no code implementations • 24 Aug 2023 • Shan Guleria, Benjamin Schwartz, Yash Sharma, Philip Fernandes, James Jablonski, Sodiq Adewole, Sanjana Srivastava, Fisher Rhoads, Michael Porter, Michelle Yeghyayan, Dylan Hyatt, Andrew Copland, Lubaina Ehsan, Donald Brown, Sana Syed
Discussion: Each of these five challenges is addressed, in part, by one of our AI-based models.
no code implementations • 18 Oct 2021 • Sodiq Adewole, Philip Fernandes, James Jablonski, Andrew Copland, Michael Porter, Sana Syed, Donald Brown
In this paper, we propose an end-to-end temporal abnormality localization for long WCE videos using only weak video level labels.
no code implementations • 18 Oct 2021 • Sodiq Adewole, Philip Fernandes, James Jablonski, Andrew Copland, Michael Porter, Sana Syed, Donald Brown
However, the search for temporal boundaries in a long video using high dimensional frame-feature matrix is computationally prohibitive and impracticable for real clinical application.
no code implementations • 11 Jan 2021 • Sodiq Adewole, Philip Fernandez, Michelle Yeghyayan, James Jablonski, Andrew Copland, Michael Porter, Sana Syed, Donald Brown
Automating the detection of frames containing specific lesion in WCE video would relieve gastroenterologists the arduous task of reviewing the entire video before making diagnosis.