Weakly supervised deep learning-based intracranial hemorrhage localization
This addresses the need for fast medical intervention in life-threatening cases by reducing annotation time, though it is incremental as it builds on weakly supervised methods.
The paper tackles the problem of localizing intracranial hemorrhage in head CT images using only slice-level labels, achieving a Dice coefficient of 58.08% on a public dataset.
Intracranial hemorrhage is a life-threatening disease, which requires fast medical intervention. Owing to the duration of data annotation, head CT images are usually available only with slice-level labeling. This paper presents a weakly supervised method of precise hemorrhage localization in axial slices using only position-free labels, which is based on multiple instance learning. An algorithm is introduced that generates hemorrhage likelihood maps and finds the coordinates of bleeding. The Dice coefficient of 58.08 % is achieved on data from a publicly available dataset.