Breast Ultrasound (BUS) imaging has been recognized as an essential imaging modality for breast masses classification in China. Current deep learning (DL) based solutions for BUS classification seek to feed ultrasound (US) images into deep convolutional neural networks (CNNs), to learn a hierarchical combination of features for discriminating malignant and benign masses. One existing problem in current DL-based BUS classification was the lack of spatial and channel-wise features weighting, which inevitably allow interference from redundant features and low sensitivity. In this study, we aim to incorporate the instructive information provided by breast imaging reporting and data system (BI-RADS) within DL-based classification. A novel DL-based BI-RADS Vector-Attention Network (BVA Net) that trains with both texture information and decoded information from BI-RADS stratifications was proposed for the task. Three baseline models, pretrained DenseNet-121, ResNet-50 and Residual-Attention Network(RA Net) were included for comparison. Experiments were conducted on a large scale private main dataset and two public datasets, UDIAT and BUSI. On the main dataset, BVA Net outperformed other models, in terms of AUC (area under the receiver operating curve, 0.908), ACC (accuracy, 0.865), sensitivity (0.812) and precision(0.795). BVA Net also achieved the high AUC (0.87 and 0.882) and ACC (0.859 and 0.843), on UDIAT and BUSI. Moreover, we proposed a method that integrates both BVA Net binary classification and BI-RADS stratification estimation, called integrated classification. The introduction of integrated classification helped improving the overall sensitivity while maintaining a high specificity.

In this paper, we investigate an automated pipeline to estimate respiration signals from videos for premature infants in neonatal intensive care units (NICUs). Two flow estimation methods, namely the conventional optical flow- and deep learning-based flow estimation methods, were employed and compared to estimate pixel motion vectors between adjacent video frames. The respiratory signal is further extracted via motion factorization. The proposed methods were evaluated by comparing our automated extracted respiration signals to that extracted from chest impedance on videos of five premature infants. The overall average cross-correlation coefficients are 0.70 for the optical flow-based method and 0.74 for the deep flow-based method. The average root mean-squared errors are 6.10 and 4.55 for the optical flow- and the deep flow-based methods, respectively. The experimental results are promising for further investigation and clinical application of the video-based respiration monitoring method for infants in NICUs.

Image-based artificial intelligence (AI) systems stand as the major modality for evaluating ophthalmic conditions. However, most of the currently available AI systems are designed for experimental research using single-central datasets. Most of them fell short of application in real-world clinical settings. In this study, we collected a dataset of 1,099 fundus images in both normal and pathologic eyes from 483 premature infants for intelligent retinopathy of prematurity (ROP) system development and validation. Dataset diversity was visualized with a spatial scatter plot. Image classification was conducted by three annotators. To the best of our knowledge, this is one of the largest fundus datasets on ROP, and we believe it is conducive to the real-world application of AI systems.

Purpose: Retinopathy of prematurity (ROP) is a retinal vascular proliferative disease common in low birth weight and premature infants and is one of the main causes of blindness in children. In the context of predictive, preventive and personalized medicine (PPPM/3PM), early screening, identification and treatment of ROP will directly contribute to improve patients’ long-term visual prognosis and reduce the risk of blindness. Thus, our objective is to establish an artificial intelligence (AI) algorithm combined with clinical demographics to create a risk model for ROP including treatment-requiring retinopathy of prematurity (TR-ROP) infants. Methods: A total of 22,569 infants who underwent routine ROP screening in Shenzhen Eye Hospital from March 2003 to September 2023 were collected, including 3335 infants with ROP and 1234 infants with TR-ROP among ROP infants. Two machine learning methods of logistic regression and decision tree and a deep learning method of multi-layer perceptron were trained by using the relevant combination of risk factors such as birth weight (BW), gestational age (GA), gender, whether multiple births (MB) and mode of delivery (MD) to achieve the risk prediction of ROP and TR-ROP. We used five evaluation metrics to evaluate the performance of the risk prediction model. The area under the receiver operating characteristic curve (AUC) and the area under the precision-recall curve (AUCPR) were the main measurement metrics. Results: In the risk prediction for ROP, the BW + GA demonstrated the optimal performance (mean ± SD, AUCPR: 0.4849 ± 0.0175, AUC: 0.8124 ± 0.0033). In the risk prediction of TR-ROP, reasonable performance can be achieved by using GA + BW + Gender + MD + MB (AUCPR: 0.2713 ± 0.0214, AUC: 0.8328 ± 0.0088). Conclusions: Combining risk factors with AI in screening programs for ROP could achieve risk prediction of ROP and TR-ROP, detect TR-ROP earlier and reduce the number of ROP examinations and unnecessary physiological stress in low-risk infants. Therefore, combining ROP-related biometric information with AI is a cost-effective strategy for predictive diagnostic, targeted prevention, and personalization of medical services in early screening and treatment of ROP.

Three-dimensional organ and cancer segmentation based on multi-sequence MRI is crucial for assisting clinical diagnosis. However, current automated segmentation methods often focus on specific sequences, specific organs, and specific cancers, i.e., lack of generality. To address this issue, we propose a universal segmentation network for multi-sequence MRI (UniMRISegNet) that can segment multiple organs and cancers. UniMRISegNet features a shared encoder-decoder architecture equipped with contextual prompt generation (CPG) and prompt-conditioned dynamic convolution (PCDC) modules. The CPG module encodes sequence-specific, position-specific, and organ/cancer-specific text prompts as prior information to inform UniMRISegNet about the specific task to be executed. The PCDC module can adaptively generate model weights based on the assigned prompts, enhancing the segmentation capabilities of the UniMRISegNet for specific tasks. To mitigate discrepancies between different sequences of the same organ and capture similarities between related sequences, we design a novel loss function called Semantic-Aware Cosine Similarity Loss (SACSL), which integrates the cosine similarity of text embeddings to reconcile discrepancies and similarities between MRI sequences of the same organ. We created a large-scale annotated multi-sequence, multi-organ, and multi-cancer segmentation workflow (MSOCS), and demonstrated that our UniMRISegNet outperforms other universal networks and single-task networks on MSOCS. Furthermore, the universal weights from MSOCS can be transferred to never-before-seen downstream tasks, achieving superior performance compared to training from scratch.