Pathological Priors Inspired Network for Vertebral Osteophytes Recognition.

Automatic vertebral osteophyte recognition in Digital Radiography is of great importance for the early prediction of degenerative disease but is still a challenge because of the tiny size and high inter-class similarity between normal and osteophyte vertebrae. Meanwhile, common sampling strategies applied in Convolution Neural Network could cause detailed context loss. All of these could lead to an incorrect positioning predicament. In this paper, based on important pathological priors, we define a set of potential lesions of each vertebra and propose a novel Pathological Priors Inspired Network (PPIN) to achieve accurate osteophyte recognition. PPIN comprises a backbone feature extractor integrating with a Wavelet Transform Sampling module for high-frequency detailed context extraction, a detection branch for locating all potential lesions and a classification branch for producing final osteophyte recognition. The Anatomical Map-guided Filter between two branches helps the network focus on the specific anatomical regions via the generated heatmaps of potential lesions in the detection branch to address the incorrect positioning problem. To reduce the inter-class similarity, a Bilateral Augmentation Module based on the graph relationship is proposed to imitate the clinical diagnosis process and to extract discriminative contextual information between adjacent vertebrae in the classification branch. Experiments on the two osteophytes-specific datasets collected from the public VinDr-Spine database show that the proposed PPIN achieves the best recognition performance among multitask frameworks and shows strong generalization. The results on a private dataset demonstrate the potential in clinical application. The Class Activation Maps also show the powerful localization capability of PPIN. The source codes are available in https://github.com/Phalo/PPIN.

BreasTDLUSeg: A coarse-to-fine framework for segmentation of breast terminal duct lobular units on histopathological whole-slide images.

Automatic segmentation of breast terminal duct lobular units (TDLUs) on histopathological whole-slide images (WSIs) is crucial for the quantitative evaluation of TDLUs in the diagnostic and prognostic analysis of breast cancer. However, TDLU segmentation remains a great challenge due to its highly heterogeneous sizes, structures, and morphologies as well as the small areas on WSIs. In this study, we propose BreasTDLUSeg, an efficient coarse-to-fine two-stage framework based on multi-scale attention to achieve localization and precise segmentation of TDLUs on hematoxylin and eosin (H&E)-stained WSIs. BreasTDLUSeg consists of two networks: a superpatch-based patch-level classification network (SPPC-Net) and a patch-based pixel-level segmentation network (PPS-Net). SPPC-Net takes a superpatch as input and adopts a sub-region classification head to classify each patch within the superpatch as TDLU positive or negative. PPS-Net takes the TDLU positive patches derived from SPPC-Net as input. PPS-Net deploys a multi-scale CNN-Transformer as an encoder to learn enhanced multi-scale morphological representations and an upsampler to generate pixel-wise segmentation masks for the TDLU positive patches. We also constructed two breast cancer TDLU datasets containing a total of 530 superpatch images with patch-level annotations and 2322 patch images with pixel-level annotations to enable the development of TDLU segmentation methods. Experiments on the two datasets demonstrate that BreasTDLUSeg outperforms other state-of-the-art methods with the highest Dice similarity coefficients of 79.97% and 92.93%, respectively. The proposed method shows great potential to assist pathologists in the pathological analysis of breast cancer. An open-source implementation of our approach can be found at https://github.com/Dian-kai/BreasTDLUSeg.

Clinical and economic evaluation of blood culture whole process optimisation in critically ill adult patients with positive blood cultures.

OBJECTIVES: Optimising blood culture processing is important to ensure that bloodstream infections are accurately diagnosed while minimising adverse events caused by antibiotic abuse. This study aimed to evaluate the impact of optimised blood culture processes on antibiotic use, clinical outcomes and economics in intensive care unit (ICU) patients with positive blood cultures. METHODS: From March 2020 to October 2021, this microbiology laboratory implemented a series of improvement measures, including the clinical utility of Fastidious Antimicrobial Neutralization (FAN® PLUS) bottles for the BacT/Alert Virtuo blood culture system, optimisation of bottle reception, graded reports and an upgraded laboratory information system. A total of 122 ICU patients were included in the pre-optimisation group from March 2019 to February 2020, while 179 ICU patients were included in the post-optimisation group from November 2021 to October 2022. RESULTS: Compared with the pre-optimisation group, the average reporting time of identification and antimicrobial sensitivity was reduced by 16.72 hours in the optimised group. The time from admission to targeted antibiotic therapy within 24 hours after receiving both the Gram stain report and the final report were both significantly less in the post-optimisation group compared with the pre-optimisation group. The average hospitalisation time was reduced by 6.49 days, the average antimicrobial drug cost lowered by $1720.85 and the average hospitalisation cost by $9514.17 in the post-optimisation group. CONCLUSIONS: Optimising blood culture processing was associated with a significantly increased positive detection rate, a remarkable reduction in the length of hospital stay and in hospital costs for ICU patients with bloodstream infections.

Massive field-of-view sub-cellular traction force videography enabled by Single-Pixel Optical Tracers (SPOT).

We report a massive field-of-view and high-speed videography platform for measuring the sub-cellular traction forces of more than 10,000 biological cells over 13 mm2 at 83 frames per second. Our Single-Pixel Optical Tracers (SPOT) tool uses 2-dimensional diffraction gratings embedded into a soft substrate to convert cells' mechanical traction force into optical colors detectable by a video camera. The platform measures the sub-cellular traction forces of diverse cell types, including tightly connected tissue sheets and near isolated cells. We used this platform to explore the mechanical wave propagation in a tightly connected sheet of Neonatal Rat Ventricular Myocytes (NRVMs) and discovered that the activation time of some tissue regions are heterogeneous from the overall spiral wave behavior of the cardiac wave.

Case report: A toxoplasmic encephalitis in an immunocompromised child detected through metagenomic next-generation sequencing.

There exist numerous pathogens that are capable of causing infections within the central nervous system (CNS); however, conventional detection and analysis methods prove to be challenging. Clinical diagnosis of CNS infections often depends on clinical characteristics, cerebrospinal fluid (CSF) analysis, imaging, and molecular detection assays. Unfortunately, these methods can be both insensitive and time consuming, which can lead to missed diagnoses and catastrophic outcomes, especially in the case of infrequent diseases. Despite the application of appropriate prophylactic regimens and evidence-based antimicrobial agents, CNS infections continue to result in significant morbidity and mortality in hospital settings. Metagenomic next-generation sequencing (mNGS) is a novel tool that enables the identification of thousands of pathogens in a target-independent manner in a single run. The role of this innovative detection method in clinical pathogen diagnostics has matured over time. In this particular research, clinicians employed mNGS to investigate a suspected CNS infection in a child with leukemia, and unexpectedly detected Toxoplasma gondii. Case: A 3-year-old child diagnosed with T-cell lymphoblastic lymphoma was admitted to our hospital due to a 2-day history of fever and headache, along with 1 day of altered consciousness. Upon admission, the patient's Glasgow Coma Scale score was 14. Brain magnetic resonance imaging revealed multiple abnormal signals. Due to the patient's atypical clinical symptoms and laboratory test results, determining the etiology and treatment plan was difficulty.Subsequently, the patient underwent next-generation sequencing examination of cerebrospinal fluid. The following day, the results indicated the presence of Toxoplasma gondii. The patient received treatment with a combination of sulfamethoxazole (SMZ) and azithromycin. After approximately 7 days, the patient's symptoms significantly improved, and they were discharged from the hospital with oral medication to continue at home. A follow-up polymerase chain reaction (PCR) testing after about 6 weeks revealed the absence of Toxoplasma. Conclusion: This case highlights the potential of mNGS as an effective method for detecting toxoplasmic encephalitis (TE). Since mNGS can identify thousands of pathogens in a single run, it may be a promising detection method for investigating the causative pathogens of central nervous system infections with atypical features.

SpineRegNet: Spine Registration Network for volumetric MR and CT image by the joint estimation of an affine-elastic deformation field.

Spine registration for volumetric magnetic resonance (MR) and computed tomography (CT) images plays a significant role in surgical planning and surgical navigation system for the radiofrequency ablation of spine intervertebral discs. The affine transformation of each vertebra and elastic deformation of the intervertebral disc exist at the same time. This situation is a major challenge in spine registration. Existing spinal image registration methods failed to solve the optimal affine-elastic deformation field (AEDF) simultaneously, only consider the overall rigid or elastic alignment with the help of a manual spine mask, and encounter difficulty in meeting the accuracy requirements of clinical registration application. In this study, we propose a novel affine-elastic registration framework named SpineRegNet. The SpineRegNet consists of a Multiple Affine Matrices Estimation (MAME) Module for multiple vertebrae alignment, an Affine-Elastic Fusion (AEF) Module for joint estimation of the overall AEDF, and a Local Rigidity Constraint (LRC) Module for preserving the rigidity of each vertebra. Experiments on T2-weighted volumetric MR and CT images show that the proposed approach achieves impressive performance with mean Dice similarity coefficients of 91.36%, 81.60%, and 83.08% for the mask of the vertebrae on Datasets A-C, respectively. The proposed technique does not require a mask or manual participation during the tests and provides a useful tool for clinical spinal disease surgical planning and surgical navigation systems.

Massively Concurrent Sub-Cellular Traction Force Videography enabled by Single-Pixel Optical Tracers (SPOTs).

We report a large field-of-view and high-speed videography platform for measuring the sub-cellular traction forces of more than 10,000 biological cells over 13mm 2 at 83 frames per second. Our Single-Pixel Optical Tracers (SPOT) tool uses 2-dimensional diffraction gratings embedded into a soft substrate to convert cells' mechanical traction stress into optical colors detectable by a video camera. The platform measures the sub-cellular traction forces of diverse cell types, including tightly connected tissue sheets and near isolated cells. We used this platform to explore the mechanical wave propagation in a tightly connected sheet of Neonatal Rat Ventricular Myocytes (NRVMs) and discovered that the activation time of some tissue regions are heterogeneous from the overall spiral wave behavior of the cardiac wave. One-Sentence Summary: An optical platform for fast, concurrent measurements of cell mechanics at 83 frames per second, over a large area of 13mm 2 .

DDTNet: A dense dual-task network for tumor-infiltrating lymphocyte detection and segmentation in histopathological images of breast cancer.

The morphological evaluation of tumor-infiltrating lymphocytes (TILs) in hematoxylin and eosin (H& E)-stained histopathological images is the key to breast cancer (BCa) diagnosis, prognosis, and therapeutic response prediction. For now, the qualitative assessment of TILs is carried out by pathologists, and computer-aided automatic lymphocyte measurement is still a great challenge because of the small size and complex distribution of lymphocytes. In this paper, we propose a novel dense dual-task network (DDTNet) to simultaneously achieve automatic TIL detection and segmentation in histopathological images. DDTNet consists of a backbone network (i.e., feature pyramid network) for extracting multi-scale morphological characteristics of TILs, a detection module for the localization of TIL centers, and a segmentation module for the delineation of TIL boundaries, where a boundary-aware branch is further used to provide a shape prior to segmentation. An effective feature fusion strategy is utilized to introduce multi-scale features with lymphocyte location information from highly correlated branches for precise segmentation. Experiments on three independent lymphocyte datasets of BCa demonstrate that DDTNet outperforms other advanced methods in detection and segmentation metrics. As part of this work, we also propose a semi-automatic method (TILAnno) to generate high-quality boundary annotations for TILs in H& E-stained histopathological images. TILAnno is used to produce a new lymphocyte dataset that contains 5029 annotated lymphocyte boundaries, which have been released to facilitate computational histopathology in the future.

PCAN: Pixel-wise classification and attention network for thoracic disease classification and weakly supervised localization.

Automatic chest X-ray (CXR) disease classification has drawn increasing public attention as CXR is widely used in thoracic disease diagnosis. Existing classification networks typically employ a global average pooling layer to produce the final feature for the subsequent classifier. This limits the classification performance owing to the characteristics of lesions in CXR images, including small relative sizes, varied absolute sizes, and different occurrence locations. In this study, we propose a pixel-wise classification and attention network (PCAN) to simultaneously perform disease classification and weakly supervised localization, which provides interpretability for disease classification. The PCAN comprises a backbone network for extracting mid-level features, a pixel-wise classification branch (pc-branch) for generating pixel-wise diagnoses, and a pixel-wise attention branch (pa-branch) for producing pixel-wise weights. The pc-branch is capable of explicitly detecting small lesions, and the pa-branch is capable of adaptively focusing on different regions when classifying different thoracic diseases. Then, the pixel-wise diagnoses are multiplied with the pixel-wise weights to obtain the disease localization map, which provides the sizes and locations of lesions in a manner of weakly supervised learning. The final image-wise diagnosis is obtained by summing up the disease localization map at the spatial dimension. Comprehensive experiments conducted on the ChestX-ray14 and CheXpert datasets demonstrate the effectiveness of the proposed PCAN, which has great potential for thoracic disease diagnosis and treatment. The source codes are available at https://github.com/fzfs/PCAN.

Distributed colorimetric interferometer for mapping the pressure distribution in a complex microfluidics network.

We demonstrate a novel platform for mapping the pressure distribution of complex microfluidics networks with high spatial resolution. Our approach utilizes colorimetric interferometers enabled by lossy optical resonant cavities embedded in a silicon substrate. Detection of local pressures in real-time within a fluid network occurs by monitoring a reflected color emanating from each optical cavity. Pressure distribution measurements spanning a 1 cm2 area with a spatial resolution of 50 µm have been achieved. We applied a machine-learning-assisted sensor calibration method to generate a dynamic measurement range from 0 to 5.0 psi, with 0.2 psi accuracy. Adjustments to this dynamic measurement range are possible to meet different application needs for monitoring flow conditions in complex microfluidics networks, for the timely detection of anomalies such as clogging or leakage at their occurring locations.

Intracellular Photothermal Delivery for Suspension Cells Using Sharp Nanoscale Tips in Microwells.

Efficient intracellular delivery of biomolecules into cells that grow in suspension is of great interest for biomedical research, such as for applications in cancer immunotherapy. Although tremendous effort has been expended, it remains challenging for existing transfer platforms to deliver materials efficiently into suspension cells. Here, we demonstrate a high-efficiency photothermal delivery approach for suspension cells using sharp nanoscale metal-coated tips positioned at the edge of microwells, which provide controllable membrane disruption for each cell in an array. Self-aligned microfabrication generates a uniform microwell array with three-dimensional nanoscale metallic sharp tip structures. Suspension cells self-position by gravity within each microwell in direct contact with eight sharp tips, where laser-induced cavitation bubbles generate transient pores in the cell membrane to facilitate intracellular delivery of extracellular cargo. A range of cargo sizes were tested on this platform using Ramos suspension B cells with an efficiency of >84% for Calcein green (0.6 kDa) and >45% for FITC-dextran (2000 kDa), with retained viability of >96% and a throughput of >100 000 cells delivered per minute. The bacterial enzyme ß-lactamase (29 kDa) was delivered into Ramos B cells and retained its biological activity, whereas a green fluorescence protein expression plasmid was delivered into Ramos B cells with a transfection efficiency of >58%, and a viability of >89% achieved.

Lift-off cell lithography for cell patterning with clean background.

We developed a highly efficient method for patterning cells by a novel and simple technique called lift-off cell lithography (LCL). Our approach borrows the key concept of lift-off lithography from microfabrication and utilizes a fully biocompatible process to achieve high-throughput, high-efficiency cell patterning with nearly zero background defects across a large surface area. Using LCL, we reproducibly achieved >70% patterning efficiency for both adherent and non-adherent cells with <1% defects in undesired areas.

Liquid Metal-Based Multifunctional Micropipette for 4D Single Cell Manipulation.

A novel manufacturing approach to fabricate liquid metal-based, multifunctional microcapillary pipettes able to provide electrodes with high electrical conductivity for high-frequency electrical stimulation and measurement is proposed. 4D single cell manipulation is realized by applying multifrequency, multiamplitude, and multiphase electrical signals to the microelectrodes near the pipette tip to create 3D dielectrophoretic trap and 1D electrorotation, simultaneously. Functions such as single cell trapping, patterning, transfer, and rotation are accomplished. Cell viability and multiday proliferation characterization has confirmed the biocompatibility of this approach. This is a simple, low-cost, and fast fabrication process that requires no cleanroom and photolithography step to manufacture 3D microelectrodes and microchannels for easy access to a wide user base for broad applications.