Pulmonary abnormality screening on chest x-rays from different machine specifications : a generalized AI-based image manipulation pipeline
© 2023. The Author(s)..
BACKGROUND: Chest x-ray is commonly used for pulmonary abnormality screening. However, since the image characteristics of x-rays highly depend on the machine specifications, an artificial intelligence (AI) model developed for specific equipment usually fails when clinically applied to various machines. To overcome this problem, we propose an image manipulation pipeline.
METHODS: A total of 15,010 chest x-rays from systems with different generators/detectors were retrospectively collected from five institutions from May 2020 to February 2021. We developed an AI model to classify pulmonary abnormalities using x-rays from a single system. Then, we externally tested its performance on chest x-rays from various machine specifications. We compared the area under the receiver operating characteristics curve (AUC) of AI models developed using conventional image processing pipelines (histogram equalization [HE], contrast-limited histogram equalization [CLAHE], and unsharp masking [UM] with common data augmentations) with that of the proposed manipulation pipeline (XM-pipeline).
RESULTS: The XM-pipeline model showed the highest performance for all the datasets of different machine specifications, such as chest x-rays acquired from a computed radiography system (n = 356, AUC 0.944 for XM-pipeline versus 0.917 for HE, 0.705 for CLAHE, 0.544 for UM, p [Formula: see text] 0.001, for all) and from a mobile x-ray generator (n = 204, AUC 0.949 for XM-pipeline versus 0.933 for HE, p = 0.042, 0.932 for CLAHE (p = 0.009), 0.925 for UM (p = 0.001).
CONCLUSIONS: Applying the XM-pipeline to AI training increased the diagnostic performance of the AI model on the chest x-rays of different machine configurations.
RELEVANCE STATEMENT: The proposed training pipeline would successfully promote a wide application of the AI model for abnormality screening when chest x-rays are acquired using various x-ray machines.
KEY POINTS: • AI models developed using x-rays of a specific machine suffer from generalization. • We proposed a new image processing pipeline to address the generalization problem. • AI models were tested using multicenter external x-ray datasets of various machines. • AI with our pipeline achieved the highest diagnostic performance than conventional methods.
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2023 |
|---|---|
| Erschienen: |
2023 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:7 |
|---|---|
| Enthalten in: |
European radiology experimental - 7(2023), 1 vom: 09. Nov., Seite 68 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Shin, Heejun [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
Artificial intelligence |
|---|
| Anmerkungen: |
Date Completed 10.11.2023 Date Revised 15.11.2023 published: Electronic Citation Status MEDLINE |
|---|
| doi: |
10.1186/s41747-023-00386-1 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM364325348 |
|---|
| LEADER | 01000naa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM364325348 | ||
| 003 | DE-627 | ||
| 005 | 20231226095211.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 231226s2023 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1186/s41747-023-00386-1 |2 doi | |
| 028 | 5 | 2 | |a pubmed24n1214.xml |
| 035 | |a (DE-627)NLM364325348 | ||
| 035 | |a (NLM)37940797 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Shin, Heejun |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Pulmonary abnormality screening on chest x-rays from different machine specifications |b a generalized AI-based image manipulation pipeline |
| 264 | 1 | |c 2023 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Completed 10.11.2023 | ||
| 500 | |a Date Revised 15.11.2023 | ||
| 500 | |a published: Electronic | ||
| 500 | |a Citation Status MEDLINE | ||
| 520 | |a © 2023. The Author(s). | ||
| 520 | |a BACKGROUND: Chest x-ray is commonly used for pulmonary abnormality screening. However, since the image characteristics of x-rays highly depend on the machine specifications, an artificial intelligence (AI) model developed for specific equipment usually fails when clinically applied to various machines. To overcome this problem, we propose an image manipulation pipeline | ||
| 520 | |a METHODS: A total of 15,010 chest x-rays from systems with different generators/detectors were retrospectively collected from five institutions from May 2020 to February 2021. We developed an AI model to classify pulmonary abnormalities using x-rays from a single system. Then, we externally tested its performance on chest x-rays from various machine specifications. We compared the area under the receiver operating characteristics curve (AUC) of AI models developed using conventional image processing pipelines (histogram equalization [HE], contrast-limited histogram equalization [CLAHE], and unsharp masking [UM] with common data augmentations) with that of the proposed manipulation pipeline (XM-pipeline) | ||
| 520 | |a RESULTS: The XM-pipeline model showed the highest performance for all the datasets of different machine specifications, such as chest x-rays acquired from a computed radiography system (n = 356, AUC 0.944 for XM-pipeline versus 0.917 for HE, 0.705 for CLAHE, 0.544 for UM, p [Formula: see text] 0.001, for all) and from a mobile x-ray generator (n = 204, AUC 0.949 for XM-pipeline versus 0.933 for HE, p = 0.042, 0.932 for CLAHE (p = 0.009), 0.925 for UM (p = 0.001) | ||
| 520 | |a CONCLUSIONS: Applying the XM-pipeline to AI training increased the diagnostic performance of the AI model on the chest x-rays of different machine configurations | ||
| 520 | |a RELEVANCE STATEMENT: The proposed training pipeline would successfully promote a wide application of the AI model for abnormality screening when chest x-rays are acquired using various x-ray machines | ||
| 520 | |a KEY POINTS: • AI models developed using x-rays of a specific machine suffer from generalization. • We proposed a new image processing pipeline to address the generalization problem. • AI models were tested using multicenter external x-ray datasets of various machines. • AI with our pipeline achieved the highest diagnostic performance than conventional methods | ||
| 650 | 4 | |a Multicenter Study | |
| 650 | 4 | |a Journal Article | |
| 650 | 4 | |a Artificial intelligence | |
| 650 | 4 | |a Chest radiography | |
| 650 | 4 | |a Computer-aided classification | |
| 650 | 4 | |a Deep learning | |
| 650 | 4 | |a Image postprocessing (computer-assisted) | |
| 700 | 1 | |a Kim, Taehee |e verfasserin |4 aut | |
| 700 | 1 | |a Park, Juhyung |e verfasserin |4 aut | |
| 700 | 1 | |a Raj, Hruthvik |e verfasserin |4 aut | |
| 700 | 1 | |a Jabbar, Muhammad Shahid |e verfasserin |4 aut | |
| 700 | 1 | |a Abebaw, Zeleke Desalegn |e verfasserin |4 aut | |
| 700 | 1 | |a Lee, Jongho |e verfasserin |4 aut | |
| 700 | 1 | |a Van, Cong Cung |e verfasserin |4 aut | |
| 700 | 1 | |a Kim, Hyungjin |e verfasserin |4 aut | |
| 700 | 1 | |a Shin, Dongmyung |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t European radiology experimental |d 2017 |g 7(2023), 1 vom: 09. Nov., Seite 68 |w (DE-627)NLM283556358 |x 2509-9280 |7 nnns |
| 773 | 1 | 8 | |g volume:7 |g year:2023 |g number:1 |g day:09 |g month:11 |g pages:68 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1186/s41747-023-00386-1 |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 7 |j 2023 |e 1 |b 09 |c 11 |h 68 | ||