A High Rate of Human Error in Early Detection of Small Brain Metastases Suggests a Basis for Development of Artificial Intelligence Recognition Technology
JSHS · 2024
Overview
Brain Metastases carry poor prognosis in cancer patients with their rapidly growing nature. Early identification of these tumors is crucial in improving patient survival. This investigation aims to study the rate of human error in missing early brain metastases and factors associated with the threshold of sensitivity of human eyes. Data from the University of Nebraska Medical Center was used. The database included patients with new brain metastasis diagnosed based on brain magnetic resonance imaging (MRI) who also had previous MRI scan(s) 1-6 months before diagnosis and no exposure to whole-brain radiotherapy. The brain MRI used for diagnosis of brain metastasis and the MRI performed 1 -6 months prior were reviewed. Based on the location of the newly diagnosed tumor, the corresponding location in the previous MRI was assessed for a missed incidence of a preexisting tumor. The sizes of the missed tumors were then measured to assess the threshold of human eyes in detecting metastases. The percentage of missed metastases was 44% (56/126). The mean size of missed metastases was 3.0 millimeters (range 1.2 to 7.7 mm). No clinical factors were significantly associated with a higher rate of missed diagnosis. The most likely reason for the missed diagnosis is the tiny size although visual distraction seems to play a role including adjacent contrast-enhancing structures such as blood vessels. The results show a high rate of human error for missing small metastases. These results justify the development of artificial intelligence-based recognition to assist neuroradiologists in diagnosis.
Competition history
- JSHS 2024
Resources
Related projects
ISEF · 2025
Revolutionizing Eye Care: AI-Assisted Detection of Choroidal Melanoma
ISEF · 2023
Detection of Malignant Lung Tumors From CT Scans Through Deep Learning-Based Artificial Intelligence Algorithms
CWSF · 2026
Beyond Accuracy: AI Brain Tumor Detection with GradCAM++ Interpretability & Clinical Deployment
ISEF · 2018
Screening Malignant Glioma Using an Electrical Differential Impedance Spectrometer and Artificial Neural Network
ISEF · 2023
NueROX: A Visual Diagnosing System Using a Fully Optimized Convolutional Neural Network Architecture for Rapid and Efficient Classification of Tumorous and Non-Tumorous Brain MRI's and a Model for the Segmentation, Dimension, and Severity Estimation of Brain Tumors
ISEF · 2019
Neural Networks and Cancer Detection
ISEF · 2022
Detection of Benign and Malignant Lung Nodules in 3D Volumes Generated From Thoracic Computed Tomography Scans Leveraging Artificial Intelligence, Year 2
ISEF · 2024
Detection of MRI Imaged Brain Tumors Using EfficientNet-Based Convolutional Neural Networks
Closest projects by meaning, across every fair and year in the corpus.