SATr: Slice Attention with Transformer for Universal Lesion Detection

Li, Han; Chen, Long; Han, Hu; Zhou, S. Kevin

doi:10.1007/978-3-031-16437-8_16

Cited by 22 publications

(12 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multiorgan/universal [32,36,40,61], partially-label [16,80,84], multi-domain [25,44] and multi-task [7,21] algorithms are emerging, but none of them addresses multi-cancer detection and diagnosis problems. DeepLesion [73] attempts to tackle the universal lesion detection task in CT scans, but their derived lesion detection methods [69,70,72] and several follow-up work [35,41,52,53,71,74] so far have reported mostly moderate multi-class lesion detection performance. Distinguishing between malignant and benign lesions in multi-class tumor setting is still far from a clinical reality.…”

Section: Related Workmentioning

confidence: 99%

“…Transformers [57] have advanced the state-of-the-art performance in various computer vision tasks [6,9,19,20,38,55,86], by capturing global interactions between image patches and having no built-in inductive prior. The success of Transformer has also been witnessed in medical image detection [34] and segmentation [8,22,67]. With the recent progress in transformers [6,59], a new variant called mask Transformers has been proposed, where segmentation predictions are represented by a set of query embeddings with their own semantic class labels, generated through the conversion of query embedding to mask embedding vectors followed by multiplying with the image features.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Towards a Single Unified Model for Effective Detection, Segmentation, and Diagnosis of Eight Major Cancers Using a Large Collection of CT Scans

Chen¹,

Xia²,

Yao³

et al. 2023

Preprint

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Towards a Single Unified Model for Effective Detection, Segmentation, and Diagnosis of Eight Major Cancers Using a Large Collection of CT Scans

Chen¹,

Xia²,

Yao³

et al. 2023

Preprint

View full text Add to dashboard Cite

“…Generally, we record all training samples loss and uncertainty of 2 SOTA ULD methods based on 25%, 50%, and 100% training data, and draw their value or their indexes into 2D dashed figures. We hereby show the results based on [54] in Fig. 3, other results are shown in the supplementary materials.…”

Section: The Relationship Analysis Between Loss and Uncertaintymentioning

confidence: 99%

“…Figure 3. Illustration of loss and uncertainty relationship based on SATr[54]. Yellow (or cyan) points denote the sample whose absolute difference between uncertainty and loss is greater (or less than) than 0.3.…”

mentioning

confidence: 99%

Mixed-order self-paced curriculum learning for universal lesion detection

Li¹,

Huang²,

Zhou³

2023

Preprint

View full text Add to dashboard Cite

Self-paced curriculum learning (SCL) has demonstrated its great potential in computer vision, natural language processing, etc. During training, it implements easy-to-hard sampling based on online estimation of data difficulty. Most SCL methods commonly adopt a loss-based strategy of estimating data difficulty and deweight the 'hard' samples in the early training stage. While achieving success in a variety of applications, SCL stills confront two challenges in a medical image analysis task, such as universal lesion detection, featuring insufficient and highly class-imbalanced data: (i) the loss-based difficulty measurer is inaccurate; ii) the hard samples are under-utilized from a deweighting mechanism. To overcome these challenges, in this paper we propose a novel mixed-order self-paced curriculum learning (Mo-SCL) method. We integrate both uncertainty and loss to better estimate difficulty online and mix both hard and easy samples in the same mini-batch to appropriately alleviate the problem of under-utilization of hard samples. We provide a theoretical investigation of our method in the context of stochastic gradient descent optimization and extensive experiments based on the DeepLesion benchmark dataset for universal lesion detection (ULD). When applied for two state-of-the-art ULD methods, the proposed mixed-order SCL method can provide a free boost to lesion detection accuracy without extra special network designs.

show abstract

“…[4][5][6] As the 3D lesion extent can be a potential indicator of response to treatment, it is crucial to detect and label the 3D lesions, such that their size can be accurately measured based on current guidelines. Prior approaches [3][4][5][7][8][9][10][11][12][13][14] for 3D lesion detection used the publicly available DeepLesion dataset, 7 but it contains incomplete annotations 5,6,15 and severe class imbalances. 6,15 In this work, we designed a self-training pipeline for 3D lesion detection and tagging.…”

Section: Introductionmentioning

confidence: 99%

3D universal lesion detection and tagging in CT with self-training

Frazier¹,

Mathai²,

Liu³

et al. 2023

Medical Imaging 2023: Computer-Aided Diagnosis

View full text Add to dashboard Cite

Radiologists routinely perform the tedious task of lesion localization, classification, and size measurement in computed tomography (CT) studies. Universal lesion detection and tagging (ULDT) can simultaneously help alleviate the cumbersome nature of lesion measurement and enable tumor burden assessment. Previous ULDT approaches utilize the publicly available DeepLesion dataset, however it does not provide the full volumetric (3D) extent of lesions and also displays a severe class imbalance. In this work, we propose a self-training pipeline to detect 3D lesions and tag them according to the body part they occur in. We used a significantly limited 30% subset of DeepLesion to train a VFNet model for 2D lesion detection and tagging. Next, the 2D lesion context was expanded into 3D, and the mined 3D lesion proposals were integrated back into the baseline training data in order to retrain the model over multiple rounds. Through the self-training procedure, our VFNet model learned from its own predictions, detected lesions in 3D, and tagged them. Our results indicated that our VFNet model achieved an average sensitivity of 46.9% at [0.125:8] false positives (FP) with a limited 30% data subset in comparison to the 46.8% of an existing approach that used the entire DeepLesion dataset. To our knowledge, we are the first to jointly detect lesions in 3D and tag them according to the body part label.

show abstract

SATr: Slice Attention with Transformer for Universal Lesion Detection

Cited by 22 publications

References 34 publications

Towards a Single Unified Model for Effective Detection, Segmentation, and Diagnosis of Eight Major Cancers Using a Large Collection of CT Scans

Towards a Single Unified Model for Effective Detection, Segmentation, and Diagnosis of Eight Major Cancers Using a Large Collection of CT Scans

Mixed-order self-paced curriculum learning for universal lesion detection

3D universal lesion detection and tagging in CT with self-training

Contact Info

Product

Resources

About