Hunting imaging biomarkers in pulmonary fibrosis: Benchmarks of the AIIB23 challenge

• DOI: 10.48550/arxiv.2312.13752
• Publication Date: 2023-01-01
• Authors: Nan Yang; Xiaodan Xing; Shiyi Wang; Zeyu Tang; Federico N Felder; Sheng Zhang; Roberta Eufrasia ...; Xiaoliu Ding; Ruiqi Yu; Weiping Liu; Feng Shi; Tianyang Sun; Zehong Cao; Minghui Zhang; Yun Gu; Hanxiao Zhang; Jian Gao; Wen Tang; Pengxin Yu; Han Sung Kang; Junqiang Chen; Xing Lü; Boyu Zhang; Michail Mamalakis; Francesco Prinzi; Gianluca Carlini; Lisa Cuneo; Abhirup Banerjee; Zhaohu Xing; Lei Zhu; Zacharia Mesbah; Dhruv Jain; Tsiry Mayet; Hongyu Yuan; Qing Lyu; Athol U. Wells; Simon Walsh; Guang Yang
• Keywords: Honeycombing

Airway-related quantitative imaging biomarkers are crucial for examination, diagnosis, and prognosis in pulmonary diseases. However, the manual delineation of airway trees remains prohibitively time-consuming. While significant efforts have been made towards enhancing modelling, current public-available datasets concentrate on lung diseases with moderate morphological variations. The intricate honeycombing patterns present tissues fibrotic disease patients exacerbate challenges, often leading to various prediction errors. To address this issue, 'Airway-Informed Quantitative CT Imaging Biomarker Fibrotic Lung Disease 2023' (AIIB23) competition was organized conjunction official 2023 International Conference Medical Image Computing Computer Assisted Intervention (MICCAI). structures were meticulously annotated by three experienced radiologists. Competitors encouraged develop automatic segmentation models high robustness generalization abilities, followed exploring most correlated QIB mortality prediction. A training set 120 high-resolution computerised tomography (HRCT) scans publicly released expert annotations status. online validation incorporated 52 HRCT from offline test included 140 cases fibrosis COVID-19 patients. results shown that capacity extracting could be enhanced introducing voxel-wise weighted general union loss continuity loss. In addition competitive image prognosis, a strong airway-derived biomarker (Hazard ratio>1.5, p<0.0001) revealed survival prognostication compared existing clinical measurements, clinician assessment AI-based biomarkers.