cs.CVNov 15, 2026

Trading Parameters for Accuracy in MambaVision for Chest Radiograph Classification

Authors: Qaiser Farooq, Tehseen Zia

Organizations: Department of Computer Science, COMSATS University Islamabad, Islamabad, Pakistan

Abstract

MambaVision, a hybrid state-space and attention backbone, places most of its parameters in its widest, fourth stage. We quantified what that stage contributes to multi-label chest radiograph classification. The fourth stage was replaced with lighter single-scan state-space blocks adapted from Spatial-Mamba, with multi-scale dilated depthwise convolutions applied to the scan output; the first three stages were retained. Tiny, Small, and Base models were evaluated across nine dataset-scale combinations drawn from two source cohorts: the National Institutes of Health ChestX-ray14 archive (full ChestMNIST and a high-resolution subset) and Indiana University Open-i. At the Tiny scale, replacement reduced parameter count by 35.7% and increased graphics processing unit throughput by 19.3%; the savings came from the stage's dense projections, not from removing quadratic attention. Without distillation, the mean area under the receiver operating characteristic curve on ChestMNIST was 0.004 lower than the baseline (95% bootstrap confidence interval, 0.002 to 0.006), and 0.003 to 0.006 lower across all nine combinations. Knowledge distillation from an unreplaced teacher narrowed the gap to the undistilled baseline to 0.001-0.002 and met the equivalence criterion on ChestMNIST, but distilling the baseline with the same recipe improved it by 0.001-0.003, so a gap of 0.002-0.005 persisted under symmetric training. At the Base scale, the unreplaced Small model was both smaller and more accurate. Fourth-stage replacement is a useful trade-off only for Tiny and Small models. These measurements let developers of chest radiograph screening tools weigh model size and speed against a quantified loss of diagnostic discrimination.

Figures & tables

Explore similar work

Apr 14, 2026

A Hybrid Architecture for Benign-Malignant Classification of Mammography ROIs

Accurate characterization of suspicious breast lesions in mammography is important for early diagnosis and treatment planning. While Convolutional Neural Networks (CNNs) are effective at extracting local visual patterns, they are less suited to modeling long-range dependencies. Vision Transformers (ViTs) address this limitation through self-attention, but their quadratic computational cost can be prohibitive. This paper presents a hybrid architecture that combines EfficientNetV2-M for local feature extraction with Vision Mamba, a State Space Model (SSM), for efficient global context modeling. The proposed model performs binary classification of abnormality-centered mammography regions of interest (ROIs) from the CBIS-DDSM dataset into benign and malignant classes. By combining a strong CNN backbone with a linear-complexity sequence model, the approach achieves strong lesion-level classification performance in an ROI-based setting.
Nov 9, 2026

Is a CNN-Transformer Ensemble Worth It When a Chest X-Ray Foundation Model Is Available?

Ensembles of ImageNet-pretrained convolutional networks and Vision Transformers are a common way to improve multi-label chest X-ray classification. We ask whether such an ensemble is still worth its cost when a model pretrained on chest radiographs is available. On the official patient-disjoint split of NIH ChestX-ray14 (25,596 test images, 2,797 patients), we train five ImageNet-pretrained backbones and the EVA-X-S foundation model with validation-only model selection, compare five fusion rules, and estimate every difference with a patient-level paired bootstrap. EVA-X-S alone reached a test macro-AUROC of 0.828; the five-model ensemble reached 0.814 (difference -0.0136, 95% CI -0.0178 to -0.0098) at about 7.8 times the summed backbone latency and ten times the parameters. Averaging the five ImageNet models with EVA-X-S changed AUROC by -0.0056 (-0.0095 to -0.0021). With all five ImageNet backbones trained with the EVA-X-S recipe, their ensemble scored 0.817, still -0.0109 (-0.0142 to -0.0076) below EVA-X-S, and across three seeds EVA-X-S varied by at most 0.001 AUROC. Under the ImageNet models' recipe, however, EVA-X-S fell to 0.798. Our full-stacking configuration never outperformed calibrated averaging. In this setting, the tested ImageNet ensembles did not improve on a single chest X-ray foundation model fine-tuned with its own recipe.
Jun 16, 2026

Vision-language models for chest radiography do not always need the image

Vision-language models that answer questions about chest radiographs are evaluated by their accuracy on labels derived from radiology reports. High benchmark accuracy is often interpreted as evidence that the model uses the image. A model that answers from the finding named in the question can score as well as a model that uses the radiograph. Keeping the question fixed, we audit eight open-weight systems by swapping in another patient's radiograph with the same or the opposite label, occluding the radiologist-marked region or an equal region elsewhere, and removing the radiograph or replacing it with noise or a photograph. On 2,548 yes-or-no questions from MIMIC-CXR, one multimodal model answers Yes regardless of the image, another multimodal model changes its answers without following the label, and four systems use the image but keep about half of their correct answers when the radiograph is swapped for an opposite-label radiograph. A medical model that receives only the question text scores 55.3% on the pooled questions, higher than two multimodal systems. It scores 91.8% where every finding is present, and answering Yes to every question scores 100% there. Where the image is necessary, the best multimodal system exceeds this model by 10.4% in balanced accuracy. The categories are unchanged on CheXpert. Confidence is not higher when a correct answer depends on the marked region. In a reader study with three radiologists, the two radiologists who read a balanced set of 200 cases score 86.0% and 82.0%, and the systems score 50.0% to 73.0%. Accuracy does not establish image use, but an intervention on the image can test it.