Accurate uncertainty estimation is essential for machine learning systems de- ployed in high-stakes domains such as medicine. Traditional approaches primarily rely on probability outputs from trained models (point predictions), which provide no formal guarantees on prediction coverage and often require additional calibra- tion techniques to improve reliability. In contrast, conformal prediction (region prediction) offers a principled alternative by generating prediction sets with finite- sample validity guarantees, ensuring that the ground truth is contained within the set at a specified confidence level. In this study, we explore the impact of pre-training approach, dataset scale and domain on both point and region-level uncertainty quantification, by studying domain-specific vision medical foundation models vs. general domain vision foundation models. We conduct a comprehensive evaluation across foundation models trained on retinal, histopathological, and Chest X-Rays data, applying various calibration techniques. Our results demonstrate that (1) pre-training on higher-quality domain-specific datasets along with self-supervised learning leads to better-calibrated point predictions than general domain pre-training, (2) stan- dard re-calibration methods alone cannot fully mitigate uncertainty discrepancies across models trained on different data sources, (3) domain-specific foundation model can lead to more efficient conformal prediction. These findings highlight the importance of careful model selection and the inte- gration of both point and region prediction to enhance the reliability and trust- worthiness of medical AI systems. Our work underscores the need for a holistic approach to uncertainty quantification in recent development of medical vision foundation model, ensuring robust and interpretable AI-driven decision-making.
Multi-organ ultrasound classifiers increasingly combine attention, mixture-of-experts routing, uncertainty gating, and evidential deep learning (EDL) objectives to address heterogeneous anatomy and acquisition. Yet a plausible design rationale does not by itself establish that an added component improves the trained system. We contribute a controlled complexity-audit framework, applied to the deployment decision between the maximal evidential candidate Full-EDL and simpler alternatives. Six candidates were evaluated on the primary dataset and three in an internal replication, using ten matched seeds, frozen image-level partitions, capacity- and optimisation-aware comparisons, symmetric temperature scaling, paired decision rules, and a separate out-of-distribution (OOD) veto. Retaining Full-EDL did not establish a reliable macro-F1 gain on either dataset, while the simplified alternatives remained inconclusive under the non-inferiority margin. Simple cross-entropy with temperature scaling (Simple-CE+TS) met the calibrated negative log-likelihood criterion on both datasets and showed favourable selective-risk ordering. The raw calibration advantage of evidential training disappeared after temperature scaling and did not recur on the second dataset. The gate had negligible observable influence at the audited checkpoints, and deleting the Full-only chain revealed no stable task or calibrated-loss benefit. Simple-CE nevertheless triggered the OOD veto against the fetal probe but not the lung probe, precluding an unconditional OOD-safety claim. We therefore selected Simple-CE+TS for the evaluated in-distribution objective while retaining Full-EDL as the maximal reference. Components should earn retention through functional and retraining-based evidence, and calibration and distribution-shift reliability should be evaluated separately.
Reliable evaluation of vision-language models (VLMs) and medical vision-language models (Medical-VLMs) requires calibrated confidence, particularly under realistic clinical conditions. However, existing efforts mainly focused on improving accuracy, leaving calibration in the medical domain underexplored. To this end, we propose MVC-Bench, a calibration-centric benchmark for medical image classification with VLMs and Medical-VLMs. MVC-Bench assesses the calibration across three axes: (i) robustness to modality, backbone, and domain shift (ii) effectiveness of calibration strategies and prompt-tuning methods (iii) stability under prompt-template and random-seed variations. The benchmark covers eight different backbones, three medical modalities, including fundus imaging, histopathology, and chest X-ray under in-domain and domain shift settings. It compares post-hoc calibration, train-time calibration, and zero-shot inference methods, together with six prompt-tuning methods. Across more than 1638 controlled experiments, we report accuracy and Expected Calibration Error (ECE) as primary metrics, and further report results with complementary calibration measures, including Maximum Calibration Error (MCE) and Adaptive Calibration Error (ACE). We further investigate the underlying causes of miscalibration in VLMs and Medical-VLMs and propose a simple train-time calibration method, Multi-Class Margin (MCM) regularization, which achieves lowest ECE on 10 out of 12 settings in in-domain and remains competitive under domain shifts. Collectively, MVC-Bench provides a structured evaluation framework and actionable guidance for improving calibration in safety-critical medical workflows.
Probability calibration aligns model confidence with predictive accuracy, enabling clinicians to identify unreliable segmentation regions. This alignment breaks down under domain shift, where artifacts and unseen protocols produce confident errors. Existing post-hoc methods adapt the correction at test time, conditioning on predictive entropy, the logit pattern, or augmentation response, but each proxy is read from the terminal prediction, the very quantity that shift corrupts. This motivates reliability evidence beyond the terminal prediction, which categorical diffusion provides in two ways. First, a generative shape prior keeps a capacity-limited reference intact when appearance is corrupted, so its disagreement with the primary segmentor highlights primary-model errors. Second, every reverse step yields a class distribution, separating persistent disagreement from transient discrepancy. Aggregated over the trajectory, this disagreement correlates with Dice at 0.788, against 0.521 for a matched discriminative control. We therefore propose CARD (Calibration via Agreement in Reverse Diffusion), which maps the temporal aggregate of this disagreement to a temperature field applied per pixel across all classes, so that confidence changes while the segmentation does not. Across cardiac, prostate and brain MRI shifts, CARD lowers calibration error in 45 of 49 comparisons against the strongest baseline in each setting.
Jai Kumar Sharma, Peeyush Tapadiyacs.CV cs.AI q-bio.QM
Frozen hematology foundation-model (FM) embeddings reach near-saturated in-domain white-blood-cell (WBC) accuracy, but clinical deployment demands reliability across scanners, sites, stains and preparation pipelines. We audit 15 frozen encoders (hematology, pathology, and general vision) across four public single-cell acquisition domains along two axes: accuracy robustness and calibration. In-domain linear-probe macro-F1 is saturated (0.98-0.997), yet cross-dataset macro-F1 drops 34-72% and rankings re-order: DinoBloom-L, the in-domain best, falls to 10th of 15 on the most-shifted target (MLL23) at the benchmark's shared 224-px input, behind RedDino and several general and pathology encoders. Rank transfer is probe-dependent: 1-NN retrieval is more stable on average than a source-fitted linear head (median $ρ$ 0.65 vs 0.45), but neither probe universally predicts target robustness. Calibration also collapses: source-trained probes are nearly calibrated in-domain (expected calibration error, ECE, 0.004) but confidently wrong off-domain (ECE 0.35), and source-fitted temperature scaling transfers poorly. We further audit pretraining exposure and identify MLL23 as DinoBloom's internal cohort; because DinoBloom's only held-out dataset is also our source domain, this benchmark cannot isolate exposure from scanner-associated shift. Label-free adaptation and marginal-entropy-based model selection appear safe under balanced evaluation but fail under realistic WBC class-prior shift. Class-Balanced Re-standardization (CBR), a training-free pseudo-label-balanced feature normalization, improves all evaluated target-prior scenario means and partially improves calibration, although encoder-level exceptions and residual miscalibration remain. Hematology FM benchmarks must therefore jointly audit accuracy, calibration, exposure, and class-prior robustness.
Mohaimenul Azam Khan Raiaan, Nur Mohammad Fahadcs.CV
Medical vision-language models (VLMs) can achieve high accuracy but remain unreliable: they are systematically overconfident, benefit little from test-time reasoning, and lack the ability to reliably calibrate trust in their own responses. We introduce EVADE (Evidence-Verified Agentic Diagnosis with Escape), an inferential, non-training method that enhances the safety of deploying a single frozen VLM. EVADE responds and, when uncertain, localises the region most diagnostically relevant, re-answers on a zoomed view, and commits only when both the entire image and the zoomed view responses agree; otherwise, it abstains. To directly address verification hallucination in single-model self-checking, our main idea is to verify gate consistency across different image views rather than re-reading the model's own text. Experimental evaluation on VQA-RAD, SLAKE, and PathVQA using Qwen2.5-VL-7B reports that EVADE is the only method that simultaneously improves both calibration and selective risk while maintaining accuracy, reducing expected calibration error (ECE) by up to 45% compared to zero-shot. Chain-of-thought, self-consistency, and self-verification all fail at least one axis. A grounding analysis reports that self-proposed regions perform better at diagnostic structure localisation than centres or random crops. However, a 7B VLM cannot use this localisation to revise answers. Therefore, reliability gains come from the consistency gate and calibrated abstention.
Leon Koole, Jiapan Guo, Matias Valdenegro-Torocs.CV cs.LG
Skin lesion classifiers can be confidently wrong on the cases that matter most, so knowing when a prediction should not be trusted is clinically as useful as the prediction. We study uncertainty quantification on a dataset pooled from many ISIC sources, with a shared backbone and two jointly learned heads: a binary malignant versus non-malignant head and a five-class diagnostic head. Five UQ methods (MC Dropout, DropConnect, Flipout, Deep Ensembles, DUQ) are compared on accuracy, calibration, uncertainty decomposition, and risk-coverage. Difficulty is largely method-agnostic: even methods with narrow entropy distributions rank the same samples as hard (per-sample entropy correlations of $0.54$ to $0.91$). The choice of method matters more for calibration and uncertainty decomposition, where Deep Ensembles is the clear winner, than for finding difficult cases. The ranking is also good enough that deferring the most uncertain cases removes a disproportionate share of errors, supporting uncertainty-based selective referral, evaluated here in-distribution only.
Deep networks segment brain tumours accurately in-distribution, but can fail silently when the input differs from their training data. That risk is central to clinical deployment and is the premise of the BraTS-GoAT generalizability task. We ask not only how well a model segments, but whether its uncertainty knows when it is wrong. On BraTS-GoAT (Task 3) we train a 5-fold cross-validated nnU-Net baseline (one held-out prediction per case) and a 3-seed deep ensemble. Both are evaluated for calibration and error detection on a per-region relevant mask, aggregated per case. In-distribution the 3-seed ensemble improves modestly over the already strong single model on the same held-out split, with the clearest gain in calibration. The separation appears under shift. In a controlled robustness study using graded synthetic corruptions as a proxy for acquisition shift, the single model's confidence stays flat while its accuracy and calibration degrade. Inter-member disagreement instead rises steeply, about a quarter to a third above the clean condition, several times the single model's response. On the official validation leaderboard the 5-fold ensemble of those folds attains whole-tumour Dice 0.87. The generalization gap is concentrated on the harder regions, with a characteristic failure of missing small, satellite lesions on unseen cohorts. In the synthetic study, disagreement among the 3-seed members is a more sensitive case-level indicator of acquisition shift than single-model confidence. Its per-voxel error localisation weakens as severity grows. The contribution is a rigorous, honest reliability comparison rather than a claim that any one uncertainty method dominates.
Pulmonary nodule malignancy prediction typically depends on image-trained specialist deep learning (DL) models that require substantial annotated imaging data and task-specific training. We investigate whether a generalist large language model (LLM), reading only a faithful natural-language rendering of standard nodule attributes, can serve as a calibrated triage layer. We propose ConfTriage, a confidence-calibrated method built on three pillars: language as the modality, calibration as the safety mechanism, and a selective specialist DL backstop for low-confidence cases. We prove two guarantees: a finite-sample combined-error bound yielding an explicit per-threshold operational certificate, and an oracle inequality showing that excess risk over the Bayes-optimal deferral classifier is controlled by the L1 calibration error of the LLM probability. A controlled seven-way input ablation across five frontier LLMs on LIDC-IDRI shows that natural-language descriptions dominate the diagnostic signal, while low-level image statistics are essentially diagnostically vacuous. ConfTriage achieved an F1 score of 88.22% and an AUC of 0.92, resolving 76.5% of cases using zero-shot LLM inference alone and referring only uncertain cases to the specialist DL backstop. These results demonstrate that clinically meaningful diagnostic information can be captured through structured radiological descriptions and leveraged by calibrated LLMs for selective referral. The framework suggests a practical pathway for combining generalist LLM prediction with specialist AI models in medical decision-support systems. Source code is publicly available at https://github.com/rabiul-ai/ConfTriage.
Dinh Tan Nguyen, Hoang Quan Dang, Chen Zhang +1cs.CV
Breast lesion detection in mammography remains a challenging task due to variations in image quality, lesion appearance, and population demographics across datasets. While current object detectors such as YOLO and DETR achieve strong results on individual datasets, their performance often degrades when trained on or applied across heterogeneous sources. To address this, we propose MammoMix, a novel framework based on Mixture-of-Experts (MoE) paradigm for robust and generalizable lesion detection. In MammoMix, each expert model is trained on a specific domain, allowing it to specialize in distinct characteristics of its source data. A gating mechanism adaptively weighs contributions from each expert based on input image, combining their outputs to enable domain-adaptive inference. To improve reliability, we further incorporate a calibration module, MoCAE, which adjusts confidence scores to reflect true predictive uncertainty. We evaluate MammoMix on 3 public mammography datasets: CSAW, DDSM, and DMID, covering diverse clinical settings. Results show that MammoMix outperforms baseline detectors in both average precision and reliability, particularly on datasets with greater variability. Our findings demonstrate that expert specialization and calibrated ensemble fusion significantly enhance model generalization and robustness. MammoMix offers a promising step toward dependable AI-assisted breast cancer screening across real-world clinical domains.
When decoder language models are used as classifiers, predicted class probabilities depend on implementation choices, including the prompt template, verbalizer (label-to-token mapping), and scoring rule, that are rarely treated as experimental variables. We present a controlled evaluation of three Mistral-7B variants (Base, BioMistral, and Instruct) on PubMed RCT sentence classification (n=2000) under FP16, INT8, and INT4 precision using four answer-text prompt templates. Our primary finding is that the probability extraction protocol dominates apparent calibration. Switching from summed to mean token log-likelihood scoring reverses the calibration ranking between models: BioMistral average expected calibration error increases from 0.097 to 0.289, whereas Instruct decreases from 0.237 to 0.096, while accuracy changes by less than 1 percentage point for the specialized models but 4-6 percentage points for the base model. Prompt template choice produces accuracy differences of 7-24 percentage points, comparable to or larger than model-level effects. On one template, BioMistral outperforms Instruct although the overall mean favors Instruct by only 1.3 percentage points. For BioMistral and Instruct, INT8 quantization changes accuracy and F1 by only 1-2 percentage points relative to FP16, whereas the base model shows larger INT8 effects on some templates (up to +4.2 percentage points). INT4 produces heterogeneous but non-catastrophic effects. Temperature scaling reduces expected calibration error under summed scoring for both models but only for that scoring rule. A fine-tuned PubMedBERT reference achieves 82.7% accuracy but uses about 176000 labeled training examples, precluding direct comparison. These results demonstrate that prompt template design and scoring normalization are first-order experimental decisions when evaluating decoder language model calibration.
Cardiac digital twins convert clinical images into physiological measurements through observation operators, yet calibration studies often assume a fixed reference convention. Across four shared-backbone echocardiographic EF front-ends, phase conditioning appears to remove CAMUS baseline bias. Matched-reference analysis rejects this gain: singleplane ground-truth EF error is statistically indistinguishable across models, while single-plane ground-truth EF exceeds CAMUS biplane clinical EF by +6.30 points, explaining nearly all baseline bias. A prespecified EchoNet-Dynamic replication, with released data and our extractor aligned to the apical four-chamber plane, removes baseline overestimation and reverses the CAMUS ranking. We also quantify haemodynamic effects, conformal residual-width budgets, and EF-stratum changes, yielding a Convention-Aware EF Audit protocol that separates genuine observation operator calibration from measurement artefacts. GitHub: EjectionFraction-Bias-in-Cardiac-Digital-Twin.git
Decoders of anesthetic state from cortical activity fail across drug classes, most notoriously ketamine, but reported accuracy cannot say whether the neural representation or only the decision threshold has failed; we separate the two in a controlled preparation with ground-truth labels. We decoded awake versus anesthetized from mouse electrocorticography (the 250-Hz-bandlimited local field potential, sampled at 1875 Hz) under leave-one-anesthetic-out evaluation across five mechanistically distinct anesthetics (isoflurane, dexmedetomidine, ketamine, propofol, and midazolam), comparing a spatially blind band-power decoder, covariance/Riemannian representations, and Riemannian domain adaptation, with all statistics at the session level and mouse-level cluster bootstrapping for the ketamine fold. The representation transfers: band-power ranks awake versus anesthetized at a session AUROC of at least 0.96 on every held-out drug, ketamine included (0.980, cluster confidence interval 0.821 to 1.000). The failure is confined to the threshold: across three representations the ketamine ranking is near-invariant while its balanced accuracy swings from chance to high, and a permutation test is significant for ranking (p = 0.0025) but not for fixed-threshold accuracy (p = 0.3795). Riemannian domain adaptation is net-negative. A causal, label-free threshold anchored to the subject's own pre-induction baseline fixes ketamine (balanced accuracy 0.50 to 0.85) and dominates domain adaptation. Because the ketamine test sessions come from three mice that also contribute training drugs, this is within-subject cross-drug transfer; we do not claim population-level transfer across subjects. In cross-drug state decoding the actionable failure is calibration, not representation.
Behraj Khan, Shabir Ahmad, Syed Ahmad Chan Bukhari +1cs.LG
Medical world models aim to learn a latent state of patient or organ physiology and a transition function that forecasts how that state evolves under interventions, supporting downstream tasks from imaging-based diagnosis to digital-twin treatment planning. Two failure modes threaten the reliability of such models in clinical deployment: (i)~\emph{covariate shift}, because training data are fragmented across hospitals, scanners, and time, so the feature distribution seen by the latent-dynamics predictor differs across fragments and from the distribution at deployment; and (ii)~\emph{confidence misalignment}, because multi-step forecasts are often overconfident exactly where clinical risk is highest. We argue that both problems admit a unified treatment via a single lightweight regularisation objective, \textbf{CalTwin}, which combines a Fisher-Information-based shift penalty adapted from our prior work on fragmented covariate-shift remediation~\cite{khan2025mitigating,khan2025causal} with a Confidence Misalignment Penalty adapted from our prior work on calibrated vision-language classification~\cite{khan2025confidence}, applied here to a GRU-based medical world model's latent transition predictor. We derive the combined objective, establish which proof steps transfer from the classification setting without modification and which require adaptation, and evaluate it on the PhysioNet 2019 Sepsis Challenge, treating the two hospital systems as sequential training fragments and the unseen system as an out-of-distribution test. CalTwin reduces OOD next-step latent-state MSE by 9.1\% relative to the no-penalty baseline (FIM penalty alone accounts for 7.0\%); the ECE reduction from the Confidence Misalignment Penalty is real but small (0.7\% for CalTwin, 1.3\% for CMP alone).
Recent work has argued normatively, on synthetic data, that evaluating survival models by discrimination alone (concordance index) yields systematically misleading model comparisons, because the metric ignores calibration and time-dependent accuracy. Whether this matters for real, published, non-clinical models has not been tested. We reproduce three published survival-ML models across three structurally distinct domains -- hard-drive failure prediction, peer-to-peer credit default, and user disengagement on digital platforms -- validate our instrument against the anchor paper's own synthetic experiment, and test five pre-registered hypotheses under a Holm-corrected family-wise error rate. Three of five reject (though one pre-registered threshold clears by a narrow margin). A model reproducing the published literature's discrimination almost exactly (C = 0.9595 vs. 0.958 reported) fails a formal calibration test at p < 0.001; a broad feature-ablation search finds no single attribute responsible for its discrimination, so the calibration failure is not a trivial shortcut artifact. A lender's estimated default risk is biased upward by roughly two percentage points, growing to nearly four in the riskiest segment, when loan prepayment is treated as non-informative censoring rather than a competing risk. A platform's churn model shows probability estimates that degrade with the horizon even as global discrimination stays within the pre-registered C-index band. A direct test of whether metric choice inverts model preference does not reject, though with limited power given two to three models per domain; the failure mode we document is better characterized as misplaced confidence in a chosen model than as choosing the wrong one. We release a pre-registered evaluation harness with full code and an annotated notebook, so these results can be verified independently and the audit extended.
S M Asif Hossain, Ruksat Khan Shayoni, M. F. Mridha +1cs.LG
Cardiometabolic diseases remain among the most persistent drivers of preventable morbidity because diabetes, hypertension, and cardiovascular disease frequently co-occur and share metabolic, vascular, demographic, and behavioral determinants. Existing machine learning studies for chronic disease prediction often emphasize discrimination on a single dataset, while underreporting label leakage, calibration, temporal robustness, external transportability, and subgroup reliability. This paper presents CardioMeta, a calibrated multi-task framework for joint prediction of diabetes, hypertension, and cardiovascular disease across population survey and electronic health record (EHR) data. The study uses NHANES for population-level model development and temporal validation, and MIMIC-IV for EHR-domain evaluation under substantial distribution shift. To reduce circular label reconstruction, the primary analysis excludes disease-defining variables from the corresponding prediction heads, while a full-clinical feature setting is retained only as sensitivity analysis. CardioMeta combines a shared cardiometabolic encoder with disease-specific gated heads and post-hoc probability calibration. In the leakage-reduced temporal validation setting, the model achieved a macro-AUROC of 0.839, macro-AUPRC of 0.536, macro-F1 of 0.614, and expected calibration error of 0.024, with modest but consistent improvements over strong gradient-boosting and neural tabular baselines. External evaluation on MIMIC-IV showed clear degradation under domain shift, while limited fine-tuning partially recovered performance. The findings indicate that the principal value of multi-task cardiometabolic modeling lies not in inflated accuracy, but in reproducible leakage control, calibrated probabilities, and transparent reliability reporting across heterogeneous healthcare data sources.
Antony Garcia, Adrian Noriega, Gabrielle Britton +1cs.LG
Black-box models limit the adoption of artificial intelligence in medicine due to their lack of interpretability and reproducibility. We introduce a statistically grounded framework that provides fully interpretable, rule-based clinical classification using the Bernoulli Naïve Bayes (BNB) model. The method applies supervised $χ^2$-guided statistical binarization to continuous variables, identifying thresholds that maximize association with clinical outcomes within the training data. This transformation allows BNB to operate effectively on continuous medical data without sacrificing its inherent transparency. The approach was evaluated on three benchmark datasets, Pima Indians Diabetes, Wisconsin Breast Cancer, and Heart Failure Prediction, achieving area-under-the-curve (AUC) scores of 0.800 for the Pima analysis, 0.984 for Wisconsin Breast Cancer, and 0.919 for Heart Failure Prediction. In addition to discrimination, probabilistic reliability was assessed using leakage-safe cross-validated calibration analysis including Brier score, calibration intercept/slope, and post-hoc beta calibration, which improved probability calibration across datasets. These results suggest that a statistically interpretable framework can achieve performance comparable to more complex models while providing explicit, clinically meaningful decision rules and calibrated risk estimates. To illustrate this transparency concretely, a complete worked example demonstrates that model inference can be reproduced using only a reference table and basic arithmetic, without access to software or proprietary tools. This work offers a practical approach to supporting trustworthy and generalizable AI in real-world healthcare settings.
Laurin Lux, Alexander H. Berger, Moritz Knolle +2cs.CV cs.AI cs.LG
Region-based loss functions, such as the Dice loss, have established themselves as the de facto standard for highly class- and region-imbalanced segmentation tasks. However, models trained using region-based loss functions are notoriously miscalibrated and typically yield over-confident predictions. In medical imaging applications, such as defining tumor resection margins, this miscalibration is hindering clinical adoption. In this work, we outline a novel gradient perspective on this overconfidence and show how it affects region-based loss functions. We propose a "surgery" on the gradient vector field as a simple, yet effective intervention to mitigate calibration issues. This surgery adds a factor to the loss's partial derivative, scaling the gradient's magnitude linearly with the prediction error. In empirical evaluations across 2D and 3D medical segmentation tasks, we demonstrate the effectiveness of this intervention while maintaining high prediction accuracy when used in conjunction with any region-based loss function.
Deep networks now subtype brain tumors on MRI about as well as specialist readers, yet accuracy is not what keeps them out of the clinic. What matters at the point of care is whether a model's confidence can be trusted to flag the cases it is likely to misclassify and defer them to a human. Deterministic estimates cannot: an auxiliary confidence head trained alongside the classifier collapses to a near-constant output that says nothing about correctness. This study proposes an uncertainty-first pipeline for four-class brain tumor MRI (glioma, meningioma, pituitary, no tumor) that reads predictive uncertainty from Monte Carlo (MC) Dropout over T = 20 passes and turns the resulting entropy into an explicit rule for deferring uncertain cases to a radiologist. We partitioned 7,200 images by perceptual-hash cluster, closing the near-duplicate leakage that inflates accuracy under naive splitting, and evaluated the pipeline on ViT-B/16 and ResNet-50 across five seeds along three axes: discrimination, calibration, and selective prediction. Both discriminate strongly (macro-AUC 0.994; accuracy 0.962 and 0.964), and no seed separates them (0 of 5 significant, p < 0.05), so the result is driven by the uncertainty pipeline, not the network. A single temperature scalar pulls the deterministic softmax into tight calibration (expected calibration error 0.016-0.020), and deferring the most uncertain 5% of cases lifts accuracy on the rest to about 0.98 on both (area under the risk-coverage curve 0.010-0.011). MC-Dropout uncertainty here is thus calibrated, non-collapsing, and directly actionable through a concrete deferral rule, providing an architecture-agnostic basis for calibrated, defer-to-human brain tumor MRI triage under internal validation.
Background: Deep learning models can classify thyroid nodules on ultrasound, but reliable clinical decision support also requires calibrated probabilities, uncertainty estimation, and selective referral, particularly under dataset shift. Methods: We developed a calibrated deterministic five-member deep ensemble for ROI-based thyroid nodule classification and selective image-based triage. TN5000 was used for model development, five-fold cross-validation, member-wise vector-scaling calibration, and fold-specific threshold selection. TN3K served as an independent external dataset-shift evaluation. The framework used ConvNeXt-Tiny with squeeze-and-excitation attention, ensemble-mean malignancy probability, and mutual information (MI) as an ensemble-disagreement score. A three-tier policy assigned images to No-FNA suggestion, FNA recommendation, or radiologist review. Results: On pooled out-of-fold TN5000 predictions, the ensemble achieved AUC-ROC 0.9395, AP 0.9715, ECE 0.0088, and Brier score 0.0813. At 50% nominal MI retention, 7.2% of cases received a No-FNA suggestion, 39.9% an FNA recommendation, and 52.9% radiologist review, with 98.3% No-FNA NPV and 99.83% malignancy capture. On TN3K, AUC-ROC decreased to 0.7870, AP to 0.7254, ECE increased to 0.1899, and Brier score to 0.2281. The frozen TN5000 policy assigned 83.7% to review, 1.0% to No-FNA, and 15.3% to FNA recommendation. No malignant image entered the No-FNA pathway, but FNA-recommendation PPV fell to 76.6%. Conclusion: The framework showed strong internal discrimination and calibration, but limited external threshold transportability. Selective prediction may help identify images unsuitable for automated triage, but local recalibration, threshold validation, and prospective clinical evaluation are required before deployment.
Lin Li, Chaochao Zhou, Benjamin Aubert +2eess.IV cs.CV cs.LG math.NA physics.comp-ph
Accurate geometric calibration is essential for fluoroscopy-guided spinal imaging, digitally reconstructed radiograph (DRR) generation, and 2D--3D vertebral registration. Although calibration quality is typically evaluated using reconstruction-based metrics such as reprojection error, its influence on projection-domain consistency remains poorly understood. This study presents a synthetic framework for evaluating how intrinsic calibration perturbations affect vertebral fluoroscopic projections and downstream registration performance. CT-derived vertebral models and controlled cone-beam imaging geometry were used to generate DRRs with both ground-truth and perturbed intrinsic calibration parameters while maintaining identical anatomy and acquisition pose. Projection-domain changes were quantified using anatomical landmark displacement, contour distance, silhouette overlap, image similarity, and landmark-based 2D--3D registration accuracy in anterior--posterior (AP) and lateral (LAT) views. Results show that even small intrinsic calibration perturbations produce measurable changes in vertebral projection geometry, contour morphology, landmark localization, and DRR appearance. Sensitivity is strongly view dependent, with LAT projections exhibiting substantially greater deformation and anatomical displacement than AP projections. These projection inconsistencies also degrade downstream 2D--3D registration, particularly rotational alignment accuracy. The findings demonstrate that projection-domain consistency complements conventional reconstruction-based calibration metrics and provides a practical framework for assessing calibration robustness. This approach may improve the reliability of DRR generation and fluoroscopy-guided vertebral registration in image-guided spinal applications.
Cervical cytology classification models are typically evaluated on curated, class-balanced benchmarks, but real-world liquid-based cytology (LBC) collections are often small and class-imbalanced. This paper presents a class-imbalance-aware and calibration-aware ensemble classification study on the Mendeley LBC dataset, using its native four-class Bethesda taxonomy (NILM, LSIL, HSIL, SCC) rather than a collapsed binary formulation. Three lightweight architectures (Swin-Tiny, TinyViT-5M, DenseNet121) are trained directly on Mendeley LBC using weighted random sampling to counteract class imbalance, and compared against two soft-voting ensembles (Hybrid-2, Hybrid-3). Post-hoc temperature scaling is fit on a held-out calibration subset carved out of the training portion of each cross-validation fold, distinct from both the training data used to fit model weights and the evaluation fold used for final metrics, avoiding the optimistic calibration estimates that result when the same data is used for both purposes. Calibration substantially reduces expected calibration error, Brier score, and negative log-likelihood for every model and ensemble configuration tested, while discrimination metrics (accuracy, macro-F1, macro-AUROC) remain essentially unchanged. Ensemble size shows no consistent additional reliability benefit over the best individual model once all configurations are properly calibrated. Confusion matrices show that all classification errors, across every configuration, are confined to the boundary between high-grade lesions (HSIL) and carcinoma (SCC); no errors involve the negative (NILM) or low-grade (LSIL) categories. These results suggest that, for this dataset, calibration is the dominant lever for reliability, not ensemble size, though this conclusion should be read in light of the dataset's modest size.
Conformal prediction is being adopted in drug discovery to put an honest number on model reliability: pick an error rate alpha, and the method returns prediction sets containing the true label with probability at least 1 - alpha. We show this guarantee can be dangerous on imbalanced datasets. Across four datasets, standard (marginal) conformal prediction hits its global 90% coverage target while leaving the minority class badly exposed: realized minority coverage falls to 64.8% on blood-brain-barrier penetration and to 4.2% on clinical-trial toxicity, where the rare class is nearly abandoned. The failure is not tied to one model: a random forest, a graph network, and a frozen chemical language model all reproduce it (p < 0.001 in every case), with severity tracking baseline calibration on rare labels rather than architecture. A conservation identity explains the effect: the minority's shortfall equals the majority's surplus amplified by the imbalance ratio, predicting the measured gap to within one point and ordering severity across datasets. The failure survives realistic scaffold splits and a second conformal score, while aggregate accuracy and overall coverage stay reassuringly high, which is exactly why it is easy to miss. Class-conditional (Mondrian) conformal prediction closes the gap on every dataset, restoring minority coverage to target for a modest increase in prediction-set size. We localize the failures to generic molecular scaffolds - plain benzene and pyridine cores occurring in both classes - propose a one-number diagnostic, and show with a cost model that abstaining on affected compounds flips a screening campaign from net-negative to net-positive utility. Our contribution is demonstrating on real chemistry how severe and invisible this known conformal-theory gap becomes under imbalance, and laying out a practical protocol restoring per-class reliability.
In this study, we examine learned preprocessing pipelines in the context of triage-oriented orthopedic abnormality detection task using elbow radiographs from MURA dataset. The evaluation focuses on patient-level detection of musculoskeletal abnormalities under a leakage-aware protocol. We compare multiple preprocessing pipelines, with and without a lightweight DnCNN module as a learned preprocessing component, to assess their impact on discrimination and calibration. Performance is assessed using discrimination metrics (AUROC, PR-AUC), calibration measures (ECE, Brier score), and validation-selected operating point analysis targeting high specificity. Results show that differences across preprocessing strategies are modest and configuration-dependent, with no consistent discrimination advantage over the raw-input DenseNet121 baseline. The raw and diverse inputs combined with the DnCNN front-end showed reduced ECE and Brier score, while CLAHE combined with DnCNN did not improve calibration. Overall, the results suggest that under patient-level evaluation, preprocessing gains are modest and configuration-dependent; the raw-input DenseNet121 baseline remains competitive throughout, and no tested preprocessing strategy produced a consistent discrimination advantage across all metrics.
We introduce the Cross-Platform Fairness Evaluation (CPFE) framework -- a five-axis audit protocol covering discriminative performance, calibration, statistical significance, prediction equity, and attribution stability -- and apply it to four transformer models (BERT, RoBERTa, Emotion-DistilRoBERTa, GoEmotions-RoBERTa) trained on a Kaggle mental health corpus (n=35,556) and evaluated on Reddit (n=6,257) and Twitter (n=2,883) test sets with emotion labels mapped to clinical proxies. All three independently evaluated models exhibit consistent and substantial cross-platform AUC degradation (30.3-35.4% on Reddit, 37.9-39.5% on Twitter) relative to within-platform performance (AUC 0.983-0.987), confirmed across five independent training seeds. Calibration failure is concurrent and severe: ECE rises from 0.056-0.060 in-domain to 0.196-0.229 on Reddit and 0.499-0.542 on Twitter. Platform-specific temperature scaling reduces mean ECE by 88.0% without altering discriminative performance (mean |delta AUC|<0.01), confirming separable failure modes. Prediction equity analysis reveals large cross-platform disparities (raw DI < 0.17; prior-shift-adjusted DI: 0.11-0.29 on Reddit), with equalized odds differences of 0.753-0.830 for mental health proxy classes on Reddit and 0.755-0.831 for anxiety on Twitter. Attribution stability analysis shows near-complete vocabulary divergence across platforms (Jaccard J=0 in 14/16 model-class pairs at K=10). These findings support treating cross-platform validation across all five CPFE axes as a standard requirement for mental health NLP systems in heterogeneous environments. In a single-seed fine-tuning experiment, mean AUC improved by 0.216, suggesting target-platform labels provide greater benefit as training signal than as calibration signal.
In many prediction problems in medical applications, target labels exhibit an inherent ordinal structure, where class ordering reflects clinically meaningful severity levels. The cost associated with misclassification is often non-uniform and asymmetric, as errors between distant ordinal categories may have substantially more severe consequences than errors between adjacent ones, and overestimating disease severity may have different clinical implications than underestimating it. Traditional loss functions such as multi-class cross-entropy treat all misclassifications equally and fail to incorporate this ordering information. Recent advances in ordinal regression aim to address this limitation by integrating rank-based structures into deep learning models. In this work, we introduce the \textbf{Ordinal Cross-Entropy (OCE)} framework, a general and architecture-independent approach for learning from ordinal data. The proposed method extends the standard cross-entropy formulation to account for misclassification severity through an ordinal cost matrix while preserving the probabilistic interpretation and optimization benefits of the conventional loss. We provide a theoretical analysis of the OCE gradient behavior and show that it yields smoother optimization dynamics and improved ordinal consistency. Experiments on benchmark datasets show that our method achieves lower prediction error costs and better calibration compared to existing state-of-the-art ordinal approaches, establishing OCE as a simple yet effective solution for ordinal regression in deep neural networks.
Manar Alsaid, Mandip Shrestha, Mohammad Abbaseess.IV cs.CV cs.LG
Lesion segmentation in breast ultrasound involves two related challenges. In images with lesions, speckle noise, low tissue contrast, and posterior acoustic shadowing cause boundary leakage and incomplete contour delineation. In images without lesions, those same artifacts generate false-positive activations in regions resembling solid lesion tissue. This study addresses both failure modes through a single modification to the training objective. Rather than weighting every boundary pixel equally, the proposed loss scales contour penalties by per-pixel predictive entropy and the ground-truth boundary map, concentrating gradient emphasis on lesion margin locations where the network remains uncertain. The loss was evaluated on the BUSI dataset through a controlled ablation against two baselines: a model without boundary supervision and a model with uniformly weighted boundary binary cross-entropy. Across 97 lesion-containing test images, mean Dice scores were statistically indistinguishable between the proposed method and the no-boundary baseline (0.7624 versus 0.7616, paired Wilcoxon p = 0.27), confirming that lesion segmentation quality is preserved. The primary effect appears in specificity. False-positive activations on 20 no-lesion test images fell from 14 of 20 and 19 of 20 for the two baselines to 5 of 20 with the proposed approach (McNemar p = 0.012 and 0.0005). Non-overlapping Wilson 95% confidence intervals confirm the difference is both statistically significant and practically substantial. A post-hoc spatial temperature scaling step further reduced expected calibration error from 0.0201 to 0.0095 without altering segmentation masks. Entropy-guided boundary supervision and spatial calibration thus function as complementary training-level and inference-level refinements that improve specificity and probability reliability within a U-Net framework.
Multimodal medical imaging fuses complementary anatomical and functional information, yet modalities frequently disagree in pathologically heterogeneous regions. Current segmentation models handle this in one of two inadequate ways: deterministic fusion that averages away disagreement, or post-hoc uncertainty estimation decoupled from the fusion process that produces it. Both obscure the clinically critical question: why is this prediction unreliable? We present EnTrust, a framework that treats inter-modal conflict as the primary source of predictive uncertainty. Our EnFuse module decomposes multimodal features into three disentangled components: shared anatomical consensus (F_c), modality-specific cues (F_{u,m}), and spatially localized conflict signals (F_{cf}), with independence enforced via a cross-covariance objective. This structured decomposition conditions SegDiff, a diffusion-based generative segmentation model whose sampled hypotheses diverge specifically in regions of modal disagreement. TrustMap then translates this hypothesis divergence into calibrated, pixel-wise uncertainty using ensemble entropy, conflict-guided perturbation probing, and a learned calibration head, enabling clinicians to understand not only where predictions are uncertain, but why. Across four benchmarks spanning brain, cardiac, lesion, and oncology domains, EnTrust achieves state-of-the-art segmentation accuracy while reducing calibration error by 40% compared to the strongest baseline. Notably, it outperforms 5x deep ensembles using a single model at roughly half the memory footprint. Code and checkpoints are available at https://github.com/GenMI-Lab/EnTrust.git.
Large language models (LLMs) are increasingly applied to structured clinical data, yet whether they can recognize the limits of their own knowledge on such tasks remains unexplored. We study this question through the lens of cross-model attribution divergence with the goal of reducing epistemic uncertainty for structured tasks, comparing Qwen 2.5 7B and XGBoost on a prediction task via attribution divergence analysis. We report four findings. First, LLM verbalized confidence is epistemically vacuous, it outputs a near-constant (0.856-0.937) regardless of whether accuracy is 49% or 75.3%, tracking prompt format rather than prediction quality. Second, the LLM exhibits an inverse difficulty effect: accuracy drops to 64.8% when XGBoost is 99% correct, but matches XGBoost (73.8% vs. 73.1%) when it is moderately uncertain. Third, few-shot examples and SHAP-derived feature evidence are orthogonal, super-additive interventions: they reduce the Attribution Disagreement Score (ADS) from 1.54 to 0.38 and improve accuracy from 49% to 75.3% without training. Fourth, a cross-model calibrator that determined LLM reliability using attribution divergence signals reduces expected calibration error from 0.254 to 0.080, replacing uninformative verbalized confidence with patient-specific reliability estimates, without accessing model internals or requiring repeated inference. We frame these findings as a cold start problem for LLMs on structured data and outline a path toward genuine epistemic self-awareness.
Csaba Kiss, Roland Molontay, Gabriele Pergolacs.LG cs.CL
Distinguishing causal adverse drug events (ADEs) from spurious correlations remains a central challenge in pharmacovigilance. The InferBERT framework integrates transformer models with Do-calculus, but its success hinges on the underlying classification model. This study evaluates the impact of model choice in InferBERT, assessing whether simpler models suffice, if domain-specific pre-training helps, whether scaling to LLMs improves causal detection, and the effect of post-hoc calibration. We performed a comparative study on two benchmarks: Analgesics-induced Acute Liver Failure (AILF) and Tramadol-related Mortalities (TRAM). Four models were evaluated-XGBoost (baseline), ALBERT (original InferBERT), BioBERT (biomedical transformer), and Med-LLaMA (medical LLM)-using 5-fold cross-validation repeated over 20 runs. We measured accuracy, Expected Calibration Error (ECE) pre- and post-isotonic regression, and Jaccard concordance of causal terms with PRR, ROR, and EBGM; significance was tested with paired t-tests. BioBERT achieved the highest accuracy on both datasets, while Med-LLaMA underperformed despite its size and parameter-efficient fine-tuning. Domain-specific pre-training was decisive. Calibration improved ECE but had mixed effects on accuracy and causal discovery. BioBERT's superiority also yielded the strongest concordance with traditional pharmacovigilance signals. These results show that domain-specific pre-training provides a clear advantage over simpler baselines and larger LLMs. Investing in manageable, domain-aware models is more effective for computational pharmacovigilance than simply scaling model size.