Exploring Generalizable Distillation for Efficient Medical Image Segmentation.

Efficient medical image segmentation aims to provide accurate pixel-wise predictions with a lightweight implementation framework. However, existing lightweight networks generally overlook the generalizability of the cross-domain medical segmentation tasks. In this paper, we propose Generalizable Knowledge Distillation (GKD), a novel framework for enhancing the performance of lightweight networks on cross-domain medical segmentation by generalizable knowledge distillation from powerful teacher networks. Considering the domain gaps between different medical datasets, we propose the Model-Specific Alignment Networks (MSAN) to obtain the domain-invariant representations. Meanwhile, a customized Alignment Consistency Training (ACT) strategy is designed to promote the MSAN training. Based on the domain-invariant vectors in MSAN, we propose two generalizable distillation schemes, Dual Contrastive Graph Distillation (DCGD) and Domain-Invariant Cross Distillation (DICD). In DCGD, two implicit contrastive graphs are designed to model the intra-coupling and inter-coupling semantic correlations. Then, in DICD, the domain-invariant semantic vectors are reconstructed from two networks (i.e., teacher and student) with a crossover manner to achieve simultaneous generalization of lightweight networks, hierarchically. Moreover, a metric named Fréchet Semantic Distance (FSD) is tailored to verify the effectiveness of the regularized domain-invariant features. Extensive experiments conducted on the Liver, Retinal Vessel and Colonoscopy segmentation datasets demonstrate the superiority of our method, in terms of performance and generalization ability on lightweight networks.

AnyFace++: A Unified Framework for Free-style Text-to-Face Synthesis and Manipulation.

Human faces contain rich semantic information that could hardly be described without a large vocabulary and complex sentence patterns. However, most existing text-to-image synthesis methods could only generate meaningful results based on limited sentence templates with words contained in the training set, which heavily impairs the generalization ability of these models. In this paper, we define a novel 'free-style' text-to-face generation and manipulation problem, and propose an effective solution, named AnyFace++, which is applicable to a much wider range of open-world scenarios. The CLIP model is involved in AnyFace++ for learning an aligned language-vision feature space, which also expands the range of acceptable vocabulary as it is trained on a large-scale dataset. To further improve the granularity of semantic alignment between text and images, a memory module is incorporated to convert the description with arbitrary length, format, and modality into regularized latent embeddings representing discriminative attributes of the target face. Moreover, the diversity and semantic consistency of generation results are improved by a novel semi-supervised training scheme and a series of newly proposed objective functions. Compared to state-of-the-art methods, AnyFace++ is capable of synthesizing and manipulating face images based on more flexible descriptions and producing realistic images with higher diversity.

Biphasic Face Photo-Sketch Synthesis via Semantic-Driven Generative Adversarial Network With Graph Representation Learning.

Biphasic face photo-sketch synthesis has significant practical value in wide-ranging fields such as digital entertainment and law enforcement. Previous approaches directly generate the photo-sketch in a global view, they always suffer from the low quality of sketches and complex photograph variations, leading to unnatural and low-fidelity results. In this article, we propose a novel semantic-driven generative adversarial network to address the above issues, cooperating with graph representation learning. Considering that human faces have distinct spatial structures, we first inject class-wise semantic layouts into the generator to provide style-based spatial information for synthesized face photographs and sketches. In addition, to enhance the authenticity of details in generated faces, we construct two types of representational graphs via semantic parsing maps upon input faces, dubbed the intraclass semantic graph (IASG) and the interclass structure graph (IRSG). Specifically, the IASG effectively models the intraclass semantic correlations of each facial semantic component, thus producing realistic facial details. To preserve the generated faces being more structure-coordinated, the IRSG models interclass structural relations among every facial component by graph representation learning. To further enhance the perceptual quality of synthesized images, we present a biphasic interactive cycle training strategy by fully taking advantage of the multilevel feature consistency between the photograph and sketch. Extensive experiments demonstrate that our method outperforms the state-of-the-art competitors on the CUHK Face Sketch (CUFS) and CUHK Face Sketch FERET (CUFSF) datasets.

An Automated Framework for Histopathological Nucleus Segmentation with Deep Attention Integrated Networks.

Clinical management and accurate disease diagnosis are evolving from qualitative stage to the quantitative stage, particularly at the cellular level. However, the manual process of histopathological analysis is lab-intensive and time-consuming. Meanwhile, the accuracy is limited by the experience of the pathologist. Therefore, deep learning-empowered computer-aided diagnosis (CAD) is emerging as an important topic in digital pathology to streamline the standard process of automatic tissue analysis. Automated accurate nucleus segmentation can not only help pathologists make more accurate diagnosis, save time and labor, but also achieve consistent and efficient diagnosis results. However, nucleus segmentation is susceptible to staining variation, uneven nucleus intensity, background noises, and nucleus tissue differences in biopsy specimens. To solve these problems, we propose Deep Attention Integrated Networks (DAINets), which mainly built on self-attention based spatial attention module and channel attention module. In addition, we also introduce a feature fusion branch to fuse high-level representations with low-level features for multi-scale perception, and employ the mark-based watershed algorithm to refine the predicted segmentation maps. Furthermore, in the testing phase, we design Individual Color Normalization (ICN) to settle the dyeing variation problem in specimens. Quantitative evaluations on the multi-organ nucleus dataset indicate the priority of our automated nucleus segmentation framework.

Graph Flow: Cross-Layer Graph Flow Distillation for Dual Efficient Medical Image Segmentation.

With the development of deep convolutional neural networks, medical image segmentation has achieved a series of breakthroughs in recent years. However, high-performance convolutional neural networks always mean numerous parameters and high computation costs, which will hinder the applications in resource-limited medical scenarios. Meanwhile, the scarceness of large-scale annotated medical image datasets further impedes the application of high-performance networks. To tackle these problems, we propose Graph Flow, a comprehensive knowledge distillation framework, for both network-efficiency and annotation-efficiency medical image segmentation. Specifically, the Graph Flow Distillation transfers the essence of cross-layer variations from a well-trained cumbersome teacher network to a non-trained compact student network. In addition, an unsupervised Paraphraser Module is integrated to purify the knowledge of the teacher, which is also beneficial for the training stabilization. Furthermore, we build a unified distillation framework by integrating the adversarial distillation and the vanilla logits distillation, which can further refine the final predictions of the compact network. With different teacher networks (traditional convolutional architecture or prevalent transformer architecture) and student networks, we conduct extensive experiments on four medical image datasets with different modalities (Gastric Cancer, Synapse, BUSI, and CVC-ClinicDB). We demonstrate the prominent ability of our method on these datasets, which achieves competitive performances. Moreover, we demonstrate the effectiveness of our Graph Flow through a novel semi-supervised paradigm for dual efficient medical image segmentation. Our code will be available at Graph Flow.

Accurate Retinal Vessel Segmentation in Color Fundus Images via Fully Attention-Based Networks.

Automatic retinal vessel segmentation is important for the diagnosis and prevention of ophthalmic diseases. The existing deep learning retinal vessel segmentation models always treat each pixel equally. However, the multi-scale vessel structure is a vital factor affecting the segmentation results, especially in thin vessels. To address this crucial gap, we propose a novel Fully Attention-based Network (FANet) based on attention mechanisms to adaptively learn rich feature representation and aggregate the multi-scale information. Specifically, the framework consists of the image pre-processing procedure and the semantic segmentation networks. Green channel extraction (GE) and contrast limited adaptive histogram equalization (CLAHE) are employed as pre-processing to enhance the texture and contrast of retinal blood images. Besides, the network combines two types of attention modules with the U-Net. We propose a lightweight dual-direction attention block to model global dependencies and reduce intra-class inconsistencies, in which the weights of feature maps are updated based on the semantic correlation between pixels. The dual-direction attention block utilizes horizontal and vertical pooling operations to produce the attention map. In this way, the network aggregates global contextual information from semantic-closer regions or a series of pixels belonging to the same object category. Meanwhile, we adopt the selective kernel (SK) unit to replace the standard convolution for obtaining multi-scale features of different receptive field sizes generated by soft attention. Furthermore, we demonstrate that the proposed model can effectively identify irregular, noisy, and multi-scale retinal vessels. The abundant experiments on DRIVE, STARE, and CHASE_DB1 datasets show that our method achieves state-of-the-art performance.

Prediction of BAP1 Expression in Uveal Melanoma Using Densely-Connected Deep Classification Networks.

Uveal melanoma is the most common primary intraocular malignancy in adults, with nearly half of all patients eventually developing metastases, which are invariably fatal. Manual assessment of the level of expression of the tumor suppressor BRCA1-associated protein 1 (BAP1) in tumor cell nuclei can identify patients with a high risk of developing metastases, but may suffer from poor reproducibility. In this study, we verified whether artificial intelligence could predict manual assessments of BAP1 expression in 47 enucleated eyes with uveal melanoma, collected from one European and one American referral center. Digitally scanned pathology slides were divided into 8176 patches, each with a size of 256 × 256 pixels. These were in turn divided into a training cohort of 6800 patches and a validation cohort of 1376 patches. A densely-connected classification network based on deep learning was then applied to each patch. This achieved a sensitivity of 97.1%, a specificity of 98.1%, an overall diagnostic accuracy of 97.1%, and an F1-score of 97.8% for the prediction of BAP1 expression in individual high resolution patches, and slightly less with lower resolution. The area under the receiver operating characteristic (ROC) curves of the deep learning model achieved an average of 0.99. On a full tumor level, our network classified all 47 tumors identically with an ophthalmic pathologist. We conclude that this deep learning model provides an accurate and reproducible method for the prediction of BAP1 expression in uveal melanoma.

A support vector machine approach for AF classification from a short single-lead ECG recording.

OBJECTIVE: In this paper, a support vector machine (SVM) approach using statistical features, P wave absence, spectrum features, and length-adaptive entropy are presented to classify ECG rhythms as four types: normal rhythm, atrial fibrillation (AF), other rhythm, and too noisy to classify. APPROACH: The proposed algorithm consisted of three steps: (1) signal pre-processing based on the wavelet method; (2) feature extraction, the extracted features including one power feature, two spectrum features, two entropy features, 17 RR interval-related features, and 11 P wave features; and (3) classification using the SVM classifier. MAIN RESULTS: The algorithm was trained by 8528 single-lead ECG recordings lasting from 9 s to just over 60 s and then tested on a hidden test set consisting of 3658 recordings of similar lengths, which were all provided by the PhysioNet/Computing in Cardiology Challenge 2017. The scoring for this challenge used an F 1 measure, and the final F 1 score was defined as the average of F 1n (the F 1 score of normal rhythm), F 1a (the F 1 score of AF rhythm), and F 1o (the F 1 score of other rhythm). The results confirmed the high accuracy of our proposed method, which obtained 90.27%, 86.37%, and 75.08% for F 1n , F 1a , and F 1n and the final F 1 score of 84% on the training set. In the final test to assess the performance of all of the hidden data, the obtained F 1n , F 1a , F 1o and the average F 1 were 90.82%, 78.56%, 71.77% and 80%, respectively. SIGNIFICANCE: The proposed algorithm targets a large number of raw, short single ECG data rather than a small number of carefully selected, often clean ECG records, which have been studied in most of the previous literature. It breaks through the limitation in applicability and provides reliable AF detection from a short single-lead ECG.

Chinese Biotheology and Ethical spirit of Christianity / 中国医学伦理学

This paper disscuss religious base of bioethics and chinese biotheology.