy: a report by the ASGE Task Force

Unravelling the effect of data augmentation transformations in polyp segmentation

A Framework With a Fully Convolutional Neural Network for Semi-Automatic Colon Polyp Annotation

Contact Info

Product

Resources

About