Related Experiment Video
Updated: May 31, 2025
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
Published on: April 13, 2013
Impact of Dataset Size on 3D CNN Performance in Intracranial Hemorrhage Classification
Chun-Chao Huang1,2, Hsin-Fan Chiang1,2,3, Cheng-Chih Hsieh1,2,3
1Department of Radiology, MacKay Memorial Hospital, Taipei 104, Taiwan.
Insights
Larger sample sizes improve artificial intelligence (AI) model performance for detecting intracranial hemorrhage (ICH) using 3D convolutional neural networks (3DCNNs). Limited data can lead to inaccurate predictions, highlighting the need for sufficient cases for reliable AI development in medical imaging.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Neurology
Background:
- Evaluating the impact of sample size on developing 3D convolutional neural network (3DCNN) models for intracranial hemorrhage (ICH) detection.
- Investigating the binary classification of intraparenchymal, subarachnoid, and subdural hemorrhage (IPH, SAH, SDH).
Purpose of the Study:
- To assess how varying sample sizes affect the performance of 3DCNN models in classifying different types of ICH.
- To determine the minimum sample size required for reliable AI model development in ICH detection.
Main Methods:
- Compiled brain CT scan images into 3D images for 3DCNN model training.
- Varied non-hemorrhage and ICH case numbers to test model performance.
- Utilized cross-validation to compute average area under the curve (AUC) and accuracy for model evaluation.
Main Results:
- Larger sample sizes resulted in stable and acceptable AI model performance.
- Training with limited cases risked falsely high AUC values or accuracy.
- Model performance varied across ICH types (IPH, SAH, SDH), with more stability observed at larger sample sizes.
Conclusions:
- 3DCNN models can detect ICH even with limited data, but a minimum case number is essential.
- AI model performance for ICH detection is sample size-dependent and varies by hemorrhage type.
- Larger datasets enhance the stability and reliability of AI models for diagnosing intracranial hemorrhages.
Abstract:
Background: This study aimed to evaluate the effect of sample size on the development of a three-dimensional convolutional neural network (3DCNN) model for predicting the binary classification of three types of intracranial hemorrhage (ICH): intraparenchymal, subarachnoid, and subdural (IPH, SAH, SDH, respectively). Methods: During the training, we compiled all images of each brain computed tomography scan into a single 3D image, which was then fed into the model to classify the presence of ICH. We divided the non-hemorrhage quantities into 20, 30, 40, 50, 100, and 150 and the ICH quantities into 20, 30, 40, and 50. Cross-validation was performed to compute the average area under the curve (AUC) over the last five iterations. The AUC and accuracy were used to evaluate the performance of the models. Results: Fifty patients, each with the three ICH types, and 150 non-hemorrhage cases were enrolled. Larger sample sizes achieved stable and acceptable performance in the artificial intelligence (AI) models, whereas training with a limited number of cases posed the risk of falsely high AUC values or accuracy. The overall trends and fluctuations in AUC values were similar between IPH and SDH but different for SAH. The accuracy of the results was relatively consistent among the three ICH types. Conclusions: The 3DCNN technique can be used to develop AI models capable of detecting ICH from limited case numbers. However, a minimal case number must be provided. The performance of AI models varies across different ICH types and is more stable with larger sample sizes.
