Matthew S. Brown
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Matthew S. Brown has a total of 164 co-authors. Here are the top co-authors by number of shared publications:
- Jonathan Gerald Goldin
The University of California, Los Angeles (UCLA), David Geffen School of Medicine at UCLA, Center for Computer Vision and Imaging Biomarkers, Department of Radiological Sciences, Los Angeles, California, United States
20 shared publications - Michael Francis McNitt-Gray
Univ. of California, Los Angeles (United States)
13 shared publications - Grace Hyun J. Kim
University of California Los Angeles
10 shared publications Denise R. Aberle
Center for Computer Vision and Imaging Biomarkers, Department of Radiological Sciences, David Geffen School of Medicine at UCLA, 924 Westwood Blvd., Suite 615, Los Angeles, CA, 90024, USA
6 shared publicationsHyun J. Kim
Center for Computer Vision and Imaging Biomarkers, University of California, Los Angeles, Los Angeles, CA
6 shared publications- Fereidoun Abtin
David Geffen School of Medicine at UCLA
5 shared publications
Latest Publications
- Open Access
- 3 November 2026
- Published by Oxford University Press (OUP) in Radiology Advances
- Vol. 3 (6)
Adaptive deep learning model with anatomical priors for 3D segmentation of fine hepatic vasculature on CT
Background: Accurate segmentation of fine hepatic vessels on CT remains challenging because peripheral branches are small and low contrast. Purpose: To develop and test PriM Fusion Net, a deep learning model that adaptively integrates an optimized anatomical prior to improve 3D CT segmentation of portal and hepatic veins, emphasizing fine peripheral branches for preoperative 3D modeling. Materials and methods: This retrospective study used the public Medical Segmentation Decathlon (MSD) Task08-HepaticVessel dataset of 303 abdominal CT volumes. PriM Fusion Net uses a dual-encoder architecture: a Swin Transformer processes CT images and a parallel convolutional pathway processes a dataset-optimized Frangi vesselness prior. Gated multimodal units act as spatially varying voxel-wise gates that fuse the 2 feature streams at multiple scales. A rule-based SimpleMind post-processing module refined anatomical plausibility. Performance was evaluated with 5-fold cross-validation and zero-shot external inference on 3DIRCADb (3D Image Reconstruction for Comparison of Algorithm Database). Results: PriM Fusion Net achieved a mean Dice similarity coefficient (DSC) of 77.68% (95% confidence interval [CI]: 77.11, 78.25) and 95th percentile Hausdorff distance (HD95) of 11.55 mm (95% CI, 9.94, 13.16), outperforming all baselines (all adjusted P < .02). For vessels with radius ≤ 1.5 mm, DSC improved by 13.5% versus SwinUNETR (59.2% [95% CI, 54.48, 63.92] vs 45.7% [95% CI, 40.49, 50.91]). In ablation, replacing rigid concatenation with gated fusion improved DSC from 75.86% (95% CI, 75.31, 76.41) to 77.65% (95% CI, 76.92, 78.38), and SimpleMind post-processing reduced HD95 from 14.98 mm (95% CI, 10.91, 19.05) to 11.55 mm (95% CI, 9.94, 13.16). On 3DIRCADb, zero-shot inference yielded DSC of 68.42% (95% CI, 65.82, 71.02) and HD95 of 11.78 mm (95% CI, 8.50, 15.06). Conclusion: Adaptive deep fusion of an optimized anatomical prior improved CT segmentation of portal and hepatic veins, especially fine peripheral branches, supporting further evaluation in 3D vascular modeling workflows for liver surgery and intervention.
- 3 November 2026
- Published by Oxford University Press (OUP) in Radiology Advances
- Vol. 3 (6)
Open AccessAdaptive deep learning model with anatomical priors for 3D segmentation of fine hepatic vasculature on CT
Background: Accurate segmentation of fine hepatic vessels on CT remains challenging because peripheral branches are small and low contrast. Purpose: To develop and test PriM Fusion Net, a deep learning model that adaptively integrates an optimized anatomical prior to improve 3D CT segmentation of portal and hepatic veins, emphasizing fine peripheral branches for preoperative 3D modeling. Materials and methods: This retrospective study used the public Medical Segmentation Decathlon (MSD) Task08-HepaticVessel dataset of 303 abdominal CT volumes. PriM Fusion Net uses a dual-encoder architecture: a Swin Transformer processes CT images and a parallel convolutional pathway processes a dataset-optimized Frangi vesselness prior. Gated multimodal units act as spatially varying voxel-wise gates that fuse the 2 feature streams at multiple scales. A rule-based SimpleMind post-processing module refined anatomical plausibility. Performance was evaluated with 5-fold cross-validation and zero-shot external inference on 3DIRCADb (3D Image Reconstruction for Comparison of Algorithm Database). Results: PriM Fusion Net achieved a mean Dice similarity coefficient (DSC) of 77.68% (95% confidence interval [CI]: 77.11, 78.25) and 95th percentile Hausdorff distance (HD95) of 11.55 mm (95% CI, 9.94, 13.16), outperforming all baselines (all adjusted P < .02). For vessels with radius ≤ 1.5 mm, DSC improved by 13.5% versus SwinUNETR (59.2% [95% CI, 54.48, 63.92] vs 45.7% [95% CI, 40.49, 50.91]). In ablation, replacing rigid concatenation with gated fusion improved DSC from 75.86% (95% CI, 75.31, 76.41) to 77.65% (95% CI, 76.92, 78.38), and SimpleMind post-processing reduced HD95 from 14.98 mm (95% CI, 10.91, 19.05) to 11.55 mm (95% CI, 9.94, 13.16). On 3DIRCADb, zero-shot inference yielded DSC of 68.42% (95% CI, 65.82, 71.02) and HD95 of 11.78 mm (95% CI, 8.50, 15.06). Conclusion: Adaptive deep fusion of an optimized anatomical prior improved CT segmentation of portal and hepatic veins, especially fine peripheral branches, supporting further evaluation in 3D vascular modeling workflows for liver surgery and intervention.
- 1 January 2026
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 13 (01) , 1-15
Characterizing the effects of noncontrast head CT reconstruction kernel and slice thickness parameters on the performance of an automated AI algorithm in the evaluation of ischemic stroke
PurposeThere are multiple commercially available, Food and Drug Administration (FDA)-cleared, artificial intelligence (AI)-based tools automating stroke evaluation in noncontrast computed tomography (NCCT). This study assessed the impact of variations in reconstruction kernel and slice thickness on two outputs of such a system: hypodense volume and Alberta Stroke Program Early CT Score (ASPECTS).ApproachThe NCCT series image data of 67 patients imaged with a CT stroke protocol were reconstructed with four kernels (H10s-smooth, H40s-medium, H60s-sharp, and H70h-very sharp) and three slice thicknesses (1.5, 3.0, and 5.0 mm) to create 1 reference condition (H40s/5.0 mm) and 11 nonreference conditions. The 12 reconstructions per patient were processed with a commercially available FDA-cleared software package that yields total hypodense volume (mL) and ASPECTS. A mixed-effect model was used to test the difference in hypodense volume, and an ordered logistic model was used to test the difference in e-ASPECTS.ResultsHypodense volume differences from the reference condition ranged from −14.6 to 1.1 mL and were significant for all nonreference kernels (H10s p=0.025, H60s p<0.001, and H70h p<0.001) and for thinner slices (1.5 mm p0.05).ConclusionsAutomated hypodense volume measured with a commercially available, FDA-cleared software package is substantially impacted by reconstruction kernel and slice thickness. Conversely, automated ASPECTS is invariant to these reconstruction parameters.
- 1 January 2026
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 13 (01) , 1-15
Characterizing the effects of noncontrast head CT reconstruction kernel and slice thickness parameters on the performance of an automated AI algorithm in the evaluation of ischemic stroke
PurposeThere are multiple commercially available, Food and Drug Administration (FDA)-cleared, artificial intelligence (AI)-based tools automating stroke evaluation in noncontrast computed tomography (NCCT). This study assessed the impact of variations in reconstruction kernel and slice thickness on two outputs of such a system: hypodense volume and Alberta Stroke Program Early CT Score (ASPECTS).ApproachThe NCCT series image data of 67 patients imaged with a CT stroke protocol were reconstructed with four kernels (H10s-smooth, H40s-medium, H60s-sharp, and H70h-very sharp) and three slice thicknesses (1.5, 3.0, and 5.0 mm) to create 1 reference condition (H40s/5.0 mm) and 11 nonreference conditions. The 12 reconstructions per patient were processed with a commercially available FDA-cleared software package that yields total hypodense volume (mL) and ASPECTS. A mixed-effect model was used to test the difference in hypodense volume, and an ordered logistic model was used to test the difference in e-ASPECTS.ResultsHypodense volume differences from the reference condition ranged from −14.6 to 1.1 mL and were significant for all nonreference kernels (H10s p=0.025, H60s p<0.001, and H70h p<0.001) and for thinner slices (1.5 mm p0.05).ConclusionsAutomated hypodense volume measured with a commercially available, FDA-cleared software package is substantially impacted by reconstruction kernel and slice thickness. Conversely, automated ASPECTS is invariant to these reconstruction parameters.
- Open Access
TRACE: Textual Reasoning for Affordance Coordinate Extraction
Vision-Language Models (VLMs) struggle to translate high-level instructions into the precise spatial affordances required for robotic manipulation. While visual Chain-of-Thought (CoT) methods exist, they are often computationally intensive. In this work, we introduce TRACE (Textual Reasoning for Affordance Coordinate Extraction), a novel methodology that integrates a textual Chain of Reasoning (CoR) into the affordance prediction process. We use this methodology to create the TRACE dataset, a large-scale collection created via an autonomous pipeline that pairs instructions with explicit textual rationales. By fine-tuning a VLM on this data, our model learns to externalize its spatial reasoning before acting. Our experiments show that our TRACE-tuned model achieves state-of-the-art performance, reaching 48.1% accuracy on the primary Where2Place (W2P) benchmark (a 9.6% relative improvement) and 55.0% on the more challenging W2P(h) subset. Crucially, an ablation study demonstrates that performance scales directly with the amount of reasoning data used, confirming the CoR's effectiveness. Furthermore, analysis of the model's attention maps reveals an interpretable reasoning process where focus shifts dynamically across reasoning steps. This work shows that training VLMs to generate a textual CoR is an effective and robust strategy for enhancing the precision, reliability, and interpretability of VLM-based robot control. Our dataset and code are available at https://github.com/jink-ucla/TRACE
Open AccessTRACE: Textual Reasoning for Affordance Coordinate Extraction
Vision-Language Models (VLMs) struggle to translate high-level instructions into the precise spatial affordances required for robotic manipulation. While visual Chain-of-Thought (CoT) methods exist, they are often computationally intensive. In this work, we introduce TRACE (Textual Reasoning for Affordance Coordinate Extraction), a novel methodology that integrates a textual Chain of Reasoning (CoR) into the affordance prediction process. We use this methodology to create the TRACE dataset, a large-scale collection created via an autonomous pipeline that pairs instructions with explicit textual rationales. By fine-tuning a VLM on this data, our model learns to externalize its spatial reasoning before acting. Our experiments show that our TRACE-tuned model achieves state-of-the-art performance, reaching 48.1% accuracy on the primary Where2Place (W2P) benchmark (a 9.6% relative improvement) and 55.0% on the more challenging W2P(h) subset. Crucially, an ablation study demonstrates that performance scales directly with the amount of reasoning data used, confirming the CoR's effectiveness. Furthermore, analysis of the model's attention maps reveals an interpretable reasoning process where focus shifts dynamically across reasoning steps. This work shows that training VLMs to generate a textual CoR is an effective and robust strategy for enhancing the precision, reliability, and interpretability of VLM-based robot control. Our dataset and code are available at https://github.com/jink-ucla/TRACE
- 22 September 2025
- Published by Springer Nature
- p. 310-319
DuEL-Med: A Dual-Path Enhanced Language Model for Clinically-Aware Radiology Report Generation
- 22 September 2025
- Published by Springer Nature
- p. 310-319
DuEL-Med: A Dual-Path Enhanced Language Model for Clinically-Aware Radiology Report Generation
- 1 March 2025
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 12 (02) , 1-14
Using a fully automated, quantitative fissure integrity score extracted from chest CT scans of emphysema patients to predict endobronchial valve response
PurposeWe aim to develop and validate a prediction model using a previously developed fully automated quantitative fissure integrity score (FIS) extracted from pre-treatment CT images to identify suitable candidates for endobronchial valve (EBV) treatment.ApproachWe retrospectively collected 96 anonymized pre- and post-treatment chest computed tomography (CT) exams from patients with moderate to severe emphysema and who underwent EBV treatment. We used a previously developed fully automated, deep learning–based approach to quantitatively assess the completeness of each fissure by obtaining the FIS for each fissure from each patient’s pre-treatment CT exam. The response to EBV treatment was recorded as the amount of targeted lobe volume reduction (TLVR) compared with target lobe volume prior to treatment as assessed on the pre- and post-treatment CT scans. EBV placement was considered successful with a TLVR of ≥350 cc. The dataset was split into a training set (N=58) and a test set (N=38) to train and validate a logistic regression model using fivefold cross-validation; the extracted FIS of each patient’s targeted treatment lobe was the primary CT predictor. Using the training set, a receiver operating characteristic (ROC) curve analysis and predictive values were quantified over a range of FIS thresholds to determine an optimal cutoff value that would distinguish complete and incomplete fissures, which was used to evaluate predictive values of the test set cases.ResultsROC analysis of the training set provided an AUC of 0.83, and the determined FIS threshold was 89.5%. Using this threshold on the test set achieved an accuracy of 81.6%, specificity (Sp) of 90.9%, sensitivity (Sn) of 77.8%, positive predictive value (PPV) of 62.5%, and negative predictive value of 95.5%.ConclusionsA model using the quantified FIS shows potential as a predictive biomarker for whether a targeted lobe will achieve successful volume reduction from EBV treatment.
- 1 March 2025
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 12 (02) , 1-14
Using a fully automated, quantitative fissure integrity score extracted from chest CT scans of emphysema patients to predict endobronchial valve response
PurposeWe aim to develop and validate a prediction model using a previously developed fully automated quantitative fissure integrity score (FIS) extracted from pre-treatment CT images to identify suitable candidates for endobronchial valve (EBV) treatment.ApproachWe retrospectively collected 96 anonymized pre- and post-treatment chest computed tomography (CT) exams from patients with moderate to severe emphysema and who underwent EBV treatment. We used a previously developed fully automated, deep learning–based approach to quantitatively assess the completeness of each fissure by obtaining the FIS for each fissure from each patient’s pre-treatment CT exam. The response to EBV treatment was recorded as the amount of targeted lobe volume reduction (TLVR) compared with target lobe volume prior to treatment as assessed on the pre- and post-treatment CT scans. EBV placement was considered successful with a TLVR of ≥350 cc. The dataset was split into a training set (N=58) and a test set (N=38) to train and validate a logistic regression model using fivefold cross-validation; the extracted FIS of each patient’s targeted treatment lobe was the primary CT predictor. Using the training set, a receiver operating characteristic (ROC) curve analysis and predictive values were quantified over a range of FIS thresholds to determine an optimal cutoff value that would distinguish complete and incomplete fissures, which was used to evaluate predictive values of the test set cases.ResultsROC analysis of the training set provided an AUC of 0.83, and the determined FIS threshold was 89.5%. Using this threshold on the test set achieved an accuracy of 81.6%, specificity (Sp) of 90.9%, sensitivity (Sn) of 77.8%, positive predictive value (PPV) of 62.5%, and negative predictive value of 95.5%.ConclusionsA model using the quantified FIS shows potential as a predictive biomarker for whether a targeted lobe will achieve successful volume reduction from EBV treatment.
- Open Access
Anatomically Guided Deep Learning System for Right Internal Jugular Line (RIJL) Segmentation and Tip Localization in Chest X-Ray
The right internal jugular line (RIJL) is a type of central venous catheter (CVC) inserted into the right internal jugular vein to deliver medications and monitor vital functions in ICU patients. The placement of RIJL is routinely checked by a clinician in a chest X-ray (CXR) image to ensure its proper function and patient safety. To reduce the workload of clinicians, deep learning-based automated detection algorithms have been developed to detect CVCs in CXRs. Although RIJL is the most widely used type of CVCs, there is a paucity of investigations focused on its accurate segmentation and tip localization. In this study, we propose a deep learning system that integrates an anatomical landmark segmentation, an RIJL segmentation network, and a postprocessing function to segment the RIJL course and detect the tip with accuracy and precision. We utilized the nnU-Net framework to configure the segmentation network. The entire system was implemented on the SimpleMind Cognitive AI platform, enabling the integration of anatomical knowledge and spatial reasoning to model relationships between objects within the image. Specifically, the trachea was used as an anatomical landmark to extract a subregion in a CXR image that is most relevant to the RIJL. The subregions were used to generate cropped images, which were used to train the segmentation network. The segmentation results were recovered to original dimensions, and the most inferior point’s coordinates in each image were defined as the tip. With guidance from the anatomical landmark and customized postprocessing, the proposed method achieved improved segmentation and tip localization compared to the baseline segmentation network: the mean average symmetric surface distance (ASSD) was decreased from 2.72 to 1.41 mm, and the mean tip distance was reduced from 11.27 to 8.29 mm.
Open AccessAnatomically Guided Deep Learning System for Right Internal Jugular Line (RIJL) Segmentation and Tip Localization in Chest X-Ray
The right internal jugular line (RIJL) is a type of central venous catheter (CVC) inserted into the right internal jugular vein to deliver medications and monitor vital functions in ICU patients. The placement of RIJL is routinely checked by a clinician in a chest X-ray (CXR) image to ensure its proper function and patient safety. To reduce the workload of clinicians, deep learning-based automated detection algorithms have been developed to detect CVCs in CXRs. Although RIJL is the most widely used type of CVCs, there is a paucity of investigations focused on its accurate segmentation and tip localization. In this study, we propose a deep learning system that integrates an anatomical landmark segmentation, an RIJL segmentation network, and a postprocessing function to segment the RIJL course and detect the tip with accuracy and precision. We utilized the nnU-Net framework to configure the segmentation network. The entire system was implemented on the SimpleMind Cognitive AI platform, enabling the integration of anatomical knowledge and spatial reasoning to model relationships between objects within the image. Specifically, the trachea was used as an anatomical landmark to extract a subregion in a CXR image that is most relevant to the RIJL. The subregions were used to generate cropped images, which were used to train the segmentation network. The segmentation results were recovered to original dimensions, and the most inferior point’s coordinates in each image were defined as the tip. With guidance from the anatomical landmark and customized postprocessing, the proposed method achieved improved segmentation and tip localization compared to the baseline segmentation network: the mean average symmetric surface distance (ASSD) was decreased from 2.72 to 1.41 mm, and the mean tip distance was reduced from 11.27 to 8.29 mm.
- Open Access
- 18 October 2024
- Published by MDPI AG in Medical Sciences
- Vol. 12 (4) , 57
Predicting Invasiveness in Lepidic Pattern Adenocarcinoma of Lung: Analysis of Visual Semantic and Radiomic Features
Objectives: To differentiate invasive lepidic predominant adenocarcinoma (iLPA) from adenocarcinoma in situ (AIS)/minimally invasive adenocarcinoma (MIA) of lung utilizing visual semantic and computer-aided detection (CAD)-based texture features on subjects initially diagnosed as AIS or MIA with CT-guided biopsy. Materials and Methods: From 2011 to 2017, all patients with CT-guided biopsy results of AIS or MIA who subsequently underwent resection were identified. CT scan before the biopsy was used to assess visual semantic and CAD texture features, totaling 23 semantic and 95 CAD-based quantitative texture variables. The least absolute shrinkage and selection operator (LASSO) method or forward selection was used to select the most predictive feature and combination of semantic and texture features for detection of invasive lung adenocarcinoma. Results: Among the 33 core needle-biopsied patients with AIS/MIA pathology, 24 (72.7%) had invasive LPA and 9 (27.3%) had AIS/MIA on resection. On CT, visual semantic features included 21 (63.6%) part-solid, 5 (15.2%) pure ground glass, and 7 (21.2%) solid nodules. LASSO selected seven variables for the model, but all were not statistically significant. “Volume” was found to be statistically significant when assessing the correlation between independent variables using the backward selection technique. The LASSO selected “tumor_Perc95”, “nodule surround”, “small cyst-like spaces”, and “volume” when assessing the correlation between independent variables. Conclusions: Lung biopsy results showing noninvasive LPA underestimate invasiveness. Although statistically non-significant, some semantic features showed potential for predicting invasiveness, with septal stretching absent in all noninvasive cases, and solid consistency present in a significant portion of invasive cases.
- 18 October 2024
- Published by MDPI AG in Medical Sciences
- Vol. 12 (4) , 57
Open AccessPredicting Invasiveness in Lepidic Pattern Adenocarcinoma of Lung: Analysis of Visual Semantic and Radiomic Features
Objectives: To differentiate invasive lepidic predominant adenocarcinoma (iLPA) from adenocarcinoma in situ (AIS)/minimally invasive adenocarcinoma (MIA) of lung utilizing visual semantic and computer-aided detection (CAD)-based texture features on subjects initially diagnosed as AIS or MIA with CT-guided biopsy. Materials and Methods: From 2011 to 2017, all patients with CT-guided biopsy results of AIS or MIA who subsequently underwent resection were identified. CT scan before the biopsy was used to assess visual semantic and CAD texture features, totaling 23 semantic and 95 CAD-based quantitative texture variables. The least absolute shrinkage and selection operator (LASSO) method or forward selection was used to select the most predictive feature and combination of semantic and texture features for detection of invasive lung adenocarcinoma. Results: Among the 33 core needle-biopsied patients with AIS/MIA pathology, 24 (72.7%) had invasive LPA and 9 (27.3%) had AIS/MIA on resection. On CT, visual semantic features included 21 (63.6%) part-solid, 5 (15.2%) pure ground glass, and 7 (21.2%) solid nodules. LASSO selected seven variables for the model, but all were not statistically significant. “Volume” was found to be statistically significant when assessing the correlation between independent variables using the backward selection technique. The LASSO selected “tumor_Perc95”, “nodule surround”, “small cyst-like spaces”, and “volume” when assessing the correlation between independent variables. Conclusions: Lung biopsy results showing noninvasive LPA underestimate invasiveness. Although statistically non-significant, some semantic features showed potential for predicting invasiveness, with septal stretching absent in all noninvasive cases, and solid consistency present in a significant portion of invasive cases.
- 1 May 2024
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 11 (03) , 1-19
Quantifying lung fissure integrity using a three-dimensional patch-based convolutional neural network on CT images for emphysema treatment planning
PurposeEvaluation of lung fissure integrity is required to determine whether emphysema patients have complete fissures and are candidates for endobronchial valve (EBV) therapy. We propose a deep learning (DL) approach to segment fissures using a three-dimensional patch-based convolutional neural network (CNN) and quantitatively assess fissure integrity on CT to evaluate it in subjects with severe emphysema.ApproachFrom an anonymized image database of patients with severe emphysema, 129 CT scans were used. Lung lobe segmentations were performed to identify lobar regions, and the boundaries among these regions were used to construct approximate interlobar regions of interest (ROIs). The interlobar ROIs were annotated by expert image analysts to identify voxels where the fissure was present and create a reference ROI that excluded non-fissure voxels (where the fissure is incomplete). A CNN configured by nnU-Net was trained using 86 CT scans and their corresponding reference ROIs to segment the ROIs of left oblique fissure (LOF), right oblique fissure (ROF), and right horizontal fissure (RHF). For an independent test set of 43 cases, fissure integrity was quantified by mapping the segmented fissure ROI along the interlobar ROI. A fissure integrity score (FIS) was then calculated as the percentage of labeled fissure voxels divided by total voxels in the interlobar ROI. Predicted FIS (p-FIS) was quantified from the CNN output, and statistical analyses were performed comparing p-FIS and reference FIS (r-FIS).ResultsThe absolute percent error mean (±SD) between r-FIS and p-FIS for the test set was 4.0% (±4.1%), 6.0% (±9.3%), and 12.2% (±12.5%) for the LOF, ROF, and RHF, respectively.ConclusionsA DL approach was developed to segment lung fissures on CT images and accurately quantify FIS. It has potential to assist in the identification of emphysema patients who would benefit from EBV treatment.
- 1 May 2024
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 11 (03) , 1-19
Quantifying lung fissure integrity using a three-dimensional patch-based convolutional neural network on CT images for emphysema treatment planning
PurposeEvaluation of lung fissure integrity is required to determine whether emphysema patients have complete fissures and are candidates for endobronchial valve (EBV) therapy. We propose a deep learning (DL) approach to segment fissures using a three-dimensional patch-based convolutional neural network (CNN) and quantitatively assess fissure integrity on CT to evaluate it in subjects with severe emphysema.ApproachFrom an anonymized image database of patients with severe emphysema, 129 CT scans were used. Lung lobe segmentations were performed to identify lobar regions, and the boundaries among these regions were used to construct approximate interlobar regions of interest (ROIs). The interlobar ROIs were annotated by expert image analysts to identify voxels where the fissure was present and create a reference ROI that excluded non-fissure voxels (where the fissure is incomplete). A CNN configured by nnU-Net was trained using 86 CT scans and their corresponding reference ROIs to segment the ROIs of left oblique fissure (LOF), right oblique fissure (ROF), and right horizontal fissure (RHF). For an independent test set of 43 cases, fissure integrity was quantified by mapping the segmented fissure ROI along the interlobar ROI. A fissure integrity score (FIS) was then calculated as the percentage of labeled fissure voxels divided by total voxels in the interlobar ROI. Predicted FIS (p-FIS) was quantified from the CNN output, and statistical analyses were performed comparing p-FIS and reference FIS (r-FIS).ResultsThe absolute percent error mean (±SD) between r-FIS and p-FIS for the test set was 4.0% (±4.1%), 6.0% (±9.3%), and 12.2% (±12.5%) for the LOF, ROF, and RHF, respectively.ConclusionsA DL approach was developed to segment lung fissures on CT images and accurately quantify FIS. It has potential to assist in the identification of emphysema patients who would benefit from EBV treatment.
- Open Access
- 6 January 2024
- Published by MDPI AG in Biomedicines
- Vol. 12 (1) , 120
Quantitative Computed Tomography Lung COVID Scores with Laboratory Markers: Utilization to Predict Rapid Progression and Monitor Longitudinal Changes in Patients with Coronavirus 2019 (COVID-19) Pneumonia
Coronavirus disease 2019 (COVID-19), is an ongoing issue in certain populations, presenting rapidly worsening pneumonia and persistent symptoms. This study aimed to test the predictability of rapid progression using radiographic scores and laboratory markers and present longitudinal changes. This retrospective study included 218 COVID-19 pneumonia patients admitted at the Chungnam National University Hospital. Rapid progression was defined as respiratory failure requiring mechanical ventilation within one week of hospitalization. Quantitative COVID (QCOVID) scores were derived from high-resolution computed tomography (CT) analyses: (1) ground glass opacity (QGGO), (2) mixed diseases (QMD), and (3) consolidation (QCON), and the sum, quantitative total lung diseases (QTLD). Laboratory data, including inflammatory markers, were obtained from electronic medical records. Rapid progression was observed in 9.6% of patients. All QCOVID scores predicted rapid progression, with QMD showing the best predictability (AUC = 0.813). In multivariate analyses, the QMD score and interleukin(IL)-6 level were important predictors for rapid progression (AUC = 0.864). With >2 months follow-up CT, remained lung lesions were observed in 21 subjects, even after several weeks of negative reverse transcription polymerase chain reaction test. AI-driven quantitative CT scores in conjugation with laboratory markers can be useful in predicting the rapid progression and monitoring of COVID-19.
- 6 January 2024
- Published by MDPI AG in Biomedicines
- Vol. 12 (1) , 120
Open AccessQuantitative Computed Tomography Lung COVID Scores with Laboratory Markers: Utilization to Predict Rapid Progression and Monitor Longitudinal Changes in Patients with Coronavirus 2019 (COVID-19) Pneumonia
Coronavirus disease 2019 (COVID-19), is an ongoing issue in certain populations, presenting rapidly worsening pneumonia and persistent symptoms. This study aimed to test the predictability of rapid progression using radiographic scores and laboratory markers and present longitudinal changes. This retrospective study included 218 COVID-19 pneumonia patients admitted at the Chungnam National University Hospital. Rapid progression was defined as respiratory failure requiring mechanical ventilation within one week of hospitalization. Quantitative COVID (QCOVID) scores were derived from high-resolution computed tomography (CT) analyses: (1) ground glass opacity (QGGO), (2) mixed diseases (QMD), and (3) consolidation (QCON), and the sum, quantitative total lung diseases (QTLD). Laboratory data, including inflammatory markers, were obtained from electronic medical records. Rapid progression was observed in 9.6% of patients. All QCOVID scores predicted rapid progression, with QMD showing the best predictability (AUC = 0.813). In multivariate analyses, the QMD score and interleukin(IL)-6 level were important predictors for rapid progression (AUC = 0.864). With >2 months follow-up CT, remained lung lesions were observed in 21 subjects, even after several weeks of negative reverse transcription polymerase chain reaction test. AI-driven quantitative CT scores in conjugation with laboratory markers can be useful in predicting the rapid progression and monitoring of COVID-19.
- 1 May 2023
- Published by Radiological Society of North America (RSNA) in RadioGraphics
- Vol. 43 (5) , e220105
Translating AI to Clinical Practice: Overcoming Data Shift with Explainability
To translate artificial intelligence (AI) algorithms into clinical practice requires generalizability of models to real-world data. One of the main obstacles to generalizability is data shift, a data distribution mismatch between model training and real environments. Explainable AI techniques offer tools to detect and mitigate the data shift problem and develop reliable AI for clinical practice. Most medical AI is trained with datasets gathered from limited environments, such as restricted disease populations and center-dependent acquisition conditions. The data shift that commonly exists in the limited training set often causes a significant performance decrease in the deployment environment. To develop a medical application, it is important to detect potential data shift and its impact on clinical translation. During AI training stages, from premodel analysis to in-model and post hoc explanations, explainability can play a key role in detecting model susceptibility to data shift, which is otherwise hidden because the test data have the same biased distribution as the training data. Performance-based model assessments cannot effectively distinguish the model overfitting to training data bias without enriched test sets from external environments. In the absence of such external data, explainability techniques can aid in translating AI to clinical practice as a tool to detect and mitigate potential failures due to data shift. ©RSNA, 2023 Quiz questions for this article are available in the supplemental material.
- 1 May 2023
- Published by Radiological Society of North America (RSNA) in RadioGraphics
- Vol. 43 (5) , e220105
Translating AI to Clinical Practice: Overcoming Data Shift with Explainability
To translate artificial intelligence (AI) algorithms into clinical practice requires generalizability of models to real-world data. One of the main obstacles to generalizability is data shift, a data distribution mismatch between model training and real environments. Explainable AI techniques offer tools to detect and mitigate the data shift problem and develop reliable AI for clinical practice. Most medical AI is trained with datasets gathered from limited environments, such as restricted disease populations and center-dependent acquisition conditions. The data shift that commonly exists in the limited training set often causes a significant performance decrease in the deployment environment. To develop a medical application, it is important to detect potential data shift and its impact on clinical translation. During AI training stages, from premodel analysis to in-model and post hoc explanations, explainability can play a key role in detecting model susceptibility to data shift, which is otherwise hidden because the test data have the same biased distribution as the training data. Performance-based model assessments cannot effectively distinguish the model overfitting to training data bias without enriched test sets from external environments. In the absence of such external data, explainability techniques can aid in translating AI to clinical practice as a tool to detect and mitigate potential failures due to data shift. ©RSNA, 2023 Quiz questions for this article are available in the supplemental material.