A survey on Organoid Image Analysis Platforms

By: Alireza Ranjbaran, Azadeh Nazemi

An in-vitro cell culture system is used for biological discoveries and hypothesis-driven research on a particular cell type to understand mechanistic or test pharmaceutical drugs. Conventional in-vitro cultures have been applied to primary cells and immortalised cell lines plated on 2D surfaces. However, they are unreliable in complex physiological environments and can not always predict in-vivo behaviour correctly. Organoids are multicellu... more
An in-vitro cell culture system is used for biological discoveries and hypothesis-driven research on a particular cell type to understand mechanistic or test pharmaceutical drugs. Conventional in-vitro cultures have been applied to primary cells and immortalised cell lines plated on 2D surfaces. However, they are unreliable in complex physiological environments and can not always predict in-vivo behaviour correctly. Organoids are multicellular spheroids of a primary donor or stem cells that are replaced in vitro cell culture systems and are widely used in biological, biomedical and translational studies. Native heterogeneity, microanatomy, and functionality of an organ or diseased tissue can be represented by three-dimensional in-vitro tissue models such as organoids. Organoids are essential in in-vitro models for drug discovery and personalised drug screening. Many imaging artefacts such as organoid occlusion, overlap, out-of-focus spheroids and considerable heterogeneity in size cause difficulty in conventional image processing. Despite the power of organoid models for biology, their size and shape have mostly not been considered. Drug responses depend on dynamic changes in individual organoid morphology, number and size, which means differences in organoid shape and size, movement through focal planes, and live-cell staining with limited options cause challenges for drug response and growth analysis. This study primarily introduces the importance of the role of the organoid culture system in different disciplines of medical science and various scopes of utilising organoids. Then studies the challenges of operating organoids, followed by reviewing image analysis systems or platforms applied to organoids to address organoid utilising challenges. less
PocketNet: A Smaller Neural Network for Medical Image Analysis

By: Adrian Celaya, Jonas A. Actor, Rajarajeswari Muthusivarajan, Evan Gates, Caroline Chung, Dawid Schellingerhout, Beatrice Riviere, David Fuentes

Medical imaging deep learning models are often large and complex, requiring specialized hardware to train and evaluate these models. To address such issues, we propose the PocketNet paradigm to reduce the size of deep learning models by throttling the growth of the number of channels in convolutional neural networks. We demonstrate that, for a range of segmentation and classification tasks, PocketNet architectures produce results comparable... more
Medical imaging deep learning models are often large and complex, requiring specialized hardware to train and evaluate these models. To address such issues, we propose the PocketNet paradigm to reduce the size of deep learning models by throttling the growth of the number of channels in convolutional neural networks. We demonstrate that, for a range of segmentation and classification tasks, PocketNet architectures produce results comparable to that of conventional neural networks while reducing the number of parameters by multiple orders of magnitude, using up to 90% less GPU memory, and speeding up training times by up to 40%, thereby allowing such models to be trained and deployed in resource-constrained settings. less