Assessment of Current Practices Across Alzheimer's Disease Research Centers Biorepositories.

In 1984, the National Institute on Aging developed the Alzheimer's disease centers program. The main goal of these centers is to advance the understanding of Alzheimer's disease and related dementias (ADRD) through comprehensive patient evaluations and cutting-edge research in pathology, laboratory medicine, education, and scientific discovery. The neuropathology core of the Alzheimer's Disease Research Centers (ADRCs) collects postmortem brain tissue from consented donors ranging from cognitively normal individuals to those with late-stage dementia, whose samples and data can be shared around the world to further advance knowledge, diagnosis, and to eventually find cures for ADRD. Although recommended guidelines for biorepositories exist, we aimed to understand the current practices within neuropathology cores across the ADRCs. A survey was developed that focused on information related to sample processing methods, biospecimen requests, financial costs related to the repository, and data management. This survey was distributed to 28 current and former ADRC neuropathology cores. The survey obtained a response rate of 82% (23/28). Although most centers were consistent in responses related to sample processing and storage, they varied widely in processes by which neuropathological samples are shared and cost recovery mechanisms. The results of this survey provide benchmark data on practices within neuropathology cores across ADRCs and the overlap with biorepository best practices. Future studies focused on understanding factors that may influence current practices (such as available funds and personnel) are need to aid in minimizing barriers to optimally follow best practices. Sharing these data among ADRCs will allow for improvement in workflows and working toward cures for ADRD.

BrainSec: Automated Brain Tissue Segmentation Pipeline for Scalable Neuropathological Analysis.

As neurodegenerative disease pathological hallmarks have been reported in both grey matter (GM) and white matter (WM) with different density distributions, automating the segmentation process of GM/WM would be extremely advantageous for aiding in neuropathologic deep phenotyping. Standard segmentation methods typically involve manual annotations, where a trained researcher traces the delineation of GM/WM in ultra-high-resolution Whole Slide Images (WSIs). This method can be time-consuming and subjective, preventing a scalable analysis on pathology images. This paper proposes an automated segmentation pipeline (BrainSec) combining a Convolutional Neural Network (CNN) module for segmenting GM/WM regions and a post-processing module to remove artifacts/residues of tissues. The final output generates XML annotations that can be visualized via Aperio ImageScope. First, we investigate two baseline models for medical image segmentation: FCN, and U-Net. Then we propose a patch-based approach, BrainSec, to classify the GM/WM/background regions. We demonstrate BrainSec is robust and has reliable performance by testing it on over 180 WSIs that incorporate numerous unique cases as well as distinct neuroanatomic brain regions. We also apply gradient-weighted class activation mapping (Grad-CAM) to interpret the segmentation masks and provide relevant explanations and insights. In addition, we have integrated BrainSec with an existing Amyloid-ß pathology classification model into a unified framework (without incurring significant computation complexity) to identify pathologies, visualize their distributions, and quantify each type of pathologies in segmented GM/WM regions, respectively.