Cognitive functioning in untreated glioma patients: The limited predictive value of clinical variables.

BACKGROUND: Previous research identified many clinical variables that are significantly related to cognitive functioning before surgery. It is not clear whether such variables enable accurate prediction for individual patients' cognitive functioning because statistical significance does not guarantee predictive value. Previous studies did not test how well cognitive functioning can be predicted for (yet) untested patients. Furthermore, previous research is limited in that only linear or rank-based methods with small numbers of variables were used. METHODS: We used various machine learning models to predict preoperative cognitive functioning for 340 patients with glioma across 18 outcome measures. Predictions were made using a comprehensive set of clinical variables as identified from the literature. Model performances and optimized hyperparameters were interpreted. Moreover, Shapley additive explanations were calculated to determine variable importance and explore interaction effects. RESULTS: Best-performing models generally demonstrated above-random performance. Performance, however, was unreliable for 14 out of 18 outcome measures with predictions worse than baseline models for a substantial number of train-test splits. Best-performing models were relatively simple and used most variables for prediction while not relying strongly on any variable. CONCLUSIONS: Preoperative cognitive functioning could not be reliably predicted across cognitive tests using the comprehensive set of clinical variables included in the current study. Our results show that a holistic view of an individual patient likely is necessary to explain differences in cognitive functioning. Moreover, they emphasize the need to collect larger cross-center and multimodal data sets.

Detecting cell division of Pseudomonas aeruginosa bacteria from bright-field microscopy images with hidden conditional random fields.

An approach to automatically detect bacteria division with temporal models is presented. To understand how bacteria migrate and proliferate to form complex multicellular behaviours such as biofilms, it is desirable to track individual bacteria and detect cell division events. Unlike eukaryotic cells, prokaryotic cells such as bacteria lack distinctive features, causing bacteria division difficult to detect in a single image frame. Furthermore, bacteria may detach, migrate close to other bacteria and may orientate themselves at an angle to the horizontal plane. Our system trains a hidden conditional random field (HCRF) model from tracked and aligned bacteria division sequences. The HCRF model classifies a set of image frames as division or otherwise. The performance of our HCRF model is compared with a Hidden Markov Model (HMM). The results show that a HCRF classifier outperforms a HMM classifier. From 2D bright field microscopy data, it is a challenge to separate individual bacteria and associate observations to tracks. Automatic detection of sequences with bacteria division will improve tracking accuracy.

Segmentation of densely populated cell nuclei from confocal image stacks using 3D non-parametric shape priors.

An approach to jointly estimate 3D shapes and poses of stained nuclei from confocal microscopy images, using statistical prior information, is presented. Extracting nuclei boundaries from our experimental images of cell migration is challenging due to clustered nuclei and variations in their shapes. This issue is formulated as a maximum a posteriori estimation problem. By incorporating statistical prior models of 3D nuclei shapes into level set functions, the active contour evolutions applied on the images is constrained. A 3D alignment algorithm is developed to build the training databases and to match contours obtained from the images to them. To address the issue of aligning the model over multiple clustered nuclei, a watershed-like technique is used to detect and separate clustered regions prior to active contour evolution. Our method is tested on confocal images of endothelial cells in microfluidic devices, compared with existing approaches.

A Bayesian filtering approach to incorporate 2D/3D time-lapse confocal images for tracking angiogenic sprouting cells interacting with the gel matrix.

We present a new approach to incorporating information from heterogeneous images of migrating cells in 3D gel. We study 3D angiogenic sprouting, where cells burrow into the gel matrix, communicate with other cells and create vascular networks. We combine time-lapse fluorescent images of stained cell nuclei and transmitted light images of the background gel to track cell trajectories. The nuclei images are sampled less frequently due to photo toxicity. Hence, 3D cell tracking can be performed more reliably when 2D sprout profiles, extracted from gel matrix images, are effectively incorporated. We employ a Bayesian filtering approach to optimally combine the two heterogeneous images with different sampling rates. We construct stochastic models to predict cell locations and sprout profiles and condition the likelihood of nuclei location by the sprout profile. The conditional distribution is non-Gaussian and the cell dynamics is non-linear. To jointly update cell and sprout estimates, we use a Rao-Blackwell particle filter. Simulation and experimental results show accurate tracking of multiple cells along with sprout formation, demonstrating synergistic effects of incorporating the two types of images.

Formation of elongated fascicle-inspired 3D tissues consisting of high-density, aligned cells using sacrificial outer molding.

The majority of muscles, nerves, and tendons are composed of fiber-like fascicle morphology. Each fascicle has a) elongated cells highly aligned with the length of the construct, b) a high volumetric cell density, and c) a high length-to-width ratio with a diameter small enough to facilitate perfusion. Fiber-like fascicles are important building blocks for forming tissues of various sizes and cross-sectional shapes, yet no effective technology is currently available for producing long and thin fascicle-like constructs with aligned, high-density cells. Here we present a method for molding cell-laden hydrogels that generate cylindrical tissue structures that are ~100 µm in diameter with an extremely high length to diameter ratio (>100 : 1). Using this method we have successfully created skeletal muscle tissue with a high volumetric density (~50%) and perfect cell alignment along the axis. A new molding technique, sacrificial outer molding, allows us to i) create a long and thin cylindrical cavity of the desired size in a sacrificial mold that is solid at a low temperature, ii) release gelling agents from the sacrificial mold material after the cell-laden hydrogel is injected into fiber cavities, iii) generate a uniform axial tension between anchor points at both ends that promotes cell alignment and maturation, and iv) perfuse the tissue effectively by exposing it to media after melting the sacrificial outer mold at 37 °C. The effects of key parameters and conditions, including initial cavity diameter, axial tension, and concentrations of the hydrogel and gelling agent upon tissue compaction, volumetric cell density, and cell alignment are presented.

In vivo label-free quantification of liver microcirculation using dual-modality microscopy.

Microcirculation lesion is a common symptom of chronic liver diseases in the form of vasculature deformation and circulation alteration. In acute to chronic liver diseases such as biliary atresia, microcirculation lesion can have an early onset. Detection of microcirculation lesion is meaningful for studying the progression of liver disease. We have combined wide-field fluorescence microscopy and a laser speckle contrast technique to characterize hepatic microcirculation in vivo without labeling in a bile-duct ligation rat fibrosis model of biliary atresia. Through quantitative image analysis of four microcirculation parameters, we observed significant microcirculation lesion in the early to middle stages of fibrosis. This bimodal imaging method is useful to assess hepatic microcirculation lesion for the study of liver diseases.