MDCT Angiography With 3D Rendering: A Novel Cinematic Rendering Algorithm for Enhanced Anatomic Detail

Johnson, Pamela T.; Schneider, Robert; Lugo-Fagundo, Carolina; Johnson, Michael B.; Fishman, Elliot K.

doi:10.2214/ajr.17.17903

Cited by 81 publications

(41 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Compared to conventional medical rendering, this technique considerably enhances depth and shape perception by putting the anatomical structures within the medical scans in a virtual lighting condition that mimics the physical lighting experienced in reality. Cinematic rendering has been used for a variety of medical imaging visualization tasks [23,41,4,40].…”

Section: Cinematically Rendered Data For Fine-tuningmentioning

confidence: 99%

Deep learning with cinematic rendering: fine-tuning deep neural networks using photorealistic medical images

et al. 2018

View full text Add to dashboard Cite

Deep learning has emerged as a powerful artificial intelligence tool to interpret medical images for a growing variety of applications. However, the paucity of medical imaging data with high-quality annotations that is necessary for training such methods ultimately limits their performance. Medical data is challenging to acquire due to privacy issues, shortage of experts available for annotation, limited representation of rare conditions and cost. This problem has previously been addressed by using synthetically generated data. However, networks trained on synthetic data often fail to generalize to real data. Cinematic rendering simulates the propagation and interaction of light passing through tissue models reconstructed from CT data, enabling the generation of photorealistic images. In this paper, we present one of the first applications of cinematic rendering in deep learning, in which we propose to fine-tune synthetic data-driven networks using cinematically rendered CT data for the task of monocular depth estimation in endoscopy. Our experiments demonstrate that: (a) convolutional neural networks (CNNs) trained on synthetic data and fine-tuned on photorealistic cinematically rendered data adapt better to real medical images and demonstrate more robust performance when compared to networks with no fine-tuning, (b) these fine-tuned networks require less training data to converge to an optimal solution, and (c) fine-tuning with data from a variety of photorealistic rendering conditions of the same scene prevents the network from learning patient-specific information and aids in generalizability of the model. Our empirical evaluation demonstrates that networks fine-tuned with cinematically rendered data predict depth with 56.87% less error for rendered endoscopy images and 27.49% less error for real porcine colon endoscopy images.

show abstract

Section: Cinematically Rendered Data For Fine-tuningmentioning

confidence: 99%

Deep learning with cinematic rendering: fine-tuning deep neural networks using photorealistic medical images

et al. 2018

View full text Add to dashboard Cite

show abstract

“…The global lighting model is the basis for the photorealism in CR images. 5 Additional considerations that refine CR photorealism include the incorporation of shadowing effects that derive from assuming that the light source can be obscured by objects within the visualized volume. Such shadowing is not typically included in traditional VR, although it is possible to incorporate.…”

Section: Cinematic Rendering Image Creationmentioning

confidence: 99%

“…texture maps that are related to the brightnesses of surfaces within a virtual scene and more accurately reflect realistic lighting scenarios). 5 Ultimately, real-time interaction between the radiologist and the workstation is needed to optimize CR visualizations, just as is true with any 3D methodology. Appropriate selection of window presets using either a ramp (for tissues differing markedly in attenuation such as bone in comparison to adjacent soft tissue) or trapezoid method (for differentiation of similar attenuation structures such as adjacent soft tissues) and slab thickness will allow for appropriate display of pathology and will ensure that a finding of interest is not obscured by an overlying structure.…”

Section: Cinematic Rendering Image Creationmentioning

confidence: 99%

“…Recently, a new rendering method has become available that is similar to VR, but incorporates a more complex lighting model to create photorealistic images. This method, known as cinematic rendering (CR), [3][4][5] has not yet been widely studied but has shown initial promise in the depiction of complex anatomy and pathology in a range of different body regions. 6,7 As with traditional VR, this new technique provides particularly detailed visualizations of high density objects, including contrastenhanced vessels and highly vascularized structures, as well as soft tissues, organs and musculoskeletal structures.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cinematic rendering of cardiac CT volumetric data: Principles and initial observations

Rowe

Johnson

Fishman

2018

Journal of Cardiovascular Computed Tomography

Self Cite

View full text Add to dashboard Cite

show abstract

“…Given the complexity of placental and fetal anatomy, visualizing any acquired CT dataset in 3D may provide valuable insight [5] . In this report, we describe our first experience with fetal and placental imaging with CT utilizing a new 3D cinematic rendering (CR) technique that has recently been U.S. Food and Drug Administration approved and that allows for evaluation of anatomic detail at a photorealistic level [6] , [7] , [8] .…”

Section: Introductionmentioning

confidence: 99%

Fetal and placental anatomy visualized with cinematic rendering from volumetric CT data

Rowe

Fishman

2018

Radiology Case Reports

Self Cite

View full text Add to dashboard Cite

Avoiding unnecessary radiation exposure in children, including fetuses or embryos, is of paramount importance. However, emergent clinical situations will arise that necessitate the use of ionizing radiation-based modalities, such as computed tomography (CT), in this patient population. In such circumstances, the use of advanced visualization methods may provide optimum diagnostic utility. We present the case of a pregnant patient with Loeys-Dietz syndrome who was evaluated with CT angiography to rule out an acute aortic syndrome. The CT data from the fetus and placenta were reconstructed using the new cinematic rendering technique that allows for photorealistic display. The potential advantages of cinematic rendering relative to traditional volume rendering are discussed.

show abstract

MDCT Angiography With 3D Rendering: A Novel Cinematic Rendering Algorithm for Enhanced Anatomic Detail

Cited by 81 publications

References 7 publications

Deep learning with cinematic rendering: fine-tuning deep neural networks using photorealistic medical images

Deep learning with cinematic rendering: fine-tuning deep neural networks using photorealistic medical images

Cinematic rendering of cardiac CT volumetric data: Principles and initial observations

Fetal and placental anatomy visualized with cinematic rendering from volumetric CT data

Contact Info

Product

Resources

About