In electroencephalography (EEG)-based emotion recognition systems, the distribution between the training samples and the testing samples may be mismatched if they are sampled from different experimental sessions or subjects because of user fatigue, different electrode placements, varying impedances, etc. Therefore, it is difficult to directly classify the EEG patterns with a conventional classifier. The domain adaptation method, which is aimed at obtaining a common representation across training and test domains, is an effective method for reducing the distribution discrepancy. However, the existing domain adaptation strategies either employ a linear transformation or learn the nonlinearity mapping without a consistency constraint; they are not sufficiently powerful to obtain a similar distribution from highly non-stationary EEG signals. To address this problem, in this paper, a novel component, called the subspace alignment auto-encoder (SAAE), is proposed. Taking advantage of both nonlinear transformation and a consistency constraint, we combine an auto-encoder network and a subspace alignment solution in a unified framework. As a result, the source domain can be aligned with the target domain together with its class label, and any supervised method can be applied to the new source domain to train a classifier for classification in the target domain, as the aligned source domain follows a distribution similar to that of the target domain. We compared our SAAE method with six typical approaches using a public EEG dataset containing three affective states: positive, neutral, and negative. Subject-to-subject and session-to-session evaluations were performed. The subject-to-subject experimental results demonstrate that our component achieves a mean accuracy of 77.88% in comparison with a state-of-the-art method, TCA, which achieves 73.82% on average. In addition, the average classification accuracy of SAAE in the session-to-session evaluation for all the 15 subjects in a dataset is 81.81%, an improvement of up to 1.62% on average as compared to the best baseline TCA. The experimental results show the effectiveness of the proposed method relative to state-of-the-art methods. It can be concluded that SAAE is a useful and effective tool for decreasing domain discrepancy and reducing performance degradation across subjects and sessions in the EEG-based emotion recognition field.

Electroencephalography (EEG)-based emotion recognition is an important element in psychiatric health diagnosis for patients. However, the underlying EEG sensor signals are always non-stationary if they are sampled from different experimental sessions or subjects. This results in the deterioration of the classification performance. Domain adaptation methods offer an effective way to reduce the discrepancy of marginal distribution. However, for EEG sensor signals, both marginal and conditional distributions may be mismatched. In addition, the existing domain adaptation strategies always require a high level of additional computation. To address this problem, a novel strategy named adaptive subspace feature matching (ASFM) is proposed in this paper in order to integrate both the marginal and conditional distributions within a unified framework (without any labeled samples from target subjects). Specifically, we develop a linear transformation function which matches the marginal distributions of the source and target subspaces without a regularization term. This significantly decreases the time complexity of our domain adaptation procedure. As a result, both marginal and conditional distribution discrepancies between the source domain and unlabeled target domain can be reduced, and logistic regression (LR) can be applied to the new source domain in order to train a classifier for use in the target domain, since the aligned source domain follows a distribution which is similar to that of the target domain. We compare our ASFM method with six typical approaches using a public EEG dataset with three affective states: positive, neutral, and negative. Both offline and online evaluations were performed. The subject-to-subject offline experimental results demonstrate that our component achieves a mean accuracy and standard deviation of 80.46% and 6.84%, respectively, as compared with a state-of-the-art method, the subspace alignment auto-encoder (SAAE), which achieves values of 77.88% and 7.33% on average, respectively. For the online analysis, the average classification accuracy and standard deviation of ASFM in the subject-to-subject evaluation for all the 15 subjects in a dataset was 75.11% and 7.65%, respectively, gaining a significant performance improvement compared to the best baseline LR which achieves 56.38% and 7.48%, respectively. The experimental results confirm the effectiveness of the proposed method relative to state-of-the-art methods. Moreover, computational efficiency of the proposed ASFM method is much better than standard domain adaptation; if the numbers of training samples and test samples are controlled within certain range, it is suitable for real-time classification. It can be concluded that ASFM is a useful and effective tool for decreasing domain discrepancy and reducing performance degradation across subjects and sessions in the field of EEG-based emotion recognition.

No abstract

No abstract

In multiple plan adaptive radiotherapy (ART) strategies of bladder cancer, a library of plans corresponding to different bladder volumes is created based on images acquired in early treatment sessions. Subsequently, the plan for the smallest PTV safely covering the bladder on cone-beam CT (CBCT) is selected as the plan of the day. The aim of this study is to develop an automatic bladder segmentation approach suitable for CBCT scans and test its ability to select the appropriate plan from the library of plans for such an ART procedure. Twenty-three bladder cancer patients with a planning CT and on average 11.6 CBCT scans were included in our study. For each patient, all CBCT scans were matched to the planning CT on bony anatomy. Bladder contours were manually delineated for each planning CT (for model building) and CBCT (for model building and validation). The automatic segmentation method consisted of two steps. A patient-specific bladder deformation model was built from the training data set of each patient (the planning CT and the first five CBCT scans). Then, the model was applied to automatically segment bladders in the validation data of the same patient (the remaining CBCT scans). Principal component analysis (PCA) was applied to the training data to model patient-specific bladder deformation patterns. The number of PCA modes for each patient was chosen such that the bladder shapes in the training set could be represented by such number of PCA modes with less than 0.1 cm mean residual error. The automatic segmentation started from the bladder shape of a reference CBCT, which was adjusted by changing the weight of each PCA mode. As a result, the segmentation contour was deformed consistently with the training set to fit the bladder in the validation image. A cost function was defined by the absolute difference between the directional gradient field of reference CBCT sampled on the corresponding bladder contour and the directional gradient field of validation CBCT sampled on the segmentation contour candidate. The cost function measured the goodness of fit of the segmentation on the validation image and was minimized using a simplex optimizer. For each validation CBCT image, the segmentations were done five times using a different reference CBCT. The one with the lowest cost function was selected as the final bladder segmentation. Volume- and distance-based metrics and the accuracy of plan selection were evaluated to quantify the performance. Two to four PCA modes were needed to represent the bladder shape variation with less than 0.1 cm average residual error for the training data of each patient. The automatically segmented bladders had a 78.5% mean conformity index with the manual delineations. The mean SD of the local residual error over all patients was 0.24 cm. The agreement of plan selection between automatic and manual bladder segmentations was 77.5%. PCA is an efficient method to describe patient-specific bladder deformation. The statistical-shape-based segmentation approach is robust to handle the r...

Purpose To describe the clinical characteristics and outcomes of patients with dual-phenotype hepatocellular carcinoma (DPHCC) and investigate the use of radiomics to establish an image-based signature for preoperative differential diagnosis. Methods This study included 50 patients with a postoperative pathological diagnosis of DPHCC (observation group) and 50 patients with CK7-and CK19-negative HCC (control group) who attended our hospital between January 2015 and December 2018. All patients underwent Gd-EOB-DTPA-enhanced MRI within 1 month before surgery. Arterial phase (AP), portal venous phase (PVP), delayed phase (DP) and hepatobiliary phase (HBP) images were transferred into a radiomics platform. Volumes of interest covered the whole tumor. The dimensionality of the radiomics features were reduced using LASSO. Four classifiers, including multi-layer perceptron (MLP), support vector machines (SVM), logistic regression (LR) and K-nearest neighbor (KNN) were used to distinguish DPHCC from CK7-and CK19-negative HCC. Kaplan-Meier survival analysis was used to assess 1-year disease-free survival (DFS) and overall survival (OS) in the observation and control groups. Results The best preoperative diagnostic power for DPHCC will likely be derived from a combination of different phases and classifiers. The sensitivity, specificity and accuracy of LR in PVP (0.740, 0.780, 0.766), DP (0.893, 0.700, 0.798), HBP (0.800, 0.720, 0.756) and MLP in PVP (0.880, 0.720, 0.798) were better performance. The 1-year DFS and OS of the patients in the observation group were 69% and 78%, respectively. The 1-year DFS and OS of the patients in the control group were 83% and 85%, respectively. Kaplan-Meier survival analysis showed no statistical difference in DFS and OS between groups (P = 0.231 and 0.326), but DFS and OS were numerically lower in patients with DPHCC. Conclusion The radiomics features extracted from Gd-EOB-DTPA-enhanced MR images can be used to diagnose preoperative DPHCC. DPHCC is more likely to recur and cause death than HCC, suggesting that active postoperative management of patients with DPHCC is required.

No abstract

No abstract