Search

Histopathological examinations heavily rely on hematoxylin and eosin (HE) and immunohistochemistry (IHC) staining. IHC staining can offer more accurate diagnostic details but it brings significant financial and time costs. Furthermore, either re-staining HE-stained slides or using adjacent slides for IHC may compromise the accuracy of pathological diagnosis due to information loss. To address these challenges, we develop PST-Diff, a method for generating virtual IHC images from HE images based on diffusion models, which allows pathologists to simultaneously view multiple staining results from the same tissue slide. To maintain the pathological consistency of the stain transfer, we propose the asymmetric attention mechanism (AAM) and latent transfer (LT) module in PST-Diff. Specifically, the AAM can retain more local pathological information of the source domain images, while ensuring the model’s flexibility in generating virtual stained images that highly confirm to the target domain. Subsequently, the LT module transfers the implicit representations across different domains, effectively alleviating the bias introduced by direct connection and further enhancing the pathological consistency of PST-Diff. Furthermore, to maintain the structural consistency of the stain transfer, the conditional frequency guidance (CFG) module is proposed to precisely control image generation and preserve structural details according to the frequency recovery process. To conclude, the pathological and structural consistency constraints provide PST-Diff with effectiveness and superior generalization in generating stable and functionally pathological IHC images with the best evaluation score. In general, PST-Diff offers prospective application in clinical virtual staining and pathological image analysis.

Radiomics refers to the high-throughput extraction of quantitative features from medical images, and is widely used to construct machine learning models for the prediction of clinical outcomes, while feature engineering is the most important work in radiomics. However, current feature engineering methods fail to fully and effectively utilize the heterogeneity of features when dealing with different kinds of radiomics features. In this work, latent representation learning is first presented as a novel feature engineering approach to reconstruct a set of latent space features from original shape, intensity and texture features. This proposed method projects features into a subspace called latent space, in which the latent space features are obtained by minimizing a unique hybrid loss function including a clustering-like loss and a reconstruction loss. The former one ensures the separability among each class while the latter one narrows the gap between the original features and latent space features. Experiments were performed on a multi-center non-small cell lung cancer (NSCLC) subtype classification dataset from 8 international open databases. Results showed that compared with four traditional feature engineering methods (baseline, PCA, Lasso and L2,1-norm minimization), latent representation learning could significantly improve the classification performance of various machine learning classifiers on the independent test set (all p < 0.001). Further on two additional test sets, latent representation learning also showed a significant improvement in generalization performance. Our research shows that latent representation learning is a more effective feature engineering method, which has the potential to be used as a general technology in a wide range of radiomics researches.

Current cuff-less blood pressure (BP) monitoring methods have significant prediction errors, especially in the hypertensive population. Therefore, we proposed a two-stage framework based on the photoplethysmography (PPG) signal, termed SMART-BP, which can independently model in different BP intervals (e.g., hypotensive, normotensive, and hypertensive), thereby achieving high-precision noninvasive BP measurement. Specifically, SMART-BP utilizes a two-stage framework for noninvasive BP measurement based on PPG signals. The first stage involves a deep learning model (named SEM-ResNet) that identifies the BP interval, referred to as the coarse-grained classification phase (CCP). The second stage employs an automated machine learning (AutoML) pipeline to estimate the BP in the corresponding interval, referred to as the fine-grained regression phase (FRP). To improve the BP prediction accuracy, we designed a PPG morphological feature learning (PMFL) algorithm to obtain a highly correlated feature subset. These discriminative features can be used as prior knowledge for coarse-grained classification networks and to provide stable results for fine-grained regression pipelines. We further demonstrated the robustness of the proposed approach using two independent multicenter datasets by employing transfer learning techniques. We first fine-tuned SMART-BP with a large high-quality dataset (Mindray dataset) and then transferred the pretrained model to the target domain (MIMIC dataset). Experimental tests showed that the SMART-BP had an estimation error of −0.01 ± 1.85 and 0.01 ± 3.50 mmHg for diastolic and systolic BP (SBP), respectively, which met the advancement of the medical instrumentation standard. These results demonstrated the high reliability and robustness of the SMART-BP in measuring BP values.