Contrastive domain adaptation with consistency match for automated pneumonia diagnosis

Feng, Yangqin; Wang, Zizhou; Xu, Xinxing; Wang, Yan; Fu, Huazhu; Li, Shaohua; Zhen, Liangli; Lei, Xiaofeng; Cui, Yingnan; Sim, Jordan Zheng Ting; Ting, Yonghan; Zhou, Joey Tianyi; Liu, Yong; Goh, Rick Siow Mong; Tan, Cher Heng

doi:10.1016/j.media.2022.102664

Cited by 18 publications

(8 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Domain shift is a common issue in practical applications of AI, especially in clinical practice. The clinical practice presents much more heterogeneous acquisition conditions [ 17 ] which may lead to the performance of the trained model on external test datasets (or a dataset it encounters when deployed) degrading.…”

Section: Discussionmentioning

confidence: 99%

“…In recent years, the application of deep learning models to clinical problems has shown significant potential in facilitating auto-diagnosis of diseases and providing real-time procedural support, particularly in the field of healthcare [ 17 , 18 , 19 ]. Several studies have explored the use of deep learning models for diagnosing COVID-19 through analysis of chest X-rays [ 20 , 21 ].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Neural Network Augments Performance of Junior Residents in Diagnosing COVID-19 Pneumonia on Chest Radiographs

Feng

Sim

et al. 2023

Diagnostics

Self Cite

View full text Add to dashboard Cite

Chest X-rays (CXRs) are essential in the preliminary radiographic assessment of patients affected by COVID-19. Junior residents, as the first point-of-contact in the diagnostic process, are expected to interpret these CXRs accurately. We aimed to assess the effectiveness of a deep neural network in distinguishing COVID-19 from other types of pneumonia, and to determine its potential contribution to improving the diagnostic precision of less experienced residents. A total of 5051 CXRs were utilized to develop and assess an artificial intelligence (AI) model capable of performing three-class classification, namely non-pneumonia, non-COVID-19 pneumonia, and COVID-19 pneumonia. Additionally, an external dataset comprising 500 distinct CXRs was examined by three junior residents with differing levels of training. The CXRs were evaluated both with and without AI assistance. The AI model demonstrated impressive performance, with an Area under the ROC Curve (AUC) of 0.9518 on the internal test set and 0.8594 on the external test set, which improves the AUC score of the current state-of-the-art algorithms by 1.25% and 4.26%, respectively. When assisted by the AI model, the performance of the junior residents improved in a manner that was inversely proportional to their level of training. Among the three junior residents, two showed significant improvement with the assistance of AI. This research highlights the novel development of an AI model for three-class CXR classification and its potential to augment junior residents’ diagnostic accuracy, with validation on external data to demonstrate real-world applicability. In practical use, the AI model effectively supported junior residents in interpreting CXRs, boosting their confidence in diagnosis. While the AI model improved junior residents’ performance, a decline in performance was observed on the external test compared to the internal test set. This suggests a domain shift between the patient dataset and the external dataset, highlighting the need for future research on test-time training domain adaptation to address this issue.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Deep Neural Network Augments Performance of Junior Residents in Diagnosing COVID-19 Pneumonia on Chest Radiographs

Feng

Sim

et al. 2023

Diagnostics

Self Cite

View full text Add to dashboard Cite

show abstract

“…In our analysis our aim was to compare different UDA methods on the same (well-established) ResNet-50 backbone. However, the selection of the backbone has an impact on the overall performance Feng et al [2023]. One further extension can be the comparison of different backbones and how it influences the performances.…”

Section: Discussionmentioning

confidence: 99%

“…Another reason for the usage of AUROC for evaluation is that various works on UDA methods compare their results either with AUROC or accuracy. In particular in the medical field it is common practice to compare methods based on AUROC scores Purushotham et al [2017], Zhou et al [2022], Feng et al [2023]. Most of the domain adaptation studies use accuracy as their metric for comparisons of methods, but none of these studies discuss the possible imbalance in their datasets.…”

Section: Performance Of Uda Methods On Non-dermoscopic Datasetsmentioning

confidence: 99%

Mitigating the Influence of Domain Shift in Skin Lesion Classification: A Benchmark Study of Unsupervised Domain Adaptation Methods on Dermoscopic Images

Chamarthi

Fogelberg

Maron

et al. 2023

Preprint

View full text Add to dashboard Cite

“…For instance, LLMs have been evaluated in radiology with impressive performance in both identifying important radiographic findings and augmenting the resident training experience. 17,18 Technical Overview of LLMs An LLM operates by training on large amounts of text data. This training phase exposes the model to information and patterns of linguistics.…”

Section: Ai In Medicinementioning

confidence: 99%

Academic Surgery in the Era of Large Language Models

Rengers,

Thiels,

Salehinejad

2024

JAMA Surg

View full text Add to dashboard Cite

ImportanceThis review aims to assess the benefits and risks of implementing large language model (LLM) solutions in an academic surgical setting.ObservationsThe integration of LLMs and artificial intelligence (AI) into surgical practice has generated international attention with the emergence of OpenAI’s ChatGPT and Google’s Bard. From an administrative standpoint, LLMs have the potential to revolutionize academic practices by reducing administrative burdens and improving efficiency. LLMs have the potential to facilitate surgical research by increasing writing efficiency, building predictive models, and aiding in large dataset analysis. From a clinical standpoint, LLMs can enhance efficiency by triaging patient concerns and generating automated responses. However, challenges exist, such as the need for improved LLM generalization performance, validating content, and addressing ethical concerns. In addition, patient privacy, potential bias in training, and legal responsibility are important considerations that require attention. Research and precautionary measures are necessary to ensure safe and unbiased use of LLMs in surgery.Conclusions and RelevanceAlthough limitations exist, LLMs hold promise for enhancing surgical efficiency while still prioritizing patient care. The authors recommend that the academic surgical community further investigate the potential applications of LLMs while being cautious about potential harms.

show abstract

Contrastive domain adaptation with consistency match for automated pneumonia diagnosis

Cited by 18 publications

References 37 publications

Deep Neural Network Augments Performance of Junior Residents in Diagnosing COVID-19 Pneumonia on Chest Radiographs

Deep Neural Network Augments Performance of Junior Residents in Diagnosing COVID-19 Pneumonia on Chest Radiographs

Mitigating the Influence of Domain Shift in Skin Lesion Classification: A Benchmark Study of Unsupervised Domain Adaptation Methods on Dermoscopic Images

Academic Surgery in the Era of Large Language Models

Contact Info

Product

Resources

About