Robust Vascular Segmentation for Raw Complex Images of Laser Speckle Contrast Based on Weakly Supervised Learning.

Laser speckle contrast imaging (LSCI) is widely used for in vivo real-time detection and analysis of local blood flow microcirculation due to its non-invasive ability and excellent spatial and temporal resolution. However, vascular segmentation of LSCI images still faces a lot of difficulties due to numerous specific noises caused by the complexity of blood microcirculation's structure and irregular vascular aberrations in diseased regions. In addition, the difficulties of LSCI image data annotation have hindered the application of deep learning methods based on supervised learning in the field of LSCI image vascular segmentation. To tackle these difficulties, we propose a robust weakly supervised learning method, which selects the threshold combinations and processing flows instead of labor-intensive annotation work to construct the ground truth of the dataset, and design a deep neural network, FURNet, based on UNet++ and ResNeXt. The model obtained from training achieves high-quality vascular segmentation and captures multi-scene vascular features on both constructed and unknown datasets with good generalization. Furthermore, we intravital verified the availability of this method on a tumor before and after embolization treatment. This work provides a new approach for realizing LSCI vascular segmentation and also makes a new application-level advance in the field of artificial intelligence-assisted disease diagnosis.

Unpaired fundus image enhancement based on constrained generative adversarial networks.

Fundus photography (FP) is a crucial technique for diagnosing the progression of ocular and systemic diseases in clinical studies, with wide applications in early clinical screening and diagnosis. However, due to the nonuniform illumination and imbalanced intensity caused by various reasons, the quality of fundus images is often severely weakened, brings challenges for automated screening, analysis, and diagnosis of diseases. To resolve this problem, we developed strongly constrained generative adversarial networks (SCGAN). The results demonstrate that the quality of various datasets were more significantly enhanced based on SCGAN, simultaneously more effectively retaining tissue and vascular information under various experimental conditions. Furthermore, the clinical effectiveness and robustness of this model were validated by showing its improved ability in vascular segmentation as well as disease diagnosis. Our study provides a new comprehensive approach for FP and also possesses the potential capacity to advance artificial intelligence-assisted ophthalmic examination.

A novel hybrid machine learning model for auxiliary diagnosing myocardial ischemia.

Introduction: Accurate identification of the myocardial texture features of fat around the coronary artery on coronary computed tomography angiography (CCTA) images are crucial to improve clinical diagnostic efficiency of myocardial ischemia (MI). However, current coronary CT examination is difficult to recognize and segment the MI characteristics accurately during earlier period of inflammation. Materials and methods: We proposed a random forest model to automatically segment myocardium and extract peripheral fat features. This hybrid machine learning (HML) model is integrated by CCTA images and clinical data. A total of 1,316 radiomics features were extracted from CCTA images. To further obtain the features that contribute the most to the diagnostic model, dimensionality reduction was applied to filter features to three: LNS, GFE, and WLGM. Moreover, statistical hypothesis tests were applied to improve the ability of discriminating and screening clinical features between the ischemic and non-ischemic groups. Results: By comparing the accuracy, recall, specificity and AUC of the three models, it can be found that HML had the best performance, with the value of 0.848, 0.762, 0.704 and 0.729. Conclusion: In sum, this study demonstrates that ML-based radiomics model showed good predictive value in MI, and offer an enhanced tool for predicting prognosis with greater accuracy.

Robust vessel segmentation in laser speckle contrast images based on semi-weakly supervised learning.

Objective.The goal of this study is to develop a robust semi-weakly supervised learning strategy for vessel segmentation in laser speckle contrast imaging (LSCI), addressing the challenges associated with the low signal-to-noise ratio, small vessel size, and irregular vascular aberration in diseased regions, while improving the performance and robustness of the segmentation method.Approach.For the training dataset, the healthy vascular images denoted as normal-vessel samples were manually labeled, while the diseased LSCI images involving tumor or embolism were denoted as abnormal-vessel samples and annotated as pseudo labels by the traditional semantic segmentation methods. In the training phase, the pseudo labels were constantly updated to improve the segmentation accuracy based on DeepLabv3+. Objective evaluation was conducted on the normal-vessel test set, while subjective evaluation was performed on the abnormal-vessel test set.Main results.The proposed method achieved an IOU of 0.8671, a Dice of 0.9288, and a mean relative percentage difference (mRPD) with supervised learning of 0.5% in the objective evaluation. In the subjective evaluation, our method significantly outperformed other methods in main vessel segmentation, tiny vessel segmentation, and blood vessel connection. Additionally, our method exhibited robustness when abnormal-vessel style noise was added to normal-vessel samples using a style translation network.Significance.The proposed semi-weakly supervised learning strategy demonstrates high efficiency and excellent robustness for vascular segmentation in LSCI, providing a potential tool for assessing the morphological and structural features of vessels in clinical applications.