Lena Philipp
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Lena Philipp has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Alessa Hering
Radboudumc
4 shared publications - Fennie van der Graaf
From the Department of Medical Imaging, Radboud Institute of Medical Innovation, Radboud University Medical Center, Nijmegen, the Netherlands
3 shared publications - Colin Jacobs
From the Department of Medical Imaging, Radboud Institute of Medical Innovation, Radboud University Medical Center, Nijmegen, the Netherlands
2 shared publications - Ewoud J. Smit
Department of Radiology and Nuclear Medicine, Radboud University Medical Center, Nijmegen, The Netherlands
2 shared publications - Jan Tagscherer
Department of Medical Imaging, Radboudumc, Nijmegen, The Netherlands
2 shared publications - Sarah de Boer
Department of Medical Imaging, Radboudumc, Nijmegen, The Netherlands
2 shared publications
Latest Publications
- Open Access
- 27 August 2026
- Published by Springer Nature in International Journal of Computer Assisted Radiology and Surgery
- p. 1-12
Systematic evaluation of foundation models for organ-level classification on CT scans
Purpose: Foundation models are increasingly used as frozen feature extractors for CT classification tasks, yet the determinants of their downstream performance remain unclear. We assess how foundation model choice, feature aggregation strategy, and abnormality type are associated with performance in organ-level abnormality classification. We further evaluate whether more expressive aggregation strategies outperform simple pooling, and whether performance varies by abnormality type. Methods: We evaluate five state-of-the-art medical imaging foundation models on binary organ-level abnormality classification (normal vs. abnormal) across six abdominal organs. Local patch-level embeddings are aggregated using six strategies, including simple pooling methods (e.g., mean pooling) and attention-based multiple instance learning. A linear classifier is trained on feature embeddings, and performance is evaluated on an annotated test set of 200 CT scans. Results: Among the evaluated models, 3D CT-native models generally outperformed 2D multi-modal models. None of the tested aggregation strategies significantly outperformed mean pooling, including attention-based multiple instance learning (best-performing aggregation vs. mean: $$\Delta $$ AUC = 0.008, 95% CI $$[-0.002, 0.022]$$ ). In our organ-level classification setting, performance differed between abnormality type for all well-performing models (SPECTRE, TAP-CT, CT-FM), with lower AUCs observed for focal abnormalities compared to diffuse abnormalities (largest difference: $$\Delta $$ AUC = 0.108, 95% CI [0.080, 0.135]). This gap was not reduced by the evaluated aggregation strategies. Conclusion: In our experiments, downstream performance varied more across foundation models than across aggregation strategies. The choice of aggregation strategy, including attention-based multiple instance learning, did not significantly impact performance in this setting. The persistent gap for focal abnormalities suggests that current representations may insufficiently encode localized disease patterns, motivating the development of localization-aware pre-training approaches.- 27 August 2026
- Published by Springer Nature in International Journal of Computer Assisted Radiology and Surgery
- p. 1-12
Open AccessSystematic evaluation of foundation models for organ-level classification on CT scans
Purpose: Foundation models are increasingly used as frozen feature extractors for CT classification tasks, yet the determinants of their downstream performance remain unclear. We assess how foundation model choice, feature aggregation strategy, and abnormality type are associated with performance in organ-level abnormality classification. We further evaluate whether more expressive aggregation strategies outperform simple pooling, and whether performance varies by abnormality type. Methods: We evaluate five state-of-the-art medical imaging foundation models on binary organ-level abnormality classification (normal vs. abnormal) across six abdominal organs. Local patch-level embeddings are aggregated using six strategies, including simple pooling methods (e.g., mean pooling) and attention-based multiple instance learning. A linear classifier is trained on feature embeddings, and performance is evaluated on an annotated test set of 200 CT scans. Results: Among the evaluated models, 3D CT-native models generally outperformed 2D multi-modal models. None of the tested aggregation strategies significantly outperformed mean pooling, including attention-based multiple instance learning (best-performing aggregation vs. mean: $$\Delta $$ AUC = 0.008, 95% CI $$[-0.002, 0.022]$$ ). In our organ-level classification setting, performance differed between abnormality type for all well-performing models (SPECTRE, TAP-CT, CT-FM), with lower AUCs observed for focal abnormalities compared to diffuse abnormalities (largest difference: $$\Delta $$ AUC = 0.108, 95% CI [0.080, 0.135]). This gap was not reduced by the evaluated aggregation strategies. Conclusion: In our experiments, downstream performance varied more across foundation models than across aggregation strategies. The choice of aggregation strategy, including attention-based multiple instance learning, did not significantly impact performance in this setting. The persistent gap for focal abnormalities suggests that current representations may insufficiently encode localized disease patterns, motivating the development of localization-aware pre-training approaches. - Open Access
- 26 June 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Organ masks applied in feature space improve weakly supervised scan-level CT classification
Automated analysis of computed tomography (CT) scans is an active area of medical imaging research. Classification tasks often rely on scan-level labels without providing spatial information. When using frozen encoders from pretrained models, such as foundation models, enforcing anatomical focus by input-space cropping can shift the input distribution and degrade performance. We investigated whether organ masks can provide scalable anatomical guidance for weakly supervised classification using a frozen self-supervised Swin Transformer. Two strategies were evaluated: input-space organ centering and feature-space cropping, which applies organ masks to intermediate feature maps before pooling. Across three datasets and seven binary tasks, feature-space cropping matched or improved performance relative to full-image baselines, whereas input-space centering showed task-dependent effects. Feature-space cropping achieved a pooled improvement of 0.018 (95% confidence interval: 0.009–0.028) in area under the receiver operating characteristic curve, with the largest gains for liver lesion and pericardial effusion classification. Feature-space cropping reduced embedding dimensionality without loss of performance while preserving input distribution. In an ablation experiment of input-space centering, tighter crops reduced performance, highlighting the importance of peri-organ context. These findings demonstrate that feature-level anatomical guidance offers an efficient strategy to improve weakly supervised CT classification without retraining the encoder.
- 26 June 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessOrgan masks applied in feature space improve weakly supervised scan-level CT classification
Automated analysis of computed tomography (CT) scans is an active area of medical imaging research. Classification tasks often rely on scan-level labels without providing spatial information. When using frozen encoders from pretrained models, such as foundation models, enforcing anatomical focus by input-space cropping can shift the input distribution and degrade performance. We investigated whether organ masks can provide scalable anatomical guidance for weakly supervised classification using a frozen self-supervised Swin Transformer. Two strategies were evaluated: input-space organ centering and feature-space cropping, which applies organ masks to intermediate feature maps before pooling. Across three datasets and seven binary tasks, feature-space cropping matched or improved performance relative to full-image baselines, whereas input-space centering showed task-dependent effects. Feature-space cropping achieved a pooled improvement of 0.018 (95% confidence interval: 0.009–0.028) in area under the receiver operating characteristic curve, with the largest gains for liver lesion and pericardial effusion classification. Feature-space cropping reduced embedding dimensionality without loss of performance while preserving input distribution. In an ablation experiment of input-space centering, tighter crops reduced performance, highlighting the importance of peri-organ context. These findings demonstrate that feature-level anatomical guidance offers an efficient strategy to improve weakly supervised CT classification without retraining the encoder.
- 12 March 2026
- Published by Springer Nature
- p. 48-54
Modular Pipeline for Rapidly Evaluating Foundation Models in Medical Imaging
- 12 March 2026
- Published by Springer Nature
- p. 48-54
Modular Pipeline for Rapidly Evaluating Foundation Models in Medical Imaging
- Open Access
- 12 March 2026
- Published by Springer Nature
- p. 465
Abstract: Annotation-efficient 3D Body Composition Segmentation
- 12 March 2026
- Published by Springer Nature
- p. 465
Open AccessAbstract: Annotation-efficient 3D Body Composition Segmentation
- Open Access
- 21 August 2025
- Published by Machine Learning for Biomedical Imaging in Machine Learning for Biomedical Imaging
- Vol. 3 (Special on) , 559-593
Fairness Evaluation of Risk Estimation Models for Lung Cancer Screening
Lung cancer is the leading cause of cancer-related mortality in adults worldwide. Screening high-risk individuals with annual low-dose CT (LDCT) can support earlier detection and reduce deaths, but widespread implementation may strain the already limited radiology workforce. Artificial intelligence (AI) models have shown potential in estimating lung cancer risk from LDCT scans. However, high-risk populations for lung cancer are diverse, and these models’ performance across diverse demographic groups remains an open question. In this study, we used the JustEFAB ethical framework to evaluate potential performance disparities and fairness in two AI-based risk estimation models for lung cancer screening: the Sybil lung cancer risk model and the Venkadesh21 nodule risk estimator. We also examined disparities in the PanCan2b logistic regression model recommended in the British Thoracic Society nodule management guideline. Both AI-based models were trained on data from the U.S.-based National Lung Screening Trial (NLST), and assessed on a held-out NLST validation set. We evaluated area under the ROC curve (AUROC), sensitivity, and specificity across demographic subgroups, and explored potential confounding from clinical risk factors. We observed a statistically significant AUROC difference in Sybil’s performance between women (0.88, 95% CI: 0.86, 0.90) and men (0.81, 95% CI: 0.78, 0.84, p < .001). At 90% specificity, Venkadesh21 showed lower sensitivity for Black (0.39, 95% CI: 0.23, 0.59) than White participants (0.69, 95% CI: 0.65, 0.73). These differences were not explained by available clinical confounders and may be classified as unfair biases according to JustEFAB. Our findings highlight the importance of improving and monitoring model performance across underrepresented subgroups in lung cancer screening, as well as further research on algorithmic fairness in this field.
- 21 August 2025
- Published by Machine Learning for Biomedical Imaging in Machine Learning for Biomedical Imaging
- Vol. 3 (Special on) , 559-593
Open AccessFairness Evaluation of Risk Estimation Models for Lung Cancer Screening
Lung cancer is the leading cause of cancer-related mortality in adults worldwide. Screening high-risk individuals with annual low-dose CT (LDCT) can support earlier detection and reduce deaths, but widespread implementation may strain the already limited radiology workforce. Artificial intelligence (AI) models have shown potential in estimating lung cancer risk from LDCT scans. However, high-risk populations for lung cancer are diverse, and these models’ performance across diverse demographic groups remains an open question. In this study, we used the JustEFAB ethical framework to evaluate potential performance disparities and fairness in two AI-based risk estimation models for lung cancer screening: the Sybil lung cancer risk model and the Venkadesh21 nodule risk estimator. We also examined disparities in the PanCan2b logistic regression model recommended in the British Thoracic Society nodule management guideline. Both AI-based models were trained on data from the U.S.-based National Lung Screening Trial (NLST), and assessed on a held-out NLST validation set. We evaluated area under the ROC curve (AUROC), sensitivity, and specificity across demographic subgroups, and explored potential confounding from clinical risk factors. We observed a statistically significant AUROC difference in Sybil’s performance between women (0.88, 95% CI: 0.86, 0.90) and men (0.81, 95% CI: 0.78, 0.84, p < .001). At 90% specificity, Venkadesh21 showed lower sensitivity for Black (0.39, 95% CI: 0.23, 0.59) than White participants (0.69, 95% CI: 0.65, 0.73). These differences were not explained by available clinical confounders and may be classified as unfair biases according to JustEFAB. Our findings highlight the importance of improving and monitoring model performance across underrepresented subgroups in lung cancer screening, as well as further research on algorithmic fairness in this field.