Generating Realistic Brain MRIs via a Conditional Diffusion Probabilistic Model.

As acquiring MRIs is expensive, neuroscience studies struggle to attain a sufficient number of them for properly training deep learning models. This challenge could be reduced by MRI synthesis, for which Generative Adversarial Networks (GANs) are popular. GANs, however, are commonly unstable and struggle with creating diverse and high-quality data. A more stable alternative is Diffusion Probabilistic Models (DPMs) with a fine-grained training strategy. To overcome their need for extensive computational resources, we propose a conditional DPM (cDPM) with a memory-efficient process that generates realistic-looking brain MRIs. To this end, we train a 2D cDPM to generate an MRI subvolume conditioned on another subset of slices from the same MRI. By generating slices using arbitrary combinations between condition and target slices, the model only requires limited computational resources to learn interdependencies between slices even if they are spatially far apart. After having learned these dependencies via an attention network, a new anatomy-consistent 3D brain MRI is generated by repeatedly applying the cDPM. Our experiments demonstrate that our method can generate high-quality 3D MRIs that share a similar distribution to real MRIs while still diversifying the training set. The code is available at https://github.com/xiaoiker/mask3DMRI_diffusion and also will be released as part of MONAI, at https://github.com/Project-MONAI/GenerativeModels.

A performance improvement of enzyme-based electrochemical lactate sensor fabricated by electroplating novel PdCu mediator on a laser induced graphene electrode.

A lactate sensor for lactate sensing using porous laser-induced graphene (LIG) electrodes with an electrodeposited PdCu catalyst was developed in this study. CO2 laser was used to convert the polyimide film surface to multilayered LIG. The morphology and composition of LIG were analyzed through field-emission scanning electron microscopy and Raman spectroscopy, respectively, to confirm that the fabricated LIG electrode was composed of porous and stacked graphene layers. PdCu was electrodeposited on the LIG electrode and lactate oxidase (LOx) was immobilized on the LIG surface to create a LOx/PdCu/LIG structure. According to the Randles-Sevcík equation, the calculated active surface area of the fabricated PdCu/LIG electrode was ∼12.8 mm2, which was larger than the apparent area of PdCu/LIG (1.766 mm2) by a factor of 7.25. The measured sensitivities of the fabricated lactate sensors with the LOx/PdCu/LIG electrode were -51.91 µA/mM·cm2 (0.1-5 mM) and -17.18 µA/mM·cm2 (5-30 mM). The calculated limit of detection was 0.28 µM. The selectivity of the fabricated lactate sensor is excellent toward various potentially interfering materials such as ascorbic acid, uric acid, lactose, sucrose, K+ and Na+.

Weakly supervised segmentation on neural compressed histopathology with self-equivariant regularization.

In digital pathology, segmentation is a fundamental task for the diagnosis and treatment of diseases. Existing fully supervised methods often require accurate pixel-level annotations that are both time-consuming and laborious to generate. Typical approaches first pre-process histology images into patches to meet memory constraints and later perform stitching for segmentation; at times leading to lower performance given the lack of global context. Since image level labels are cheaper to acquire, weakly supervised learning is a more practical alternative for training segmentation algorithms. In this work, we present a weakly supervised framework for histopathology segmentation using only image-level labels by refining class activation maps (CAM) with self-supervision. First, we compress gigapixel histology images with an unsupervised contrastive learning technique to retain high-level spatial context. Second, a network is trained on the compressed images to jointly predict image-labels and refine the initial CAMs via self-supervised losses. In particular, we achieve refinement via a pixel correlation module (PCM) that leverages self-attention between the initial CAM and the input to encourage fine-grained activations. Also, we introduce a feature masking technique that performs spatial dropout on the compressed input to suppress low confidence predictions. To effectively train our model, we propose a loss function that includes a classification objective with image-labels, self-supervised regularization and entropy minimization between the CAM predictions. Experimental results on two curated datasets show that our approach is comparable to fully-supervised methods and can outperform existing state-of-the-art patch-based methods. https://github.com/PhilipChicco/wsshisto.

Sciatic nerve injection injury.

Nerve injury is a common complication following intramuscular injection and the sciatic nerve is the most frequently affected nerve, especially in children, the elderly and underweight patients. The neurological presentation may range from minor transient pain to severe sensory disturbance and motor loss with poor recovery. Management of nerve injection injury includes drug treatment of pain, physiotherapy, use of assistive devices and surgical exploration. Early recognition of nerve injection injury and appropriate management are crucial in order to reduce neurological deficit and to maximize recovery. Sciatic nerve injection injury is a preventable event. Total avoidance of intramuscular injection is recommended if other administration routes can be used. If the injection has to be administered into the gluteal muscle, the ventrogluteal region (gluteal triangle) has a more favourable safety profile than the dorsogluteal region (the upper outer quadrant of the buttock).