Enhancing Ulcerative Colitis Histological Image Segmentation with an Active Learning Framework
- García Torres, Fernando
- Santacroce, Giovanni
- Zammarchi, Irene
- Meseguer, Pablo
- Amor, Rocío del
- Joaquín Roca González (coord.)
- Dolores Ojados González (coord.)
- Juan Suardíaz Muro (coord.)
Editorial: Universidad Politécnica de Cartagena
ISBN: 978-84-17853-76-1
Año de publicación: 2023
Páginas: 360-363
Congreso: Congreso Anual de la Sociedad Española de Ingeniería Biomédica. CASEIB (41. 2023. Cartagena)
Tipo: Aportación congreso
Resumen
In the management of UC, a chronic inflammatory bowel disease, histological evaluation plays an essential role in tar- geting histological remission (HR) of intestinal inflammatory activity as the primary aim of treatment. Recent work in the field has proposed a novel index called PHRI that quantifies UC activity based on the presence or absence of neutrophils in different cellular compartments: lamina propria, cryptal epithelium, surface epithelium and cryptal lumen. Digiti- zation of tissue samples into whole-slide images (WSI) has permitted the implementation of computer vision algorithms. These can perform tasks such as semantic segmentation but they require detailed pixel-level annotations, which are labo- rious and time-consuming in gigapixel images such as WSI. To address this limitation, this work introduces an active learning (AL) algorithm, an innovative approach to alleviat- ing the workload of pathologists by improving the efficiency in identifying crucial ...