Cell2Grid: an efficient, spatial, and convolutional neural network-ready representation of cell segmentation data.

Purpose: Cell segmentation algorithms are commonly used to analyze large histologic images as they facilitate interpretation, but on the other hand they complicate hypothesis-free spatial analysis. Therefore, many applications train convolutional neural networks (CNNs) on high-resolution images that resolve individual cells instead, but their practical application is severely limited by computational resources. In this work, we propose and investigate an alternative spatial data representation based on cell segmentation data for direct training of CNNs. Approach: We introduce and analyze the properties of Cell2Grid, an algorithm that generates compact images from cell segmentation data by placing individual cells into a low-resolution grid and resolves possible cell conflicts. For evaluation, we present a case study on colorectal cancer relapse prediction using fluorescent multiplex immunohistochemistry images. Results: We could generate Cell2Grid images at 5 - µ m resolution that were 100 times smaller than the original ones. Cell features, such as phenotype counts and nearest-neighbor cell distances, remain similar to those of original cell segmentation tables ( p < 0.0001 ). These images could be directly fed to a CNN for predicting colon cancer relapse. Our experiments showed that test set error rate was reduced by 25% compared with CNNs trained on images rescaled to 5 µ m with bilinear interpolation. Compared with images at 1 - µ m resolution (bilinear rescaling), our method reduced CNN training time by 85%. Conclusions: Cell2Grid is an efficient spatial data representation algorithm that enables the use of conventional CNNs on cell segmentation data. Its cell-based representation additionally opens a door for simplified model interpretation and synthetic image generation.

Recommendations for Better Adoption of Medical Photography as a Clinical Tool.

The use of photography in routine clinical practice has the potential to increase the efficiency of overall patient care as well as improve clinical documentation and provider-to-provider communication. This is particularly important in the setting of provider burnout in the electronic health record era and during the COVID-19 pandemic. Despite the potential of photographs to enhance workflows and patient care, challenges remain that hinder the successful incorporation of medical photography into clinical practice, often because of inconsistent structure and implementation. Our proposed consolidated framework for clinical photography consists of five key aspects: appropriate informed consent; proper preparation and positioning; image acquisition with consideration of the field of view, orientation, focus, resolution, scale, and color calibration; streamlined and secure image storage and documentation; and interoperable file exchange. Overall, this viewpoint is a forward-looking paper on leveraging medical photography as an electronic health record tool for clinical care, research, and education.

Memristive Cluster Based Compact High-Density Nonvolatile Memory Design and Application for Image Storage.

As a new type of nonvolatile device, the memristor has become one of the most promising technologies for designing a new generation of high-density memory. In this paper, a 4-bit high-density nonvolatile memory based on a memristor is designed and applied to image storage. Firstly, a memristor cluster structure consisting of a transistor and four memristors is designed. Furthermore, the memristor cluster is used as a memory cell in the crossbar array structure to realize the memory design. In addition, when the designed non-volatile memory is applied to gray scale image storage, only two memory cells are needed for the storage of one pixel. Through the Pspice circuit simulation, the results show that compared with the state-of-the-art technology, the memory designed in this paper has better storage density and read-write speed. When it is applied to image storage, it achieves the effect of no distortion and fast storage.

Data preparation for artificial intelligence in medical imaging: A comprehensive guide to open-access platforms and tools.

The vast amount of data produced by today's medical imaging systems has led medical professionals to turn to novel technologies in order to efficiently handle their data and exploit the rich information present in them. In this context, artificial intelligence (AI) is emerging as one of the most prominent solutions, promising to revolutionise every day clinical practice and medical research. The pillar supporting the development of reliable and robust AI algorithms is the appropriate preparation of the medical images to be used by the AI-driven solutions. Here, we provide a comprehensive guide for the necessary steps to prepare medical images prior to developing or applying AI algorithms. The main steps involved in a typical medical image preparation pipeline include: (i) image acquisition at clinical sites, (ii) image de-identification to remove personal information and protect patient privacy, (iii) data curation to control for image and associated information quality, (iv) image storage, and (v) image annotation. There exists a plethora of open access tools to perform each of the aforementioned tasks and are hereby reviewed. Furthermore, we detail medical image repositories covering different organs and diseases. Such repositories are constantly increasing and enriched with the advent of big data. Lastly, we offer directions for future work in this rapidly evolving field.

Digital Photography as a Tool of Research and Documentation in Plant Tissue Culture.

Scientific photography is an important and indispensable tool in plant tissue culture research: photographs should be taken throughout a project for documentation. The aim of photography in plant tissue culture should be to illustrate clearly the differentiation, growth, and developmental stages occurring in vitro. Poor-quality scientific photography in tissue culture research and professional reports results in poor documentation. If visual aspects of the tissue culture are not well documented or not well reproduced in the image, an important part of the research is missed, the resulting report is of limited scientific value, and the research results may not be reproducible. Simple methods for improving the results of photography of materials from plant tissue culture are described and discussed, along with the necessary photographic equipment, suitable backgrounds, the construction of photographic plates, and correct use of electronic files for images. Finally, ethical concerns about image manipulation are discussed.