Shape-Aware Complementary-Task Learning for Multi-organ Segmentation

Navarro, Fernando A.; Shit, Suprosanna; Ezhov, Ivan; Paetzold, Johannes C.; Gafita, Andrei; Peeken, Jan C.; Combs, Stephanie E.; Menze, Bjoern

doi:10.1007/978-3-030-32692-0_71

Cited by 58 publications

(37 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In other words, we use an architecture, which is very similar to the commonly used 3D U-Net architecture with some modifications [1]. We decide to use U-Net, because this is a very successful architecture for diverse medical imaging tasks [3,5,8,10,11,15]. We refer to our solution as Aν-net, which are homophones of 'Aneu-net' and 'an U-net.'…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A$$\nu $$-Net: Automatic Detection and Segmentation of Aneurysm

Shit¹,

Ezhov²,

Paetzold³

et al. 2021

Cerebral Aneurysm Detection

Self Cite

View full text Add to dashboard Cite

We propose an automatic solution for the CADA 2020 challenge to detect aneurysm from Digital Subtraction Angiography (DSA) images. Our method relies on 3D U-net as the backbone and heavy data augmentation with a carefully chosen loss function. We were able to generalize well using our solution (despite training on a small dataset) that is demonstrated through accurate detection and segmentation on the test data.

show abstract

Section: Methodsmentioning

confidence: 99%

“…Since Dice loss provides an edge at handling class imbalance [8,10] and crossentropy loss is beneficial for smooth training convergence [9,14], we use both. The total loss function of our method is as follows:…”

Section: Methodsmentioning

confidence: 99%

A$$\nu $$-Net: Automatic Detection and Segmentation of Aneurysm

Shit¹,

Ezhov²,

Paetzold³

et al. 2021

Cerebral Aneurysm Detection

Self Cite

View full text Add to dashboard Cite

show abstract

“…IBBM. We use three 2D U-Net 44,[53][54][55] segmentation architectures with some modifications. We implemented the 2D segmentation architecture separately for slices of all three orientations (axial, coronal and sagittal) of the 3D dataset (analogous to Guha Roy et al 56 ).…”

Section: Technical Validationmentioning

confidence: 99%

An automatic multi-tissue human fetal brain segmentation benchmark using the Fetal Tissue Annotation Dataset

et al. 2021

Self Cite

View full text Add to dashboard Cite

It is critical to quantitatively analyse the developing human fetal brain in order to fully understand neurodevelopment in both normal fetuses and those with congenital disorders. To facilitate this analysis, automatic multi-tissue fetal brain segmentation algorithms are needed, which in turn requires open datasets of segmented fetal brains. Here we introduce a publicly available dataset of 50 manually segmented pathological and non-pathological fetal magnetic resonance brain volume reconstructions across a range of gestational ages (20 to 33 weeks) into 7 different tissue categories (external cerebrospinal fluid, grey matter, white matter, ventricles, cerebellum, deep grey matter, brainstem/spinal cord). In addition, we quantitatively evaluate the accuracy of several automatic multi-tissue segmentation algorithms of the developing human fetal brain. Four research groups participated, submitting a total of 10 algorithms, demonstrating the benefits the dataset for the development of automatic algorithms.

show abstract

“…Some studies found superior predictive performances using CNNs compared to handcrafted features [25,26]. Radiomics-based approaches also enable localization and segmentation of volumes of interest (VOI) [27,28].…”

Section: Introductionmentioning

confidence: 99%

Development and External Validation of Deep-Learning-Based Tumor Grading Models in Soft-Tissue Sarcoma Patients Using MR Imaging

Navarro

Dapper

Asadpour

et al. 2021

Cancers

Self Cite

View full text Add to dashboard Cite

Background: In patients with soft-tissue sarcomas, tumor grading constitutes a decisive factor to determine the best treatment decision. Tumor grading is obtained by pathological work-up after focal biopsies. Deep learning (DL)-based imaging analysis may pose an alternative way to characterize STS tissue. In this work, we sought to non-invasively differentiate tumor grading into low-grade (G1) and high-grade (G2/G3) STS using DL techniques based on MR-imaging. Methods: Contrast-enhanced T1-weighted fat-saturated (T1FSGd) MRI sequences and fat-saturated T2-weighted (T2FS) sequences were collected from two independent retrospective cohorts (training: 148 patients, testing: 158 patients). Tumor grading was determined following the French Federation of Cancer Centers Sarcoma Group in pre-therapeutic biopsies. DL models were developed using transfer learning based on the DenseNet 161 architecture. Results: The T1FSGd and T2FS-based DL models achieved area under the receiver operator characteristic curve (AUC) values of 0.75 and 0.76 on the test cohort, respectively. T1FSGd achieved the best F1-score of all models (0.90). The T2FS-based DL model was able to significantly risk-stratify for overall survival. Attention maps revealed relevant features within the tumor volume and in border regions. Conclusions: MRI-based DL models are capable of predicting tumor grading with good reproducibility in external validation.

show abstract

Shape-Aware Complementary-Task Learning for Multi-organ Segmentation

Cited by 58 publications

References 11 publications

A$$\nu $$-Net: Automatic Detection and Segmentation of Aneurysm

A$$\nu $$-Net: Automatic Detection and Segmentation of Aneurysm

An automatic multi-tissue human fetal brain segmentation benchmark using the Fetal Tissue Annotation Dataset

Development and External Validation of Deep-Learning-Based Tumor Grading Models in Soft-Tissue Sarcoma Patients Using MR Imaging

Contact Info

Product

Resources

About