Automated detection of brain metastases using machine learning
Dr. Syed Naqvi
Windsor Regional Hospital
FUNDER: Windsor Cancer Centre Foundation (WCCF)
Brain metastases are the most common type of malignant brain tumour in adults, with MRI being the most sensitive imaging modality for detecting intracranial metastases. However, the accurate detection of brain metastases on MRI can be challenging and time-consuming, especially when they are small, numerous, or located in difficult-to-see areas of the brain; radiologists often experience fatigue or overlook lesions due to the large number of images that need to be reviewed. To help address these challenges, automated detection of brain metastases using artificial intelligence (AI) and machine learning algorithms can be used and provide consistent, objective assessments, and an extra set of eyes to the radiologist. The current knowledge of efficiency and accuracy of AI in detecting brain metastases is limited, but this study plans to reduce that limitation.
Assessing the accuracy and efficacy of AI and machine learning algorithms for the detection of brain metastases have significant positive impacts on patient care. Improved accuracy of detecting brain metastases is critical for treatment planning and patient prognosis, increased efficacy of radiologists by reducing the time and resources required for lesion detection and therefore reducing the number of missed metastases. The use of machine learning algorithms could lead to earlier detection and treatment, leading to better patient outcomes. In future research, this form of AI and machine learning could be applied to detect and diagnose other forms of cancer.