Can the generalizability problem of artificial intelligence be overcome?: Pneumothorax detection algorithm

Verdi, E B; Kaya, A Gurun; Yilmaz, M; Turker, A; Ozbayoglu, A M; Isik, O; Karacin, A Bostanoglu; Yakut, O Velioglu; Yenigun, B M; Tasci, C; Uzun, C; Elhan, A H; Kaya, A

doi:10.1183/13993003.congress-2022.482

Cited by 1 publication

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Group 104: M-health/e-healthmentioning

confidence: 99%

“…Chest radiographs were used as the input to an AI algorithm to diagnose pulmonary embolism [ 82 ] and detect pneumothorax, airspace opacity, and mass or nodule [ 83 , 84 ]. V erdi et al [ 83 ] (Ankara, Turkey) used deep learning and multicentre datasets in a pneumothorax detection algorithm (PDA-alpha) to improve the model generalisability. Similarly, deep learning was used by G ana et al [ 84 ] (Harare, Zimbabwe) in another model that achieved a high performance, comparable to radiologists or other models approved for clinical use [ 85 , 86 ].…”

Section: Group 104: M-health/e-healthmentioning

confidence: 99%

See 1 more Smart Citation

ERS International Congress 2022: highlights from the Respiratory Clinical Care and Physiology Assembly

et al. 2023

View full text Add to dashboard Cite

It is a challenge to keep abreast of all the clinical and scientific advances in the field of respiratory medicine. This article contains an overview of laboratory-based science, clinical trials and qualitative research that were presented during the 2022 European Respiratory Society International Congress within the sessions from the five groups of the Assembly 1 – Respiratory clinical care and physiology. Selected presentations are summarised from a wide range of topics: clinical problems, rehabilitation and chronic care, general practice and primary care, electronic/mobile health (e-health/m-health), clinical respiratory physiology, exercise and functional imaging.

show abstract