What is it about?

Intracranial haemorrhage (ICH) is a medical condition that can have life-threatening consequences if it is not promptly diagnosed and treated. Diagnostic tools for medical imaging, such computed tomography (CT) scans, can help identify ICH.

Featured Image

Why is it important?

Due to the high mortality rate observed in ICH, non-contrast computed tomography (CT) early detection and categorization are crucial for assuring a good outcome and preventing the development of neurologic impairments. However, an increase in CT scans is likely to delay the early diagnosis of ICHs because radiologists who analyze the scans are slower to respond, particularly in academic institutions. As a result, a deep-learning Artificial Intelligence (AI) solution system has been developed to detect ICH . Deep learning models are often integrated with transfer learning models to reduce the computational complexity and potential to improve the accuracy.

Perspectives

In this study, we investigated the application of deep learning methods to CT scan images used for ICH diagnosis. In particular, we made use of the Block Matching 3D (BM3D) algorithm for image pre-processing, the Densenet121 neural network design, and the Hough Transformation for image transformation. The objective was to create a model that could correctly identify and categorize ICH in CT scan pictures. The region of interest (ROI) in the CT scan pictures was then located and extracted using the Hough Transformation approach. Overall, the findings of this study show the potential of deep learning methods to support ICH diagnosis from CT scan pictures. The suggested approach was highly accurate and has the potential to help radiologists identify ICH.

Ilavarasi A K
Vellore Institute of Technology, Chennai

Read the Original

This page is a summary of: Intracranial Haemorrhage Detection Based on Deep Learning Using CT Images, August 2023, ACM (Association for Computing Machinery),
DOI: 10.1145/3607947.3608034.
You can read the full text:

Read

Contributors

The following have contributed to this page