DELiVR's virtual reality training accelerates cell recognition in complex brain datasets

DELiVR's virtual reality training accelerates cell recognition in complex brain datasets

In a recent study published inNatural methods,Researchers introduced DELiVR, a virtual reality (VR)-enhanced deep learning pipeline for efficiently detecting neural activity in brain imaging, and provided an easy-to-use tool that improves the accuracy of data annotation and segmentation.

background

Analysis of protein expression is crucial for understanding physiological and disease mechanisms. Traditional immunohistochemistry provides only limited insight from tissue sections, while tissue clearing with fluorescence imaging provides a comprehensive view at the whole organism level.

Further research is needed to refine detection techniques, expand applications to different conditions, and fully understand the complex interactions within neural networks.

About the study

Researchers have developed a comprehensive method to process and analyze whole-brain immunolabeling using a modified SHANEL protocol.

This protocol involves several preparation steps including dehydration, rehydration and staining with c-Fos antibodies, which are used as a marker of neuronal activity. The process is enhanced through a series of washing and blocking steps to ensure specificity and clarity in labeling.

The team used light sheet microscopy to detect and visualize specific brain cells. This technique enables high-resolution imaging of brain tissue that has been processed transparently.

By using specific antibodies and advanced optical systems, researchers can capture detailed three-dimensional images of neuronal activity throughout the brain.

The researchers used automated and manual methods to analyze the massive amounts of data generated from these images. They developed a software pipeline called DELiVR that integrates VR and deep learning to optimize brain cell annotation and segmentation.

This system enables rapid and accurate identification of cell types and activity patterns, significantly speeding up the data analysis process. In addition to the technical advances, the study also focused on the practical application of these methods in biomedical research.

Study results

The team used the SHANEL protocol for whole brain c-Fos immunostaining, tissue clearing, and light sheet fluorescence microscopy (LSFM) to facilitate deep learning model training.

To effectively annotate these complex datasets, they switched from traditional two-dimensional (2D) layer-by-layer annotation to a more dynamic 3D approach using VR. This shift has been made possible through the use of commercial VR software such as Arivis VisionVR and syGlass, which allows annotators to fully immerse themselves in the volumetric data.

These tools significantly reduced annotation time and improved accuracy compared to the traditional methods used in ITK-SNAP.

The VR approach improved the training process of deep learning segmentation models by enabling fast and accurate annotation of regions of interest (ROIs) in three dimensions.

For example, using Arivis VisionVR, annotators could apply adaptive thresholds to defined ROIs based on their input, streamlining the annotation process. In contrast, traditional 2D annotation required segmentation of c-Fos+ cells in each image plane, a more time-consuming and error-prone method.

To fully leverage these annotated datasets, the team developed DELiVR, a comprehensive deep learning pipeline tailored for detailed analysis of neural activity across the brain.

DELiVR uses a series of steps to process and analyze brain images, from downsampling raw images to aligning segmented cells with the Allen Brain Atlas using sophisticated algorithms such as mBrainAligner.

The pipeline facilitates the detection and mapping of cells to specific brain regions, enabling a better understanding of neuronal activity that surpasses previous non-deep learning models.

The effectiveness of DELiVR was validated compared to traditional methods and showed significant improvement in detection accuracy and sensitivity. The deep learning pipeline increased the number of cells detected and improved the precision of those detections, outperforming established methods such as ClearMap.

For visualization, DELiVR creates a detailed map of segmented cells and colors each cell according to its brain region, which can be further visualized using tools such as BrainRender.

DELiVR's flexibility also extends to its use; It is packaged in an easy-to-use Docker container that runs on both Linux and Windows.

This package includes a special Fiji plugin that simplifies the use of DELiVR for researchers with different technical expertise. In addition, the system allows retraining on different datasets, improving its adaptability and precision for different research needs.

Additionally, DELiVR's capabilities were demonstrated in a study of cancer-related changes in brain activity. The pipeline was used to compare neuronal activity patterns between mice with different types of cancer, revealing significant differences in brain activity associated with cancer-related cachexia.

Conclusions

In summary, the team introduced DELiVR, a VR-enabled deep learning pipeline for whole brain cell mapping in mice, designed to be accessible to biologists without programming knowledge via a Fijian interface. DELiVR leverages VR for precise training annotations, improves segmentation accuracy, and easily integrates with existing datasets.

Traditional methods like ClearMap, which often miss subtle variations due to noise, are outperformed by DELiVR's 3D BasicUNet.

The tool demonstrated its effectiveness by profiling brain activation in mice with cancer and uncovering distinct neural patterns related to weight management. DELiVR combines high accuracy in cell mapping with user-friendly features, advancing the study of neurophysiological phenomena in the context of disease.

Sources: