End-to-End Multi-Task Learning for Lung Nodule Segmentation and Diagnosis

Chen, Wei; Wang, Qiuli; Yang, Dan; Zhang, Xiaohong; Liu, Chen; Li, Yucong

doi:10.1109/icpr48806.2021.9412218

Cited by 7 publications

(6 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Meanwhile, the accuracy of predicting malignancy approaches X-Caps [6] and already exceeds HSCNN [5], which uses 3D volume data. When using 10% annotations, our malignancy prediction accuracy surpasses all other explainable competitors using full annotations, among which MSN-JCN [15] is heavily supervised by additional information.…”

Section: Prediction Performance Of Nodule Attributes and Malignancymentioning

confidence: 81%

See 1 more Smart Citation

Reducing Annotation Need in Self-explanatory Models for Lung Nodule Diagnosis

Lu¹,

Chen²,

Krause³

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Recently, attempts have been made to reduce annotation requirements in feature-based self-explanatory models for lung nodule diagnosis. As a representative, cRedAnno achieves competitive performance with considerably reduced annotation needs by introducing self-supervised contrastive learning to do unsupervised feature extraction. However, it exhibits unstable performance under scarce annotation conditions. To improve the accuracy and robustness of cRedAnno, we propose an annotation exploitation mechanism by conducting semi-supervised active learning in the learned semantically meaningful space to jointly utilise the extracted features, annotations, and unlabelled data. The proposed approach achieves comparable or even higher malignancy prediction accuracy with 10x fewer annotations, meanwhile showing better robustness and nodule attribute prediction accuracy. Our complete code is open-source available: https: //github.com/diku-dk/credanno.

show abstract

Section: Prediction Performance Of Nodule Attributes and Malignancymentioning

confidence: 81%

“…The predictions of nodule attributes are considered correct if within ±1 of aggregated radiologists' annotation [6]. Attribute "internal structure" is excluded from the results because its heavily imbalanced classes are not very informative [2,5,6,15,16].…”

Section: Prediction Performance Of Nodule Attributes and Malignancymentioning

confidence: 99%

Reducing Annotation Need in Self-explanatory Models for Lung Nodule Diagnosis

Lu¹,

Chen²,

Krause³

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…They reported slightly improved performance when they applied this segmentation approach to the segmentation of lung nodules. Chen et al [20] proposed end-to-end multi-task learning framework which consists of joint classification and multi-channel segmentation networks. Both networks utilized the same latent representation learned by the common encoder branch which improved the lung segmentation performance.…”

Section: Related Workmentioning

confidence: 99%

DEHA-Net: Dual-Encoder based Hard Attention Network with Adaptive ROI Mechanism for Lung Nodule Segmentation

Usman¹,

Shin²

2022

Preprint

View full text Add to dashboard Cite

Measuring pulmonary nodules accurately can help with early diagnosis of lung cancer, which can improve a patient’s chances of survival. Many methods for segmenting nodules have been developed, but they all rely on input from radiologists in the form of a 3D volume of interest (VOI) or use a constant region of interest (ROI) and only consider the presence of nodule within the given VOI. These approaches limit the networks’ ability to detect nodules outside the given VOI and can also include unnecessary structures in the VOI, leading to potentially inaccurate segmentation. In this work, we propose a novel approach for 3D lung nodule segmentation by using 2D region of interest (ROI) inputted from radiologist or computer-aided detection (CADe) system. Particularly, we design a dual-encoder-based hard attention network (DEHA-Net) which incorporates the full slice of thoracic computed tomography scan along with the ROI mask to produce an accurate segmentation mask of lung nodule in the given slice. The proposed architecture exploits the adaptive region of interest (A-ROI) algorithm to automatically investigates the penetration of lung nodule into surrounding slices while eliminating the need to drawing separate ROIs in each slice. Further, the framework performs the multi-view analysis, i.e., in sagittal and coronal views, to improve the segmentation performance. The proposed scheme has been rigorously evaluated on the lung image database consortium and image database resource initiative (LIDC/IDRI) dataset and an extensive analysis of results have been performed. The quantitative analysis shows that the proposed method not only improves the existing state-of-the-art in term of dice score but also, significantly robust against the different types, shape and dimensions of lung nodules.

show abstract

“…The predictions of nodule attributes are considered correct if within ±1 of aggregated radiologists' annotation [15]. Feature "internal structure" is excluded from the results because its heavily imbalanced classes are not very informative [7,13,15,16,22].…”

Section: Prediction Performance Of Nodule Attributes and Malignancymentioning

confidence: 99%

“…Our experiments on the publicly available LIDC dataset [2] show that with fewer nodule samples and only 1% of their annotations, the proposed approach achieves comparable or better performance compared with state-of-the-art methods using full annotation [7,13,15,16,22], and reaches approximately 90% accuracy in predicting all nodule features simultaneously. By visualising the learned space, the extracted features are shown to be highly separable and correlated well with clinical knowledge.…”

Section: Introductionmentioning

confidence: 96%

Reducing Annotation Need in Self-Explanatory Models for Lung Nodule Diagnosis

Chen¹,

Krause²,

Erleben³

et al. 2022

Preprint

View full text Add to dashboard Cite

Feature-based self-explanatory methods explain their classification in terms of human-understandable features. In the medical imaging community, this semantic matching of clinical knowledge adds significantly to the trustworthiness of the AI. However, the cost of additional annotation of features remains a pressing issue. We address this problem by proposing cRedAnno, a data-/annotation-efficient self-explanatory approach for lung nodule diagnosis. cRedAnno considerably reduces the annotation need by introducing self-supervised contrastive learning to alleviate the burden of learning most parameters from annotation, replacing end-to-end training with two-stage training. When training with hundreds of nodule samples and only 1% of their annotations, cRedAnno achieves competitive accuracy in predicting malignancy, meanwhile significantly surpassing most previous works in predicting nodule attributes. Visualisation of the learned space further indicates that the correlation between the clustering of malignancy and nodule attributes coincides with clinical knowledge. Our complete code is open-source available: https://github.com/ludles/credanno.

show abstract

End-to-End Multi-Task Learning for Lung Nodule Segmentation and Diagnosis

Cited by 7 publications

References 20 publications

Reducing Annotation Need in Self-explanatory Models for Lung Nodule Diagnosis

Reducing Annotation Need in Self-explanatory Models for Lung Nodule Diagnosis

DEHA-Net: Dual-Encoder based Hard Attention Network with Adaptive ROI Mechanism for Lung Nodule Segmentation

Reducing Annotation Need in Self-Explanatory Models for Lung Nodule Diagnosis

Contact Info

Product

Resources

About