BackgroundDeep learning classifiers for breast cancer histology achieve expert-level accuracy but do not explain which tissue regions drove the diagnosis. We present GRALIS-Report, an attribution pipeline with two defining architectural properties: raw images never enter the report generation stage, and no information is fused across modalities in a latent space. Attribution signals—not pixels—are the sole input to pathology-oriented language; every modality transition is explicit, symbolic, and traceable to the stored structured attribution record, making every visual-totext inference auditable and independently verifiable.MethodsThe system: (i) trains a DenseNet-121 classifier on BreaKHis via knowledge distillation (high internal accuracy on a patient-level split); (ii) applies GRALIS—which transforms the image into a semantic attribution signal: a per-superpixel importance score ϕi computed via coalition-conditioned path integration—and (iii) converts that structured signal into a research SOAP-style report. Formal theoretical properties—canonical form, a priori convergence bound, and structural incompatibility of locality with exact completeness—are proved in a companion preprint (arXiv:2605.05480); the present paper is entirely experimental.ResultOn the BreaKHis test set (1,187 images), independent faithfulness benchmarks place GRALIS at rank 2 of 6 on pixel-level deletion AUC and rank 2 of 3 on ROAD MoRF AUC (the three methods for which ROAD was computed), with the largest MoRF-LeRF discrimination gap among the three evaluated methods. This ranking reflects a deliberate design trade-off: by operating at superpixel rather than pixel resolution, GRALIS sacrifices marginal pixel-level faithfulness relative to Integrated Gradients in exchange for region level spatial coherence, a pre-run Monte Carlo sample-size bound (not numerically instantiated in this paper), and a fully auditable attribution-to-report pipeline—properties that pixel-precise methods do not jointly provide. Cross-dataset evaluation on two held-out external subsets further characterises this trade-off: on IDC Breast Cancer (50 × 50 px patches, frozen backbone), GRALIS ranks first on both Deletion AUC and ROAD MoRF; on PatchCamelyon (96 × 96 px), GRALIS ranks sixth—a result plausibly associated with a mismatch between the fixed superpixel granularity (nseg = 30) and the finer discriminative feature scale of lymph-node patches, although a dedicated nseg ablation would be required to test this explanation. A supplementary ExpiScores profile is reported alongside these independent metrics; we caution that this metric shares authorship with the present work and should be weighted accordingly. The deterministic engine generated 1,187/1,187 syntactically complete reports with no execution failures; 1,175 (98.99%) corresponded to correct classifier predictions, with the 12 discordant cases identified retrospectively using test labels, operating fully offline. An expert perception discordance study (N = 4 anatomopathologists, 60 cases) reveals marked inter-rater variability in perceived clinical utility, suggesting that perceived explanation utility may not constitute a stable ground truth in histopathology.RESULTSThis is reported as a methodological finding for the XAI evaluation community, not as evidence of clinical utility. No clinical efficacy claims are made.
IntroductionBreast cancer remains the most commonly diagnosed malignancy in women worldwide, with approximately 2.3 million new cases and 670,000 deaths in 2022. Artificial intelligence (AI) is increasingly being applied across breast imaging for detection, characterization, segmentation, risk stratification, explainability, and clinical translation.MethodsThis systematic review synthesized 284 eligible peer-reviewed publications published between 2015 and 2025, identified through a PRISMA 2020-compliant search of four indexed databases. A supplementary relevance-ranked Google Scholar screen was reported separately. A targeted May 2026 narrative update of recent prospective and implementation studies was also conducted without adding these records to the PRISMA denominator.ResultsThe reviewed evidence encompassed classical machine learning and radiomics, convolutional neural networks, YOLO-family detectors, vision transformers and hybrid CNN-Transformer architectures, U-Net variants, Mask R-CNN, SAM, and MedSAM across mammography, tomosynthesis, MRI, ultrasound, contrast-enhanced mammography, and emerging photoacoustic imaging. The evidence highlights advances in explainable AI, multimodal and radiogenomic fusion, federated learning, and open-source deployment, while persistent challenges include data heterogeneity, class imbalance, domain shift, bias, incomplete calibration, and limited prospective validation. Retrospective benchmark findings were derived from the 284-study systematic corpus, whereas prospective clinical utility and real-world implementation were informed by the separate 2026 narrative update.DiscussionAI shows substantial potential to enhance breast cancer imaging, but broader clinical translation requires robust external and prospective validation, improved calibration, assessment of generalizability and bias, and integration into clinical workflows. Supplementary Data Sheet S1 should be interpreted as a provenance-tagged evidence map rather than a formal study-level risk-of-bias assessment, with 38 entries verified from full text, 48 based on abstracts, and 198 based on DOI/source pages or other web-accessible records.
Deepfake detection remains a challenging problem due to the ever-evolving techniques used to synthesize realistic fake images and videos. Traditional detection methods primarily operate in the pixel domain, often struggling to generalize across different forgery types. In this work, we explore an approach that integrates wavelet-based frequency analysis with deep learning to enhance deepfake detection. By decomposing input images using the Discrete Wavelet Transform (DWT), we extract multi-resolution features that capture both global structures and fine-grained manipulation artifacts; such features are then input to a neural network. Our results show that level-1 DWT features mainly induce a sensitivity-specificity tradeoff: they improve fake-class sensitivity and F1 in cross-forgery settings, while RGB models retain stronger specificity and in-forgery performance. This suggests that wavelet sub-bands expose manipulation cues that are useful for detecting unseen fake classes, but they should not be interpreted as a uniform robustness improvement.
As an answer to the need to prove the authenticity of digital images, several forensic tools for image forgery localization have been proposed in the past. However, authenticity analysis remained a challenging task, which requires an expert's knowledge to correctly interpret each forensic algorithm's output. Moreover, since different tools look for different manipulation traces, a thorough analysis requires the joint interpretation of the maps produced by several tools, which is nontrivial since each tool's reliability is possibly affected by different elements. Recently, deep learning-based forgery localization schemes were proposed, allowing for more automated reasoning; however, their accuracy significantly decreases when they are tested on forgeries that deviate from those used for the training phase. This work proposes a deep learning-based framework that merges the forgery localization maps provided by model-based image forensics tools based on the U-Net architecture. The experiments show that the proposed approach improves the quality of forgery localization maps compared to those produced by single tools and by state-of-the-art fusion frameworks while simultaneously achieving a desirable generalization capability.
This study investigates the correlation between misclassifications in DeepFake detection and high-level facial attributes. A pre-trained frame-level classifier is used to distinguish manipulated from authentic video content, and its wrong predictions are analyzed in detail. To enrich the dataset, we automatically annotate each video with additional labels, including gender, hair color, hair length, ear visibility, and ethnicity, using a semi-supervised facial-attribute recognition pipeline. We extend this analysis with controlled training-time exclusions on FaceForensics++, keeping a unified test set to isolate generalization bias. Compared to the no-exclusion baseline (Accuracy = 0.806, AUC = 0.823), excluding samples with ears visible yields the largest degradation (Accuracy = 0.741, AUC = 0.763), while excluding non-visible ears has a milder effect (Accuracy = 0.813, AUC = 0.832). Hair length shows a moderate but consistent impact that interacts with ear visibility. We also explain the observed confusion-matrix asymmetry as a consequence of fixed score thresholds and video-level k-of-n aggregation. The results demonstrate that ear visibility is a critical factor for robust FAKE vs. REAL discrimination and motivate attribute-aware training, including targeted data curation, attribute-specific augmentation, and threshold calibration. The proposed framework provides actionable guidance for bias-aware training strategies and supports the development of more interpretable and operationally reliable DeepFake detection systems.
Complete blood count (CBC) reports are widely used in clinical settings to determine hematological disorders at early stages, offering a cost-effective and rapid alternative to expensive, time-consuming disease-specific tests. This study presents an AI-based method for multi-disease classification using CBC report data. We collected real-world paper-based CBC reports from a local hospital in scanned form for digital processing. State-of-the-art OCR models were used to extract only table data by removing patients' demographics. After cleaning and pre-processing, a custom digital dataset comprising 6,500 patients with 12 clinically relevant parameters was constructed. A rule-based algorithm is designed for disease labeling. We tested machine learning models for binary and multi-disease prediction. The experimental results showed that all machine learning models yielded the highest accuracy after hyperparameter tuning. Gradient Boosting and Random Forest outperform on binary and multi-disease prediction. The multi-label classification technique identified multiple probable pathologies from extracted CBC records, providing an efficient and cost-effective substitute for disease prediction. This approach offers significant benefits in healthcare for early disease diagnosis while minimizing delays associated with traditional testing methods.
IntroductionDeep learning has achieved strong results in histopathological diagnosis, but the lack of transparency in deep neural network models limits their clinical acceptance. Explainable Artificial Intelligence (XAI) offers a principled route to address this gap by quantifying the reliability and morphological coherence of model predictions in support of clinical decision-making.MethodsWe present an integrated, explainability-augmented deep learning framework for binary histopathological classification on the BreaKHis dataset. The pipeline combines knowledge-distilled InceptionV3 → DenseNet-121 training with three complementary local explanation methods—Gradient-weighted Class Activation Mapping (Grad-CAM), SHapley Additive exPlanations (SHAP), and Local Interpretable Model-agnostic Explanations (LIME). Per-image outputs are aggregated at the dataset level into the ExpiScore composite interpretability index, which captures salience reliability, cross-method coherence, and morphological alignment. Additional regression and decision-tree modules build a trust meta-model that estimates the probability of prediction errors and supports automated triage of clinical cases. Two operational modes are evaluated and kept strictly separate throughout: a fully-automatic mode (CNN classification with no human intervention) and a human-in-the-loop mode (meta-model routing of high-risk cases to expert review).ResultsIn fully-automatic mode, the distilled DenseNet-121 student achieves accuracy 0.9916, F1-score 0.9939, Area Under the receiver-operating-characteristic Curve (AUC) 0.9989, sensitivity 0.9939, and specificity 0.9866 on the 1,187-image held-out test set, with five-fold cross-validation (BreaKHis official folds) further indicating generalization stability (mean accuracy 0.8962 ± 0.0177; mean AUC 0.9598 ± 0.0094). In human-in-the-loop mode, a meta-model trained on XAI-derived features and independently evaluated on the held-out test set routes the 20% of cases with the highest predicted error probability to expert review, raising post-routing accuracy from 0.9898 to 0.9983 (+0.85 pp) and reducing false negatives by 83.3% under the assumption of correct expert resolution of all routed cases.DiscussionThe two regimes are not directly comparable and are reported separately throughout the paper. The fully-automatic figures characterize purely algorithmic performance; the human-inthe-loop figures characterize a clinical-decision-support workflow in which the framework prioritizes cases for expert review. Together they suggest that ExpiScore-driven routing carries discriminative predictive signal for error detection and supports a principled trade-off between predictive accuracy and interpretative transparency, providing held-out test evidence for the potential clinical utility of the proposed routing mechanism within this evaluation setting. External multi-cohort validation and formal calibration of the triage thresholds are required before clinical deployment.
Breast cancer early detection using liquid biopsy, circulating tumor DNA (ctDNA), artificial intelligence (AI), and multimodal fusion offers a promising but still emerging research strategy to overcome the limitations of conventional imaging. Breast cancer screening and molecular diagnostics remain constrained by the inability of mammography and magnetic resonance imaging to detect pre-invasive disease, alongside the low sensitivity and spatial ambiguity of ctDNA in early-stage settings. This mini-review summarizes the rationale and recent advances in AI-driven multimodal frameworks that integrate imaging phenotypes with blood-derived genotypic signals through feature-, decision-, and intermediate-level fusion strategies. Such approaches improve diagnostic sensitivity and specificity by capturing complementary biological and structural information, enabling earlier detection and longitudinal risk assessment. Despite this progress, clinical translation is hindered by data heterogeneity, the lack of standardized multimodal datasets, and limited prospective validation. This study highlights the emerging biology-first, imaging-informed framework. Despite recent progress, current multimodal approaches remain largely investigational and require robust prospective evidence before clinical deployment. It outlines key future directions, including federated learning, longitudinal modeling, and large-scale validation, to support the future evolution of scalable and equitable early-detection strategies.
Video captured for action recognition often contains sensitive appearance cues such as faces, skin color, and clothing. Models trained on such data may exploit these cues rather than the underlying motion, raising privacy concerns in real-world deployment. In this work, we study action recognition under a motion-focused constraint: the model receives only motion representations that capture pixel displacement over time, while reducing appearance cues that expose identity or scene context. We focus on motion-history images and optical flow as learning-free representations that reduce identifiable appearance information while retaining action recognition accuracy. Our motion I3D model achieves approximately 31% and 52% zero-shot top-1 accuracy on HMDB-51 and UCF-101, respectively, outperforming non-CLIP direct-transfer baselines trained on Kinetics-400 despite operating without any appearance input. In 16-shot adaptation, the same model reaches 52% and 83% top-1 accuracy. In the domain adaptation setting on TP-HMDB↔TP-UCF, our motion-focused models achieve higher action recognition accuracy than prior privacy-preserving methods. Sensitive attribute predictability is reduced relative to RGB by a comparable margin, without requiring a learned privacy filter. On PA-HMDB51, optical flow is the strongest motion representation for privacy preservation, approaching chance level for skin-color prediction and remaining below RGB on most privacy attributes, indicating that motion representations retain useful action information while exposing less personal information.
In Biometrics, the presence of privacy restrictions on personal data transfer and storage poses significant challenges in creating a sufficiently comprehensive and varied dataset by leveraging various data sources for traditional batch-based training procedures. This is particularly true in the Morphing Attack Detection (MAD) task, in which data involves facial images and a limited number of public datasets of well-controlled images are available. In this context, MAD systems generally suffer from limited generalization capabilities, with low performance on new and unseen data. Therefore, in this paper, we propose Adaptive-LwF, adopting the recent paradigm of Continual Learning (CL) as a viable solution to enable incremental training across multiple sites. Indeed, CL assumes that once a model has been trained, previous data cannot be utilized in subsequent training iterations and can be deleted. In particular, we investigate the performance of different methods in this new scenario, where a model is updated each time a new chunk of data, of variable size, becomes available. We focus our attention on the well-known Learning without Forgetting (LwF) algorithm, proposing a novel adaptive approach able to automatically fine-tune its parameters in relation to the variable size of the specific input chunks. Experimental results confirm that our approach is capable of mitigating the catastrophic forgetting effects, and the superior performance of the Adaptive-LwF algorithm with respect to alternative solutions.
Modern lifestyle is strongly marked by the presence of aluminum (Al) in practically all human consumer products. Bone tissue is one of the main sites of Al accumulation, and its toxic effects are well known in individuals subjected to chronic exposure. However, there is still a gap in knowledge regarding the effects of Al on bone formation in the neonatal period. This study evaluated the effect of Al ingestion on rat tibiae during the neonatal period. Wistar rats were divided into control and Al groups. The Al group received AlCl3 (2.02 mg/kg/day) via gavage for fifteen days, then, the right tibiae were used to evaluate osteoblast and osteocyte ultrastructure and bone microarchitecture using transmission electron microscopy and computed X-ray microtomography, respectively. Al promoted swelling and altered mitochondrial crests in osteoblasts. Osteocytes showed accumulation of electron-dense lysosomes and absence of the osmiophilic lamina in the lacunae, showing characteristics similar to osteocytic osteolysis. Cortical Thickness (Ct.Th), Trabecular thickness (Tb.th) and trabecular number (Tb.N) decreased whilst trabecular spacing (Tb.Sp) increased. These results suggest that Al intake during the neonatal period may affect the function of osteoblasts and osteocytes besides compromising bone formation.
Cardiac adipose tissue (CAT) has emerged as a critical and clinically relevant factor in cardiovascular disease (CVD), yet its full impact remains largely overlooked. The amount of fat surrounding the heart can influence major blood vessels by promoting plaque formation. In conditions such as cardiac steatosis or fatty heart disease, fat infiltration or accumulation within the heart muscle compromises its function may play a role in heart failure (HF) and coronary artery disease (CAD). This review explores the different types of fat deposits surrounding the heart, focusing on the potential contribution of CAT to cardiovascular disease (CVD). Three main imaging modalities for assessing cardiac fat are discussed, including magnetic resonance imaging (MRI), computed tomography (CT), and echocardiography. The segmentation and quantification of the fat for each imaging modality are also presented, correlating these measurements with CVD risk. Each imaging modality offers distinct advantages and limitations in segmenting and quantifying fat. Despite its clinical significance, quantification and characterization of CAT remain challenging, requiring advanced imaging techniques for precise assessment. Future research should focus on unlocking the mechanistic pathways that link CAT to adverse cardiovascular outcomes, ultimately enhancing our ability to predict, prevent, and treat heart disease with greater precision. As imaging technology advances, there is a need for refined segmentation methods and consensus-driven guidelines to establish CAT as a key biomarker in CVD risk stratification.
IntroductionHomomorphic encryption (HE) enables privacy-preserving face recognition by allowing encrypted facial embeddings to be compared without decryption. While efficient, these systems often reveal comparison scores in plaintext, introducing a security risk. Revealing these scores can potentially allow adversaries to reconstruct sensitive facial embeddings and infer demographic attributes, thus compromising user privacy.MethodsThis work proposes a training-less face template recovery attack leveraging the Lagrange multiplier optimization method. The attack requires only a small set of randomly generated synthetic facial images and their associated comparison scores with a target template. The method assumes attackers use spoofed synthetic faces and lack direct access to the face recognition system, aligning with real-world threat models.ResultsExperimental evaluation demonstrates the feasibility and effectiveness of the proposed attack. It shows that between 50 and 192 comparison scores and synthetic images are sufficient to recover the target face template with 100% success under strict system thresholds. The recovered templates closely resemble the original and retain identifiable soft biometric traits.DiscussionThe findings reveal a critical vulnerability in face recognition systems employing inner product similarity measures under homomorphic encryption. Even without system access or training data, attackers can exploit leaked comparison scores to compromise facial privacy. The study underscores the need to reassess how score leakage is handled in encrypted recognition systems and explore stronger protection mechanisms against template reconstruction.
IntroductionDeepfakes have become ubiquitous in our modern society, with both their quantity and quality increasing. The current evolution of image generation techniques makes the detection of manipulated content through visual inspection increasingly difficult. This challenge has motivated researchers to analyze heart-beat-related signal to distinguish deep fakes from genuine videos.MethodsIn this study, we analyze deepfake videos of faces generated with novel methods regarding their heart-beat-related signals using remote photoplethysmography (rPPG). The rPPG signal describes the blood flow based, or rather local blood volume changes, and thus reflects the pulse signal. For our analysis, we present a pipeline that extracts rPPG signals and investigate the origin of the extracted signals in deepfake videos using correlation analyses. To validate our rPPG extraction pipeline and analyze rPPG signals of deepfakes, we captured a dataset of facial videos synchronized with an electrocardiogram (ECG) as a ground-truth pulse signal. Additionally, we generated high-quality deepfakes and incorporated publicly available datasets into our evaluation.ResultsWe prove that our heart rate extraction pipeline produces valid estimates for genuine videos by comparing the estimated results with ECG reference data. Our high-quality deepfakes exhibit valid heart rates and their rPPG signals show a significant correlation with the corresponding driver video that was used to generate them. Furthermore, we show that this also holds for deepfakes from a publicly available dataset.DiscussionPrevious research assumed that the subtle heart-beat-related signals get lost during the deepfake generation process, making them useful for deepfake detection. However, this paper shows that this assumption is no longer valid for current deepfake methods. Nevertheless, preliminary experiments indicate that analyzing spatial distribution of bloodflow regarding its plausibility can still help to detect high quality deepfakes.
We have developed a novel approach for high-resolution confocal imaging across multiple tissue planes simultaneously. By combining confocal microscopy, piezo actuators, and optogenetic sensors, we can simultaneously capture images of dynamic fluorescence signals from various cell populations in different tissue layers (Z planes). This enables the decoding of cell-to-cell communication through complex tissues, offering a significant advancement in understanding how cells in distinct layers of tissue communicate and coordinate their functions and produce integrated behaviors. For example, our technique sheds light on myogenic coordination underlying colonic motility. Examining various cell types, such as interstitial cells of Cajal (ICC) and smooth muscle cells (SMC), distributed through the thickness of muscle layers, we demonstrate distinct Ca2+ signaling patterns and organization that underlie complex colonic motor activities.
BackgroundPresurgical evaluation of the histopathological grade of soft tissue sarcoma (STS) is important for enacting treatment strategies. In this study, we plan to investigate the correlation of high-output ultrasound (US) radiomic features and the histopathological grade of STS.MethodsPatients with STS were retrospectively enrolled. The radiomic features were extracted from the US images of the STS lesions. The lesions were graded according to the Fédération Nationale des Centers de Lutte Contre le Cancer (FNCLCC) histopathological grading system. The correlation of the radiomic features and the FNCLCC grades was evaluated. We used the features correlated with the histopathological grades to build a model for predicting high-grade STS (Grade II and III).ResultsA total of 79 patients with STS were enrolled. And 15 radiomic features were found correlated with the FNCLCC grades of STSs, with the correlation coefficient ranging from 0.22 to 0.38. And 8 features showed significant difference among the three grades. The model for predicting high-grade STS based on the 8 radiomic features had an AUC value of 0.80, a sensitivity of 0.73, and a specificity of 0.78.ConclusionThe US radiomic features were correlated with the FNCLCC grade of STS. The radiomic analysis of US imaging could be potentially helpful for identifying the FNCLCC grades of STS pre-surgically.
Background Automatic digital photogrammetry produces digital reproductions of objects using photographs. The aim of this study is to analyze feasibility of photogrammetry for electroanatomic map (EAM) reconstruction from different mapping systems. Furthermore, the possibility to import the reconstructed EAMs in a common working space is evaluated. Methods All consecutive patients undergoing EAM with one of the following EAM systems were screened for the study: (1) CARTO™; (2) Ensite™ X; (3) Rhythmia™; (4) Affera™ PRISM-3. All patient geometries were reconstructed from a video acquisition within the source EAM software. The video obtained was processed with Zephyr software and a dense point cloud was obtained. An image or sequence of images was selected to build a 3D mesh. At the end, the mesh was imported in the 3D graphics software Blender. Result A total of 24 EAMs from 24 patients were included in the study. All EAMs were reconstructed with success using photogrammetry from all 4 mapping systems assessed. The process time was ≈ 25 min. In particular, EAMs were as follows: left atrium (2 Carto; 2 Ensite; 5 Rhythmia; 2 Affera), right atrium (1 Carto; 6 Ensite; 3 Affera) and left ventricles (1 Carto; 2 Ensite). All the reconstructed EAMs were imported in Blender with success. They could be visualized in Blender and all the operations were allowed including moving EAMs in a common working space and EAMs overlap. Conclusion This study demonstrated for the first time the possibility of realizing 3-D objects from digital video formats of different EAMs.
Early detection and management of diabetic peripheral neuropathy (DPN) are critical to reducing associated morbidity and mortality. Corneal Confocal Microscopy (CCM) facilitates the imaging of corneal nerves to detect early and progressive nerve damage in DPN. However, its wider adoption has been limited by the subjectivity and time-intensive nature of manual nerve fiber quantification. This study investigates the diagnostic utility of state-of-the-art Vision Transformer (ViT) models for the binary classification of CCM images to distinguish between healthy controls and individuals with DPN. The ViT model's performance was also compared to ResNet50, a convolutional neural network (CNN) previously applied for DPN detection using CCM images. Using a dataset of approximately 700 CCM images, the ViT model achieved an AUC of 0.99, a sensitivity of 98%, a specificity of 92%, and an F1-score of 95%, outperforming previously reported methods. These findings highlight the potential of the ViT model as a reliable tool for CCM-based DPN diagnosis, eliminating the need for time-consuming manual image segmentation. Moreover, the results reinforce CCM's value as a non-invasive and precise imaging modality for detecting nerve damage, particularly in neuropathy-related conditions such as DPN.
IntroductionThe sensitivity and resolution of magnetic particle imaging (MPI) depend on the choice of tracer and specific imaging parameters. For cell tracking applications with MPI, both the superparamagnetic iron oxide (SPIO) tracer and the cell labeling efficiency have a significant impact on MPI sensitivity and vary for different tracers.MethodsThis study compared three commercially available SPIO tracers (VivoTrax, Synomag-D and ProMag) and SPIO-labeled cells using magnetic particle relaxometry (MPR) and imaging. Further, the effect of imaging parameters (high and low gradient field strength and drive field amplitude) on MPI signal strength, resolution, and cell detection limits, was evaluated.ResultsThe peak MPI signal measured by MPR was much higher for Synomag-D compared to VivoTrax and ProMag. However, the signal for intracellular Synomag-D was significantly reduced. In contrast, the signal for ProMag, a micron-sized iron oxide (MPIO) particle, was not significantly different for free and intracellular particles. The cellular iron loading was higher for ProMag compared to Synomag-D. The total MPI signal measured from images of free and intracellular SPIOs was highest for ProMag. Varying imaging parameters confirmed that a lower gradient field strength and higher drive field amplitude improved tracer and cellular sensitivity.DiscussionThese results, in addition to prior work from our lab, suggest that MPIOs are a good option for cell tracking with MPI. In conclusion, the evaluation of tracers by MPR is not sufficient to predict the performance of all SPIO tracers; in particular, not for larger, polymer-encapsulated iron particles such as ProMag, or for SPIO tracers internalized in cells. Improvements in MPI sensitivity through lower gradient field strength and higher drive field amplitudes are associated with a trade-off in image resolution.
