An official website of the United States government
Purpose: Limited studies exploring concrete methods or approaches to tackle and enhance model fairness in the radiology domain. Our proposed AI model utilizes supervised contrastive learning to minimize bias in CXR diagnosis. Materials and Methods: In this retrospective study, we evaluated our proposed method on two datasets: the Medical Imaging and Data Resource Center (MIDRC) dataset with 77,887 CXR images from 27,796 patients collected as of April 20, 2023 for COVID-19 diagnosis, and the NIH Chest X-ray (NIH-CXR) dataset with 112,120 CXR images from 30,805 patients collected between 1992 and 2015. In the NIH-CXR dataset, thoracic abnormalities include atelectasis, cardiomegaly, effusion, infiltration, mass, nodule, pneumonia, pneumothorax, consolidation, edema, emphysema, fibrosis, pleural thickening, or hernia. Our proposed method utilizes supervised contrastive learning with carefully selected positive and negative samples to generate fair image embeddings, which are fine-tuned for subsequent tasks to reduce bias in chest X-ray (CXR) diagnosis. We evaluated the methods using the marginal AUC difference (δ mAUC). Results: The proposed model showed a significant decrease in bias across all subgroups when compared to the baseline models, as evidenced by a paired T-test (p<0.0001). The δ mAUC obtained by our method were 0.0116 (95\% CI, 0.0110-0.0123), 0.2102 (95% CI, 0.2087-0.2118), and 0.1000 (95\% CI, 0.0988-0.1011) for sex, race, and age on MIDRC, and 0.0090 (95\% CI, 0.0082-0.0097) for sex and 0.0512 (95% CI, 0.0512-0.0532) for age on NIH-CXR, respectively. Conclusion: Employing supervised contrastive learning can mitigate bias in CXR diagnosis, addressing concerns of fairness and reliability in deep learning-based diagnostic methods.
more »
« less
Revisiting the Trustworthiness of Saliency Methods in Radiology AI
Purpose: To determine if saliency maps in radiology artificial intelligence (AI) are vulnerable to subtle perturbations of the input, which could potentially lead to misleading interpretations, using Prediction-Saliency Correlation (PSC) for evaluating the sensitivity and robustness of saliency methods. Materials and Methods: In this retrospective study, locally trained deep learning models and a research prototype provided by a commercial vender were systematically evaluated on 191,229 chest radiographs from the CheXpert dataset(1,2) and 7,022 MRI images of human brain tumor classification dataset(3). Two radiologists performed a reader study on 270 chest radiographs pairs. A model-agnostic approach for computing the PSC coefficient was used to evaluate the sensitivity and robustness of seven commonly used saliency methods. Results: Leveraging locally trained model parameters, we revealed the saliency methods’ low sensitivity (maximum PSC = 0.25, 95% CI: 0.12, 0.38) and weak robustness (maximum PSC = 0.12, 95% CI: 0.0, 0.25) on the CheXpert dataset. Without model specifics, we also showed that the saliency maps from a commercial prototype could be irrelevant to the model output (area under the receiver operating characteristic curve dropped by 8.6% without affecting the saliency map). The human observer studies confirmed that is difficult for experts to identify the perturbed images, who had less than 44.8% correctness. Conclusion: Popular saliency methods scored low PSC values on the two datasets of perturbed chest radiographs, indicating weak sensitivity and robustness. The proposed PSC metric provides a valuable quantification tool for validating the trustworthiness of medical AI explainability. Abbreviations: AI = artificial intelligence, PSC = prediction-saliency correlation, AUC = area under the receiver operating characteristic curve, SSIM = structural similarity index measure. Summary: Systematic evaluation of saliency methods through subtle perturbations in chest radiographs and brain MRI images demonstrated low sensitivity and robustness of those methods, warranting caution when using saliency methods that may misrepresent changes in AI model prediction.
more »
« less
- Award ID(s):
- 2046708
- PAR ID:
- 10498673
- Publisher / Repository:
- Radiological Society of North America (RSNA)
- Date Published:
- Journal Name:
- Radiology: Artificial Intelligence
- Volume:
- 6
- Issue:
- 1
- ISSN:
- 2638-6100
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract During the coronavirus disease 2019 (COVID-19) pandemic, rapid and accurate triage of patients at the emergency department is critical to inform decision-making. We propose a data-driven approach for automatic prediction of deterioration risk using a deep neural network that learns from chest X-ray images and a gradient boosting model that learns from routine clinical variables. Our AI prognosis system, trained using data from 3661 patients, achieves an area under the receiver operating characteristic curve (AUC) of 0.786 (95% CI: 0.745–0.830) when predicting deterioration within 96 hours. The deep neural network extracts informative areas of chest X-ray images to assist clinicians in interpreting the predictions and performs comparably to two radiologists in a reader study. In order to verify performance in a real clinical setting, we silently deployed a preliminary version of the deep neural network at New York University Langone Health during the first wave of the pandemic, which produced accurate predictions in real-time. In summary, our findings demonstrate the potential of the proposed system for assisting front-line physicians in the triage of COVID-19 patients.more » « less
-
Abstract ObjectivesThe predictive intensive care unit (ICU) scoring system is crucial for predicting patient outcomes, particularly mortality. Traditional scoring systems rely mainly on structured clinical data from electronic health records, which can overlook important clinical information in narratives and images. Materials and MethodsIn this work, we build a deep learning-based survival prediction model that utilizes multimodality data for ICU mortality prediction. Four sets of features are investigated: (1) physiological measurements of Simplified Acute Physiology Score (SAPS) II, (2) common thorax diseases predefined by radiologists, (3) bidirectional encoder representations from transformers-based text representations, and (4) chest X-ray image features. The model was evaluated using the Medical Information Mart for Intensive Care IV dataset. ResultsOur model achieves an average C-index of 0.7829 (95% CI, 0.7620-0.8038), surpassing the baseline using only SAPS-II features, which had a C-index of 0.7470 (95% CI: 0.7263-0.7676). Ablation studies further demonstrate the contributions of incorporating predefined labels (2.00% improvement), text features (2.44% improvement), and image features (2.82% improvement). Discussion and ConclusionThe deep learning model demonstrated superior performance to traditional machine learning methods under the same feature fusion setting for ICU mortality prediction. This study highlights the potential of integrating multimodal data into deep learning models to enhance the accuracy of ICU mortality prediction.more » « less
-
BackgroundRisk-based screening for lung cancer is currently being considered in several countries; however, the optimal approach to determine eligibility remains unclear. Ensemble machine learning could support the development of highly parsimonious prediction models that maintain the performance of more complex models while maximising simplicity and generalisability, supporting the widespread adoption of personalised screening. In this work, we aimed to develop and validate ensemble machine learning models to determine eligibility for risk-based lung cancer screening. Methods and findingsFor model development, we used data from 216,714 ever-smokers recruited between 2006 and 2010 to the UK Biobank prospective cohort and 26,616 high-risk ever-smokers recruited between 2002 and 2004 to the control arm of the US National Lung Screening (NLST) randomised controlled trial. The NLST trial randomised high-risk smokers from 33 US centres with at least a 30 pack-year smoking history and fewer than 15 quit-years to annual CT or chest radiography screening for lung cancer. We externally validated our models among 49,593 participants in the chest radiography arm and all 80,659 ever-smoking participants in the US Prostate, Lung, Colorectal and Ovarian (PLCO) Screening Trial. The PLCO trial, recruiting from 1993 to 2001, analysed the impact of chest radiography or no chest radiography for lung cancer screening. We primarily validated in the PLCO chest radiography arm such that we could benchmark against comparator models developed within the PLCO control arm. Models were developed to predict the risk of 2 outcomes within 5 years from baseline: diagnosis of lung cancer and death from lung cancer. We assessed model discrimination (area under the receiver operating curve, AUC), calibration (calibration curves and expected/observed ratio), overall performance (Brier scores), and net benefit with decision curve analysis.Models predicting lung cancer death (UCL-D) and incidence (UCL-I) using 3 variables—age, smoking duration, and pack-years—achieved or exceeded parity in discrimination, overall performance, and net benefit with comparators currently in use, despite requiring only one-quarter of the predictors. In external validation in the PLCO trial, UCL-D had an AUC of 0.803 (95% CI: 0.783, 0.824) and was well calibrated with an expected/observed (E/O) ratio of 1.05 (95% CI: 0.95, 1.19). UCL-I had an AUC of 0.787 (95% CI: 0.771, 0.802), an E/O ratio of 1.0 (95% CI: 0.92, 1.07). The sensitivity of UCL-D was 85.5% and UCL-I was 83.9%, at 5-year risk thresholds of 0.68% and 1.17%, respectively, 7.9% and 6.2% higher than the USPSTF-2021 criteria at the same specificity. The main limitation of this study is that the models have not been validated outside of UK and US cohorts. ConclusionsWe present parsimonious ensemble machine learning models to predict the risk of lung cancer in ever-smokers, demonstrating a novel approach that could simplify the implementation of risk-based lung cancer screening in multiple settings.more » « less
-
null (Ed.)Abstract Background This study outlines an image processing algorithm for accurate and consistent lung segmentation in chest radiographs of critically ill adults and children typically obscured by medical equipment. In particular, this work focuses on applications in analysis of acute respiratory distress syndrome – a critical illness with a mortality rate of 40% that affects 200,000 patients in the United States and 3 million globally each year. Methods Chest radiographs were obtained from critically ill adults (n = 100), adults diagnosed with acute respiratory distress syndrome (ARDS) (n = 25), and children (n = 100) hospitalized at Michigan Medicine. Physicians annotated the lung field of each radiograph to establish the ground truth. A Total Variation-based Active Contour (TVAC) lung segmentation algorithm was developed and compared to multiple state-of-the-art methods including a deep learning model (U-Net), a random walker algorithm, and an active spline model, using the Sørensen–Dice coefficient to measure segmentation accuracy. Results The TVAC algorithm accurately segmented lung fields in all patients in the study. For the adult cohort, an averaged Dice coefficient of 0.86 ±0.04 (min: 0.76) was reported for TVAC, 0.89 ±0.12 (min: 0.01) for U-Net, 0.74 ±0.19 (min: 0.15) for the random walker algorithm, and 0.64 ±0.17 (min: 0.20) for the active spline model. For the pediatric cohort, a Dice coefficient of 0.85 ±0.04 (min: 0.75) was reported for TVAC, 0.87 ±0.09 (min: 0.56) for U-Net, 0.67 ±0.18 (min: 0.18) for the random walker algorithm, and 0.61 ±0.18 (min: 0.18) for the active spline model. Conclusion The proposed algorithm demonstrates the most consistent performance of all segmentation methods tested. These results suggest that TVAC can accurately identify lung fields in chest radiographs in critically ill adults and children.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1148/ryai.220221
