Crowd‐powered histopathological analysis of brain samples from persons with Alzheimer disease: a Citizen Science pilot study inclusive of Hispanic communities

Through methodological development, including testing of annotation interfaces and consensus methods, we found with sufficient aggregation of non-expert answers, it is possible to consistently achieve a level of agreement with gold standard data meeting or exceeding interrater agreement among expert annotators for both stages of analysis CONCLUSION: Our approaches support the rapid production of high-volume annotation datasets for use in machine learning algorithm development to provide deeper phenotyping of neurodegenerative diseases. These pilot results support the plausibility of moving forward with an openly available online platform, which will be used to execute the first ever de-novo analysis of human brain histopathology images by a crowd of volunteers. This platform will be geared towards impacted populations, such as those of Hispanic heritage, by providing a culturally responsive outreach and inclusion program for such communities to represent themselves in discussions with researchers.