A novel cascade machine learning pipeline for Alzheimer’s disease identification and prediction

Published on January 16, 2023

Imagine you’re a detective, trying to solve the complex puzzle that is Alzheimer’s disease. You know that one key piece of evidence is the size of the hippocampus, a region in the brain that is often affected by the disease. Well, our team of scientific sleuths has developed a groundbreaking machine learning pipeline called AL-SCF. It starts by using a special tool to map out and identify the hippocampus in brain images. Then, it analyzes over 850 radiomics features and selects the 37 most relevant ones to Alzheimer’s disease. These features are like clues that help distinguish between Alzheimer’s patients and healthy individuals. Finally, four sophisticated classifiers are used to make the final determination, with impressive accuracy and precision. This amazing pipeline has shown incredible promise in both training and validation sets, with an AUC of 0.97 and a Dice score of 0.93 respectively. By automating the steps from segmentation to classification, the AL-SCF pipeline can revolutionize Alzheimer’s diagnosis, helping doctors make quicker and more accurate assessments. With further research and refinement, this tool may lead to personalized treatment plans for Alzheimer’s patients. So grab your detective hat and dive into the fascinating world of our research!

IntroductionAlzheimer’s disease (AD) is a progressive and irreversible brain degenerative disorder early. Among all diagnostic strategies, hippocampal atrophy is considered a promising diagnostic method. In order to proactively detect patients with early Alzheimer’s disease, we built an Alzheimer’s segmentation and classification (AL-SCF) pipeline based on machine learning.MethodsIn our study, we collected coronal T1 weighted images that include 187 patients with AD and 230 normal controls (NCs). Our pipeline began with the segmentation of the hippocampus by using a modified U2-net. Subsequently, we extracted 851 radiomics features and selected 37 features most relevant to AD by the Hierarchical clustering method and Least Absolute Shrinkage and Selection Operator (LASSO) algorithm. At last, four classifiers were implemented to distinguish AD from NCs, and the performance of the models was evaluated by accuracy, specificity, sensitivity, and area under the curve.ResultsOur proposed pipeline showed excellent discriminative performance of classification with AD vs NC in the training set (AUC=0.97, 95% CI: (0.96-0.98)). The model was also verified in the validation set with Dice=0.93 for segmentation and accuracy=0.95 for classification.DiscussionThe AL-SCF pipeline can automate the process from segmentation to classification, which may assist doctors with AD diagnosis and develop individualized medical plans for AD in clinical practice.

Read Full Article (External Site)

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes:

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>