Virtual-Eyes: Quantitative Validation of a Lung CT Quality-Control Pipeline for Foundation-Model Cancer Risk Prediction

Md. Enamul Hoq, Linda Larson-Prior, Fred Prior
Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, PMLR 315:4639-4663, 2026.

Abstract

Robust preprocessing is rarely quantified in deep-learning pipelines for low-dose CT (LDCT) lung cancer screening. We develop and validate Virtual-Eyes, a clinically motivated, 16-bit CT quality-control pipeline for NLST, and measure its differential impact on generalist foundation models versus specialist models. Virtual-Eyes enforces strict 512 $\times$ 512 resolution, rejects short or non-diagnostic series, and extracts a contiguous lung block using Hounsfield-unit filtering and bilateral lung-coverage scoring while preserving the original 16-bit DICOM grid. Using 765 NLST patients (182 cancer, 583 non-cancer), we evaluate RAD-DINO, Merlin, Sybil, and ResNet-18 under a leakage-free protocol. For RAD-DINO, preprocessing improves slice-level AUC from 0.576 to 0.610 and patient-level AUC from 0.646 to 0.683 (mean pooling) and 0.619 to 0.735 (max pooling), with improved calibration (Brier score 0.188 $\rightarrow$ 0.112). In contrast, Sybil and ResNet-18 degrade under Virtual-Eyes, revealing reliance on contextual or shortcut features, while Merlin shows limited transferability. Sensitivity analysis and uncertainty estimation confirm the robustness and stability of these findings.

Cite this Paper

BibTeX
@InProceedings{pmlr-v315-hoq26a, title = {Virtual-Eyes: Quantitative Validation of a Lung CT Quality-Control Pipeline for Foundation-Model Cancer Risk Prediction}, author = {Hoq, Md. Enamul and Larson-Prior, Linda and Prior, Fred}, booktitle = {Proceedings of The 9th International Conference on Medical Imaging with Deep Learning}, pages = {4639--4663}, year = {2026}, editor = {Huo, Yuankai and Gao, Mingchen and Kuo, Chang-Fu and Jin, Yueming and Deng, Ruining}, volume = {315}, series = {Proceedings of Machine Learning Research}, month = {08--10 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v315/main/assets/hoq26a/hoq26a.pdf}, url = {https://proceedings.mlr.press/v315/hoq26a.html}, abstract = {Robust preprocessing is rarely quantified in deep-learning pipelines for low-dose CT (LDCT) lung cancer screening. We develop and validate Virtual-Eyes, a clinically motivated, 16-bit CT quality-control pipeline for NLST, and measure its differential impact on generalist foundation models versus specialist models. Virtual-Eyes enforces strict 512 $\times$ 512 resolution, rejects short or non-diagnostic series, and extracts a contiguous lung block using Hounsfield-unit filtering and bilateral lung-coverage scoring while preserving the original 16-bit DICOM grid. Using 765 NLST patients (182 cancer, 583 non-cancer), we evaluate RAD-DINO, Merlin, Sybil, and ResNet-18 under a leakage-free protocol. For RAD-DINO, preprocessing improves slice-level AUC from 0.576 to 0.610 and patient-level AUC from 0.646 to 0.683 (mean pooling) and 0.619 to 0.735 (max pooling), with improved calibration (Brier score 0.188 $\rightarrow$ 0.112). In contrast, Sybil and ResNet-18 degrade under Virtual-Eyes, revealing reliance on contextual or shortcut features, while Merlin shows limited transferability. Sensitivity analysis and uncertainty estimation confirm the robustness and stability of these findings.} }
Endnote
%0 Conference Paper %T Virtual-Eyes: Quantitative Validation of a Lung CT Quality-Control Pipeline for Foundation-Model Cancer Risk Prediction %A Md. Enamul Hoq %A Linda Larson-Prior %A Fred Prior %B Proceedings of The 9th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Yuankai Huo %E Mingchen Gao %E Chang-Fu Kuo %E Yueming Jin %E Ruining Deng %F pmlr-v315-hoq26a %I PMLR %P 4639--4663 %U https://proceedings.mlr.press/v315/hoq26a.html %V 315 %X Robust preprocessing is rarely quantified in deep-learning pipelines for low-dose CT (LDCT) lung cancer screening. We develop and validate Virtual-Eyes, a clinically motivated, 16-bit CT quality-control pipeline for NLST, and measure its differential impact on generalist foundation models versus specialist models. Virtual-Eyes enforces strict 512 $\times$ 512 resolution, rejects short or non-diagnostic series, and extracts a contiguous lung block using Hounsfield-unit filtering and bilateral lung-coverage scoring while preserving the original 16-bit DICOM grid. Using 765 NLST patients (182 cancer, 583 non-cancer), we evaluate RAD-DINO, Merlin, Sybil, and ResNet-18 under a leakage-free protocol. For RAD-DINO, preprocessing improves slice-level AUC from 0.576 to 0.610 and patient-level AUC from 0.646 to 0.683 (mean pooling) and 0.619 to 0.735 (max pooling), with improved calibration (Brier score 0.188 $\rightarrow$ 0.112). In contrast, Sybil and ResNet-18 degrade under Virtual-Eyes, revealing reliance on contextual or shortcut features, while Merlin shows limited transferability. Sensitivity analysis and uncertainty estimation confirm the robustness and stability of these findings.
APA
Hoq, M.E., Larson-Prior, L. & Prior, F.. (2026). Virtual-Eyes: Quantitative Validation of a Lung CT Quality-Control Pipeline for Foundation-Model Cancer Risk Prediction. Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 315:4639-4663 Available from https://proceedings.mlr.press/v315/hoq26a.html.

Related Material