Bayesian networks for Risk Assessment and postoperative deficit prediction in intraoperative neurophysiology for brain surgery.

PURPOSE: To this day there is no consensus regarding evidence of usefulness of Intraoperative Neurophysiological Monitoring (IONM). Randomized controlled trials have not been performed in the past mainly because of difficulties in recruitment control subjects. In this study, we propose the use of Bayesian Networks to assess evidence in IONM. METHODS: Single center retrospective study from January 2020 to January 2022. Patients admitted for cranial neurosurgery with intraoperative neuromonitoring were enrolled. We built a Bayesian Network with utility calculation using expert domain knowledge based on logistic regression as potential causal inference between events in surgery that could lead to central nervous system injury and postoperative neurological function. RESULTS: A total of 267 patients were included in the study: 198 (73.9%) underwent neuro-oncology surgery and 69 (26.1%) neurovascular surgery. 50.7% of patients were female while 49.3% were male. Using the Bayesian Network´s original state probabilities, we found that among patients who presented with a reversible signal change that was acted upon, 59% of patients would wake up with no new neurological deficits, 33% with a transitory deficit and 8% with a permanent deficit. If the signal change was permanent, in 16% of the patients the deficit would be transitory and in 51% it would be permanent. 33% of patients would wake up with no new postoperative deficit. Our network also shows that utility increases when corrective actions are taken to revert a signal change. CONCLUSIONS: Bayesian Networks are an effective way to audit clinical practice within IONM. We have found that IONM warnings can serve to prevent neurological deficits in patients, especially when corrective surgical action is taken to attempt to revert signals changes back to baseline properties. We show that Bayesian Networks could be used as a mathematical tool to calculate the utility of conducting IONM, which could save costs in healthcare when performed.

LiDAR-Based Intensity-Aware Outdoor 3D Object Detection.

LiDAR-based 3D object detection and localization are crucial components of autonomous navigation systems, including autonomous vehicles and mobile robots. Most existing LiDAR-based 3D object detection and localization approaches primarily use geometric or structural feature abstractions from LiDAR point clouds. However, these approaches can be susceptible to environmental noise due to adverse weather conditions or the presence of highly scattering media. In this work, we propose an intensity-aware voxel encoder for robust 3D object detection. The proposed voxel encoder generates an intensity histogram that describes the distribution of point intensities within a voxel and is used to enhance the voxel feature set. We integrate this intensity-aware encoder into an efficient single-stage voxel-based detector for 3D object detection. Experimental results obtained using the KITTI dataset show that our method achieves comparable results with respect to the state-of-the-art method for car objects in 3D detection and from a bird's-eye view and superior results for pedestrian and cyclic objects. Furthermore, our model can achieve a detection rate of 40.7 FPS during inference time, which is higher than that of the state-of-the-art methods and incurs a lower computational cost.

Effect of charge substitutions at residue his-142 on voltage gating of connexin43 channels.

Previous studies indicate that the carboxyl terminal of connexin43 (Cx43CT) is involved in fast transjunctional voltage gating. Separate studies support the notion of an intramolecular association between Cx43CT and a region of the cytoplasmic loop (amino acids 119-144; referred to as "L2"). Structural analysis of L2 shows two alpha-helical domains, each with a histidine residue in its sequence (H126 and H142). Here, we determined the effect of H142 replacement by lysine, alanine, and glutamate on the voltage gating of Cx43 channels. Mutation H142E led to a significant reduction in the frequency of occurrence of the residual state and a prolongation of dwell open time. Macroscopically, there was a large reduction in the fast component of voltage gating. These results resembled those observed for a mutant lacking the carboxyl terminal (CT) domain. NMR experiments showed that mutation H142E significantly decreased the Cx43CT-L2 interaction and disrupted the secondary structure of L2. Overall, our data support the hypothesis that fast voltage gating involves an intramolecular particle-receptor interaction between CT and L2. Some of the structural constrains of fast voltage gating may be shared with those involved in the chemical gating of Cx43.