Hippocampal Volume in Alzheimer's Disease: OASIS-1 Dataset Analysis

Overview

This study utilized data from the OASIS-1 database, comprising 416 right-handed subjects aged 18 to 96, with 3 -4 T1-weighted MRI scans per participant. The analysis focused on individuals over 60, including 91 Healthy Controls (HC) and 95 with mild cogniti ve impairment (MCI) or Alzheimer’s Disease (AD), excluding younger HC datasets and outliers with moderate dementia (CDR=2). Brain parcellation was performed using FreeSurfer (i.e., open -source software) to measure hippocampal volume, adjusted for gender and brain size differences using Atlas Scaling Factor (ASF). I hypothesize that there would be a negative correlation between bilateral hippocampal volume and Clinical Diagnosis Rating scores. To test this hypothesis, a multiple linear regression model contr olling for age, sex, and CDR scores was used to examine hippocampal volume differences between groups. Results showed significant volume reductions in AD and MCI groups compared to healthy controls. The study also found that CDR scores predicted hippocampa l volume, confirming the association between hippocampal atrophy and cognitive decline in AD. Education levels differed significantly between groups, suggesting a potential protective role against AD. The findings imply that hippocampal volume could serve as an early indicator and biomarker for AD risk and progression. However, limitations include the cross-sectional nature of the study, focus solely on hippocampal volume, and potential confounding effects of education. As OASIS data includes a longitudinal dataset, examining these brain changes with multiple assessments would be worthwhile to see these groups' structural changes and clinical diagnoses over time. Deep Learning for Crop Health: An Automated Approach to Systematic Corn Gray Leaf Spot Assessment and Management Henry Zou Johnston Senior High School, Johnston, IA Mentor Professor Jian Jin, Purdue University Gray Leaf Spot (GLS), caused by the fungal pathogens Cercospora zeae-maydis and Cercospora zeina, represents a significant challenge in corn production, resulting in annual economic losses amounting to billions of dollars. Conventional methods for assessing GLS severity are limited by high labor demands, subjectivity, and frequent inaccuracies. This study investigates the application of deep learning for automating GLS severity assessment, focusing on comparing the performances of a multi-model detection framework utilizing YOLOv8 Computer Vision Model and U-Net Convolutional Neural Network (CNN) to human experts in terms of accuracy and consistency. The YOLOv8 model was first utilized to extract leaf regions and disease type, followed by the U -Net CNN model to segment GLS lesions and calculate percent severity. The models were trained on the PlantVillage and Corn Disease & Severity datasets containing corn leaves infected with various diseases. The YOLOv8 model achieved precision accuracies of over 0.97 in classification of disease type, while the U-Net model demonstrated F1 accuracies of 0.89 in segmentation of diseased lesions. GLS severity assessments produced by the models showed a strong correlation and exhibited even higher levels of consistency than expert human evaluations, potentially making it a reliable tool for automated disease assessment. A third LLM model was integrated for field -application, providing tailored management strategies based on data from the first two models. Future studies will focus on integrating machine learning methods with field and greenhouse applications, obtaining an integrated disease score from multiple leaves and plants, and exploring their applications in other plant diseases. Illinois

Competition history

• JSHS 2025 Category not listed

Resources

• National JSHS abstract book ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.