Analysis of segmentation of lung parenchyma based on deep learning methods

Preparing CT imaging datasets for deep learning in lung nodule analysis: Insights from four well-known datasets

BACKGROUND

Deep learning is an important means to realize the automatic detection, segmentation, and classification of pulmonary nodules in computed tomography (CT) images. An entire CT scan cannot directly be used by deep learning models due to image size, image format, image dimensionality, and other factors. Between the acquisition of the CT scan and feeding the data into the deep learning model, there are several steps including data use permission, data access and download, data annotation, and data preprocessing. This paper aims to recommend a complete and detailed guide for researchers who want to engage in interdisciplinary lung nodule research of CT images and Artificial Intelligence (AI) engineering.

METHODS

The data preparation pipeline used the following four popular large-scale datasets: LIDC-IDRI (Lung Image Database Consortium image collection), LUNA16 (Lung Nodule Analysis 2016), NLST (National Lung Screening Trial) and NELSON (The Dutch-Belgian Randomized Lung Cancer Screening Trial). The dataset preparation is presented in chronological order.

FINDINGS

The different data preparation steps before deep learning were identified. These include both more generic steps and steps dedicated to lung nodule research. For each of these steps, the required process, necessity, and example code or tools for actual implementation are provided.

DISCUSSION AND CONCLUSION

Depending on the specific research question, researchers should be aware of the various preparation steps required and carefully select datasets, data annotation methods, and image preprocessing methods. Moreover, it is vital to acknowledge that each auxiliary tool or code has its specific scope of use and limitations. This paper proposes a standardized data preparation process while clearly demonstrating the principles and sequence of different steps. A data preparation pipeline can be quickly realized by following these proposed steps and implementing the suggested example codes and tools.

The application of the nnU-Net-based automatic segmentation model in assisting carotid artery stenosis and carotid atherosclerotic plaque evaluation

Objective: No new U-net (nnU-Net) is a newly-developed deep learning neural network, whose advantages in medical image segmentation have been noticed recently. This study aimed to investigate the value of the nnU-Net-based model for computed tomography angiography (CTA) imaging in assisting the evaluation of carotid artery stenosis (CAS) and atherosclerotic plaque. Methods: This study retrospectively enrolled 93 CAS-suspected patients who underwent head and neck CTA examination, then randomly divided them into the training set (N = 70) and the validation set (N = 23) in a 3:1 ratio. The radiologist-marked images in the training set were used for the development of the nnU-Net model, which was subsequently tested in the validation set. Results: In the training set, the nnU-Net had already displayed a good performance for CAS diagnosis and atherosclerotic plaque segmentation. Then, its utility was further confirmed in the validation set: the Dice similarity coefficient value of the nnU-Net model in segmenting background, blood vessels, calcification plaques, and dark spots reached 0.975, 0.974 0.795, and 0.498, accordingly. Besides, the nnU-Net model displayed a good consistency with physicians in assessing CAS (Kappa = 0.893), stenosis degree (Kappa = 0.930), the number of calcification plaque (Kappa = 0.922), non-calcification (Kappa = 0.768) and mixed plaque (Kappa = 0.793), as well as the max thickness of calcification plaque (intraclass correlation coefficient = 0.972). Additionally, the evaluation time of the nnU-Net model was shortened compared with the physicians (27.3 ± 4.4 s vs. 296.8 ± 81.1 s, p < 0.001). Conclusion: The automatic segmentation model based on nnU-Net shows good accuracy, reliability, and efficiency in assisting CTA to evaluate CAS and carotid atherosclerotic plaques.

A multi-label classification system for anomaly classification in electrocardiogram

Automatic classification of ECG signals has become a research hotspot, and most of the research work in this field is currently aimed at single-label classification. However, a segment of ECG signal may contain more than two cardiac diseases, and single-label classification cannot accurately judge all possibilities. Besides, single-label classification performs classification in units of segmented beats, which destroys the contextual relevance of signal data. Therefore, studying the multi-label classification of ECG signals becomes more critical. This study proposes a method based on the multi-label question transformation method-binary correlation and classifies ECG signals by constructing a deep sequence model. Binary correlation simplifies the learning difficulty of deep learning models and converts multi-label problems into multiple binary classification problems. The experimental results are as follows: F1 score is 0.767, Hamming Loss is 0.073, Coverage is 3.4, and Ranking Loss is 0.262. It performs better than existing work.

LawRec: Automatic Recommendation of Legal Provisions Based on Legal Text Analysis

Smart court technologies are making full use of modern science to promote the modernization of the trial system and trial capabilities, for example, artificial intelligence, Internet of things, and cloud computing. The smart court technologies can improve the efficiency of case handling and achieving convenience for the people. Article recommendation is an important part of intelligent trial. For ordinary people without legal background, the traditional information retrieval system that searches laws and regulations based on keywords is not applicable because they do not have the ability to extract professional legal vocabulary from complex case processes. This paper proposes a law recommendation framework, called LawRec, based on Bidirectional Encoder Representation from Transformers (BERT) and Skip-Recurrent Neural Network (Skip-RNN) models. It intends to integrate the knowledge of legal provisions with the case description and uses the BERT model to learn the case description text and legal knowledge, respectively. At last, laws and regulations for cases can be recommended. Experiment results show that the proposed LawRec can achieve better performance than state-of-the-art methods.

A Lightweight CNN Model Based on GhostNet

The existing deep learning models have problems such as large weight parameters and slow inference speed of equipment. In practical applications such as fire detection, they often cannot be deployed on equipment with limited resources due to the huge amount of parameters and low efficiency. In response to this problem, this paper proposes a lightweight smoke detection model based on the convolutional attention mechanism module. The model is based on the YOLOv5 lightweight framework. The backbone network draws on the GhostNet design idea, replaces the CSP structure of the FPN and head layers with the GhostBottleNeck module, adds a convolutional attention mechanism module to the backbone network layer, and uses the CIoU loss function to improve the regression accuracy. Using YOLOv5s as the benchmark model, the parameter amount of the proposed lightweight neural network model is 2.75 M, and the floating-point calculation amount is 2.56 G, which is much lower than the parameter amount and calculation amount of the benchmark model. Tested on the public fire dataset, compared with the traditional deep learning algorithm, the model proposed in the paper has better detection performance and the detection speed is significantly better than the benchmark model. Tested under the unquantized simulator, the speed of the proposed model to detect a single picture is 60 ms, which can meet the requirements of real-time engineering applications.

Deep Learning-Based Text Emotion Analysis for Legal Anomie

Text emotion analysis is an effective way for analyzing the emotion of the subjects’ anomie behaviors. This paper proposes a text emotion analysis framework (called BCDF) based on word embedding and splicing. Bi-direction Convolutional Word Embedding Classification Framework (BCDF) can express the word vector in the text and embed the part of speech tagging information as a feature of sentence representation. In addition, an emotional parallel learning mechanism is proposed, which uses the temporal information of the parallel structure calculated by Bi-LSTM to update the storage information through the gating mechanism. The convolutional layer can better extract certain components of sentences (such as adjectives, adverbs, nouns, etc.), which play a more significant role in the expression of emotion. To take advantage of convolution, a Convolutional Long Short-Term Memory (ConvLSTM) network is designed to further improve the classification results. Experimental results show that compared with traditional LSTM model, the proposed text emotion analysis model has increased 3.3 and 10.9% F1 score on psychological and news text datasets, respectively. The proposed CBDM model based on Bi-LSTM and ConvLSTM has great value in practical applications of anomie behavior analysis.

A Contrastive Predictive Coding-Based Classification Framework for Healthcare Sensor Data

Supervised learning technologies have been used in medical-data classification to improve diagnosis efficiency and reduce human diagnosis errors. A large amount of manually annotated data are required for the fully supervised learning process. However, annotating data information will consume a large amount of manpower and resources. Self-supervised learning has great advantages in solving this problem. Self-supervised learning mainly uses pretext tasks to mine its own supervised information from large-scale unsupervised data. And this constructed supervised information is used to train the network to learn valuable representations for downstream tasks. This study designs a general and efficient model for the diagnosis and classification of medical sensor data based on contrastive predictive coding (CPC) in self-supervised learning, called TCC, which consists of two steps. The first step is to design a pretext task based on the idea of CPC, which aims to extract effective features between different categories using its encoder. The second step designs a downstream classification task with lower time and space complexity to perform a supervised type of training using the features extracted by the encoder of the pretext task. Finally, to demonstrate the performance of the proposed framework in this paper, we compare the proposed framework with recent state-of-the-art works. Experiments comparing the proposed framework with supervised learning are also set up under the condition of different proportions of labeled data.