Using Machine Learning to Reduce the Need for Contrast Agents in Breast MRI through Synthetic Images.

Background Reducing the amount of contrast agent needed for contrast-enhanced breast MRI is desirable. Purpose To investigate if generative adversarial networks (GANs) can recover contrast-enhanced breast MRI scans from unenhanced images and virtual low-contrast-enhanced images. Materials and Methods In this retrospective study of breast MRI performed from January 2010 to December 2019, simulated low-contrast images were produced by adding virtual noise to the existing contrast-enhanced images. GANs were then trained to recover the contrast-enhanced images from the simulated low-contrast images (approach A) or from the unenhanced T1- and T2-weighted images (approach B). Two experienced radiologists were tasked with distinguishing between real and synthesized contrast-enhanced images using both approaches. Image appearance and conspicuity of enhancing lesions on the real versus synthesized contrast-enhanced images were independently compared and rated on a five-point Likert scale. P values were calculated by using bootstrapping. Results A total of 9751 breast MRI examinations from 5086 patients (mean age, 56 years ± 10 [SD]) were included. Readers who were blinded to the nature of the images could not distinguish real from synthetic contrast-enhanced images (average accuracy of differentiation: approach A, 52 of 100; approach B, 61 of 100). The test set included images with and without enhancing lesions (29 enhancing masses and 21 nonmass enhancement; 50 total). When readers who were not blinded compared the appearance of the real versus synthetic contrast-enhanced images side by side, approach A image ratings were significantly higher than those of approach B (mean rating, 4.6 ± 0.1 vs 3.0 ± 0.2; P < .001), with the noninferiority margin met by synthetic images from approach A (P < .001) but not B (P > .99). Conclusion Generative adversarial networks may be useful to enable breast MRI with reduced contrast agent dose. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Bahl in this issue.

Artificial Intelligence for Clinical Interpretation of Bedside Chest Radiographs.

Background Supine chest radiography for bedridden patients in intensive care units (ICUs) is one of the most frequently ordered imaging studies worldwide. Purpose To evaluate the diagnostic performance of a neural network-based model that is trained on structured semiquantitative radiologic reports of bedside chest radiographs. Materials and Methods For this retrospective single-center study, children and adults in the ICU of a university hospital who had been imaged using bedside chest radiography from January 2009 to December 2020 were reported by using a structured and itemized template. Ninety-eight radiologists rated the radiographs semiquantitatively for the severity of disease patterns. These data were used to train a neural network to identify cardiomegaly, pulmonary congestion, pleural effusion, pulmonary opacities, and atelectasis. A held-out internal test set (100 radiographs from 100 patients) that was assessed independently by an expert panel of six radiologists provided the ground truth. Individual assessments by each of these six radiologists, by two nonradiologist physicians in the ICU, and by the neural network were compared with the ground truth. Separately, the nonradiologist physicians assessed the images without and with preliminary readings provided by the neural network. The weighted Cohen κ coefficient was used to measure agreement between the readers and the ground truth. Results A total of 193 566 radiographs in 45 016 patients (mean age, 66 years ± 16 [SD]; 61% men) were included and divided into training (n = 122 294; 64%), validation (n = 31 243; 16%), and test (n = 40 029; 20%) sets. The neural network exhibited higher agreement with a majority vote of the expert panel (κ = 0.86) than each individual radiologist compared with the majority vote of the expert panel (κ = 0.81 to ≤0.84). When the neural network provided preliminary readings, the reports of the nonradiologist physicians improved considerably (aided vs unaided, κ = 0.87 vs 0.79, respectively; P < .001). Conclusion A neural network trained with structured semiquantitative bedside chest radiography reports allowed nonradiologist physicians improved interpretations compared with the consensus reading of expert radiologists. © RSNA, 2022 Supplemental material is available for this article. See also the editorial by Wielpütz in this issue.

Weakly supervised annotation-free cancer detection and prediction of genotype in routine histopathology.

Deep learning is a powerful tool in computational pathology: it can be used for tumor detection and for predicting genetic alterations based on histopathology images alone. Conventionally, tumor detection and prediction of genetic alterations are two separate workflows. Newer methods have combined them, but require complex, manually engineered computational pipelines, restricting reproducibility and robustness. To address these issues, we present a new method for simultaneous tumor detection and prediction of genetic alterations: The Slide-Level Assessment Model (SLAM) uses a single off-the-shelf neural network to predict molecular alterations directly from routine pathology slides without any manual annotations, improving upon previous methods by automatically excluding normal and non-informative tissue regions. SLAM requires only standard programming libraries and is conceptually simpler than previous approaches. We have extensively validated SLAM for clinically relevant tasks using two large multicentric cohorts of colorectal cancer patients, Darmkrebs: Chancen der Verhütung durch Screening (DACHS) from Germany and Yorkshire Cancer Research Bowel Cancer Improvement Programme (YCR-BCIP) from the UK. We show that SLAM yields reliable slide-level classification of tumor presence with an area under the receiver operating curve (AUROC) of 0.980 (confidence interval 0.975, 0.984; n = 2,297 tumor and n = 1,281 normal slides). In addition, SLAM can detect microsatellite instability (MSI)/mismatch repair deficiency (dMMR) or microsatellite stability/mismatch repair proficiency with an AUROC of 0.909 (0.888, 0.929; n = 2,039 patients) and BRAF mutational status with an AUROC of 0.821 (0.786, 0.852; n = 2,075 patients). The improvement with respect to previous methods was validated in a large external testing cohort in which MSI/dMMR status was detected with an AUROC of 0.900 (0.864, 0.931; n = 805 patients). In addition, SLAM provides human-interpretable visualization maps, enabling the analysis of multiplexed network predictions by human experts. In summary, SLAM is a new simple and powerful method for computational pathology that could be applied to multiple disease contexts. © 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.

Theranostic Trigger-Responsive Carbon Monoxide-Generating Microbubbles.

Carbon monoxide (CO) is a gaseous signaling molecule that modulates inflammation, cell survival, and recovery after myocardial infarction. However, handling and dosing of CO as a compressed gas are difficult. Here, light-triggerable and magnetic resonance imaging (MRI)-detectable CO release from dimanganese decacarbonyl (CORM-1) are demonstrated, and the development of CORM-1-loaded polymeric microbubbles (COMB) is described as an ultrasound (US)- and MRI-imageable drug delivery platform for triggerable and targeted CO therapy. COMB are synthesized via a straightforward one-step loading protocol, present a narrow size distribution peaking at 2 µm, and show excellent performance as a CORM-1 carrier and US contrast agent. Light irradiation of COMB induces local production and release of CO, as well as enhanced longitudinal and transversal relaxation rates, enabling MRI monitoring of CO delivery. Proof-of-concept studies for COMB-enabled light-triggered CO release show saturation of hemoglobin with CO in human blood, anti-inflammatory differentiation of macrophages, reduction of hypoxia-induced reactive oxygen species (ROS) production, and inhibition of ischemia-induced apoptosis in endothelial cells and cardiomyocytes. These findings indicate that CO-generating MB are interesting theranostic tools for attenuating hypoxia-associated and ROS-mediated cell and tissue damage in cardiovascular disease.

Hybrid total-body pet scanners-current status and future perspectives.

Purpose Since the 1990s, PET has been successfully combined with MR or CT systems. In the past years, especially PET systems have seen a trend towards an enlarged axial field of view (FOV), up to a factor of ten. Methods Conducting a thorough literature research, we summarize the status quo of contemporary total-body (TB) PET/CT scanners and give an outlook on possible future developments. Results Currently, three human TB PET/CT systems have been developed: The PennPET Explorer, the uExplorer, and the Biograph Vision Quadra realize aFOVs between 1 and 2 m and show a tremendous increase in system sensitivity related to their longer gantries. Conclusion The increased system sensitivity paves the way for short-term, low-dose, and dynamic TB imaging as well as new examination methods in almost all areas of imaging.

Automation of data analysis in molecular cancer imaging and its potential impact on future clinical practice.

Digitalization, especially the use of machine learning and computational intelligence, is considered to dramatically shape medical procedures in the near future. In the field of cancer diagnostics, radiomics, the extraction of multiple quantitative image features and their clustered analysis, is gaining increasing attention to obtain more detailed, reproducible, and meaningful information about the disease entity, its prognosis and the ideal therapeutic option. In this context, automation of diagnostic procedures can improve the entire pipeline, which comprises patient registration, planning and performing an imaging examination at the scanner, image reconstruction, image analysis, and feeding the diagnostic information from various sources into decision support systems. With a focus on cancer diagnostics, this review article reports and discusses how computer-assistance can be integrated into diagnostic procedures and which benefits and challenges arise from it. Besides a strong view on classical imaging modalities like x-ray, CT, MRI, ultrasound, PET, SPECT and hybrid imaging devices thereof, it is outlined how imaging data can be combined with data deriving from patient anamnesis, clinical chemistry, pathology, and different omics. In this context, the article also discusses IT infrastructures that are required to realize this integration in the clinical routine. Although there are still many challenges to comprehensively implement automated and integrated data analysis in molecular cancer imaging, the authors conclude that we are entering a new era of medical diagnostics and precision medicine.

Confounding factors in breast magnetic resonance fingerprinting: B1+ , slice profile, and diffusion effects.

PURPOSE: Magnetic resonance fingerprinting (MRF) offers rapid quantitative imaging but may be subject to confounding effects (CE) if these are not included in the model-based reconstruction. This study characterizes the influence of in-plane B1+ , slice profile and diffusion effects on T1 and T2 estimation in the female breast at 1.5T. METHODS: Simulations were used to predict the influence of each CE on the accuracy of MRF and to investigate the influence of electronic noise and spiral aliasing artefacts. The experimentally observed bias in regions of fibroglandular tissue (FGT) and fatty tissue (FT) was analyzed for undersampled spiral breast MRF data of 6 healthy volunteers by performing MRF reconstruction with and without a CE. RESULTS: Theoretic analysis predicts T1 under-/T2 overestimation if the nominal flip angles are underestimated and inversely, T1 under-/T2 overestimation if omitting slice profile correction, and T1 under-/T2 underestimation if omitting diffusion in the signal model. Averaged over repeated signal simulations, including spiral aliasing artefacts affected precision more than accuracy. Strong in-plane B1+ effects occurred in vivo, causing T2 left-right inhomogeneity between both breasts. Their correction decreased the T2 difference from 29 to 5 ms in FGT and from 29 to 9 ms in FT. Slice profile correction affected FGT T2 most strongly, resulting in -22% smaller values. For the employed spoiler gradient strengths, diffusion did not affect the parameter maps, corresponding well with theoretic predictions. CONCLUSION: Understanding CEs and their relative significance for an MRF sequence is important when defining an MRF signal model for accurate parameter mapping.

Spiral blurring correction with water-fat separation for magnetic resonance fingerprinting in the breast.

PURPOSE: Magnetic resonance fingerprinting (MRF) with spiral readout enables rapid quantification of tissue relaxation times. However, it is prone to blurring because of off-resonance effects. Hence, fat blurring into adjacent regions might prevent identification of small tumors by their quantitative T1 and T2 values. This study aims to correct for the blurring artifacts, thereby enabling fast quantitative mapping in the female breast. METHODS: The impact of fat blurring on spiral MRF results was first assessed by simulations. Then, MRF was combined with 3-point Dixon water-fat separation and spiral blurring correction based on conjugate phase reconstruction. The approach was assessed in phantom experiments and compared to Cartesian reference measurements, namely inversion recovery (IR), multi-echo spin echo (MESE), and Cartesian MRF, by normalized root-mean-square error (NRMSE) and SD calculations. Feasibility is further demonstrated in vivo for quantitative breast measurements of 6 healthy female volunteers, age range 24-31 y. RESULTS: In the phantom experiment, the blurring correction reduced the NRMSE per phantom vial on average from 16% to 8% for T1 and from 18% to 11% for T2 when comparing spiral MRF to IR/MESE sequences. When comparing to Cartesian MRF, the NRMSE reduced from 15% to 8% for T1 and from 12% to 7% for T2 . Furthermore, SDs decreased. In vivo, the blurring correction removed fat bias on T1 /T2 from a rim of ~7-8 mm width adjacent to fatty structures. CONCLUSION: The blurring correction for spiral MRF yields improved quantitative maps in the presence of water and fat.

(Hybrid) SPECT and PET Technologies.

SPECT and PET are nuclear tomographic imaging modalities that visualize functional information based on the accumulation of radioactive tracer molecules. However, SPECT and PET lack anatomical information, which has motivated their combination with an anatomical imaging modality such as CT or MRI. This chapter begins with an overview over the fundamental physics of SPECT and PET followed by a presentation of the respective detector technologies, including detection requirements, principles and different detector concepts. The reader is subsequently provided with an introduction into hybrid imaging concepts, before a dedicated section presents the challenges that arise when hybridizing SPECT or PET with MRI, namely, mutual distortions of the different electromagnetic fields in MRI on the nuclear imaging system and vice versa. The chapter closes with an overview about current hybrid imaging systems of both clinical and preclinical kind. Finally, future developments in hybrid SPECT and PET technology are discussed.

Size-isolation of superparamagnetic iron oxide nanoparticles improves MRI, MPI and hyperthermia performance.

Superparamagnetic iron oxide nanoparticles (SPION) are extensively used for magnetic resonance imaging (MRI) and magnetic particle imaging (MPI), as well as for magnetic fluid hyperthermia (MFH). We here describe a sequential centrifugation protocol to obtain SPION with well-defined sizes from a polydisperse SPION starting formulation, synthesized using the routinely employed co-precipitation technique. Transmission electron microscopy, dynamic light scattering and nanoparticle tracking analyses show that the SPION fractions obtained upon size-isolation are well-defined and almost monodisperse. MRI, MPI and MFH analyses demonstrate improved imaging and hyperthermia performance for size-isolated SPION as compared to the polydisperse starting mixture, as well as to commercial and clinically used iron oxide nanoparticle formulations, such as Resovist® and Sinerem®. The size-isolation protocol presented here may help to identify SPION with optimal properties for diagnostic, therapeutic and theranostic applications.

Flexible and modular MPI simulation framework and its use in modelling a µMPI.

The availability of thorough system simulations for detailed and accurate performance prediction and optimization of existing and future designs for a new modality such as magnetic particle imaging (MPI) are very important. Our framework aims to simulate a complete MPI system by providing a description of all (drive and receive) coils, permanent magnet configurations, magnetic nanoparticle (MNP) distributions, and characteristics of the signal processing chain. The simulation is performed on a user defined spatial and temporal discrete grid. The magnetization of the MNP is modelled by either the Langevin theory or as ideal particles with infinite steepness and ideal saturation. The magnetic fields are approximated in first order by calculating the Biot-Savart integral. Additionally the coupling constants between the excitation coils (e.g. drive field coils) and the receive coils can be determined. All coils can be described by an XML description language based on primitive geometric shapes. First simulations of a modelled µMPI system are shown. In this regard µMPI refers to a small one dimensional system for samples of a size of a few tens of a cubic millimeter and a spatial resolution of about 200 µm.

A FIELD CANCELATION SIGNAL EXTRACTION METHOD FOR MAGNETIC PARTICLE IMAGING.

In nowadays Magnetic Particle Imaging (MPI) signal detection and excitation happens at the same time. This concept, however, leads to strong coupling of the drive (excitation) field (DF) with the receive chain. As the induced DF signal is several orders of magnitude higher, special measures have to be taken to suppress this signal portion within the receive signal to keep the required dynamic range of the subsequent analog to digital conversion in a technically feasible range. For "frequency space MPI" high-order band-stop-filters have been successfully used to remove the DF signals, which unfortunately as well removes the fundamental harmonic components of the signal of the magnetic nanoparticles (MNP). According to the Langevin theory the fundamental harmonic component has a large signal contribution and is important for direct reconstruction of the particle concentration. In order to separate the fundamental harmonic component of the MNP from the induced DF signal, different concepts have been proposed using signal cancelation based on additional DF signals, also in combination with additional filtering. In this paper, we propose a field-cancelation (FC) concept in which a receive coil (RC) consists of a series connection of a primary coil in combination with an additional cancelation coil. The geometry of the primary coil was chosen to be sensitive for the MNP signal while the cancelation coil was chosen to minimize the overall inductive coupling of the FC-RC with the DF. Sensitivity plots and mutual coupling coefficients were calculated using a thin-wire approximation. A prototype FC-RC was manufactured and effectiveness of the reduction of the mutual inductive coupling (d) was tested in an existing mouse MPI scanner. The difference between simulations (ds =70 dB) and the measurements (dms =55 dB) indicated the feasibility as well as the need for further investigations.

Timing limits of ultrafast cross-luminescence emission in CsZnCl-based crystals for TOF-CT and TOF-PET.

BACKGROUND: Good timing resolution in medical imaging applications such as TOF-CT or TOF-PET can boost image quality or patient comfort significantly by reducing the influence of background noise. However, the timing resolution of state-of-the-art detectors in CT and PET are limited by their light emission process. Core-valence cross-luminescence is an alternative, but well-known compounds (e.g. BaF2) pose several problems for medical imaging applications, such as their emission wavelength in the deep UV. CsZnCl-based materials show promise to solve this issue, as they provide fast decay times of 1-2 ns and an emission wavelength around 300 nm. RESULTS: In this work, we investigated two CsZnCl-compounds: Cs2ZnCl4 and Cs3ZnCl5. We validated the previously published decay times on a time-correlated single-photon counting setup with 1.786 ± 0.016 ns for Cs2ZnCl4 and 1.034 ± 0.013 ns for Cs3ZnCl5. The setup's high resolution enabled the discovery of an additional prompt emission component with a significant abundance of 98 ± 18 (Cs2ZnCl4) and 86 ± 14 (Cs3ZnCl5) photons/MeV energy deposit. In a PET coincidence experiment, we measured the best coincidence time resolution (CTR) of 62 ps (FWHM) for Cs2ZnCL4 coupled to FBK VUV SiPMs with silicon oil. To assess the CTR for lower energies, we filtered the energy along the Compton continuum and found a deteriorated CTR that seems to be mainly influenced by photon statistics. Furthermore, this study gave us a rough estimate of e.g. 150 ps (FWHM) CTR at 100 keV energy for Cs2ZnCL4. From measurements with high activity of 14 MBq to check for pile-up effects we assume that Cs2ZnCl4 is better suited for high-rate time-of-flight applications than lutetium-based oxides. Simulations demonstrated that the stopping power of Cs2ZnCl4 is lower than for LSO:Ce,Ca, meaning that a high amount of material would be needed for TOF-PET applications. However, the stopping power seems acceptable for applications in TOF-CT. CONCLUSIONS: The fast decay time, state-of-the-art CTR in benchtop experiments and high-rate suitability make CsZnCl materials a promising candidate for time-of-flight experiments. We consider especially TOF-CT a suitable application due to its relatively low X-ray energies (~ 100 keV) and the thusly acceptable stopping power of Cs2ZnCl4. Currently, further exploration of the prompt emission and its creation mechanism is planned, as well as investigating the light transport of Cs2ZnCl4 in longer crystals.

Holistic evaluation of a machine learning-based timing calibration for PET detectors under varying data sparsity.

Objective.Modern PET scanners offer precise TOF information, improving the SNR of the reconstructed images. Timing calibrations are performed to reduce the worsening effects of the system components and provide valuable TOF information. Traditional calibration procedures often provide static or linear corrections, with the drawback that higher-order skews or event-to-event corrections are not addressed. Novel research demonstrated significant improvements in the reachable timing resolutions when combining conventional calibration approaches with machine learning, with the disadvantage of extensive calibration times infeasible for a clinical application. In this work, we made the first steps towards an in-system application and analyzed the effects of varying data sparsity on a machine learning timing calibration, aiming to accelerate the calibration time. Furthermore, we demonstrated the versatility of our calibration concept by applying the procedure for the first time to analog readout technology.Approach.We modified experimentally acquired calibration data used for training regarding their statistical and spatial sparsity, mimicking reduced measurement time and variability of the training data. Trained models were tested on unseen test data, characterized by fine spatial sampling and rich statistics. In total, 80 decision tree models with the same hyperparameter settings, were trained and holistically evaluated regarding data scientific, physics-based, and PET-based quality criteria.Main results.The calibration procedure can be heavily reduced from several days to some minutes without sacrificing quality and still significantly improving the timing resolution from(304±5)psto(216±1)pscompared to conventionally used analytical calibration methods.Significance.This work serves as the first step in making the developed machine learning-based calibration suitable for an in-system application to profit from the method's capabilities on the system level. Furthermore, this work demonstrates the functionality of the methodology on detectors using analog readout technology. The proposed holistic evaluation criteria here serve as a guideline for future evaluations of machine learning-based calibration approaches.

A finely segmented semi-monolithic detector tailored for high-resolution PET.

BACKGROUND: Preclinical research and organ-dedicated applications use and require high (spatial-)resolution positron emission tomography (PET) detectors to visualize small structures (early) and understand biological processes at a finer level of detail. Researchers seeking to improve detector and image spatial resolution have explored various detector designs. Current commercial high-resolution systems often employ finely pixelated or monolithic scintillators, each with its limitations. PURPOSE: We present a semi-monolithic detector, tailored for high-resolution PET applications with a spatial resolution in the range of 1 mm or better, merging concepts of monolithic and pixelated crystals. The detector features LYSO slabs measuring (24 × 10 × 1) mm3, coupled to a 12 × 12 readout channel photosensor with 4 mm pitch. The slabs are grouped in two arrays of 44 slabs each to achieve a higher optical photon density despite the fine segmentation. METHODS: We employ a fan beam collimator for fast calibration to train machine-learning-based positioning models for all three dimensions, including slab identification and depth-of-interaction (DOI), utilizing gradient tree boosting (GTB). The data for all dimensions was acquired in less than 2 h. Energy calculation was based on a position-dependent energy calibration. Using an analytical timing calibration, time skews were corrected for coincidence timing resolution (CTR) estimation. RESULTS: Leveraging machine-learning-based calibration in all three dimensions, we achieved high detector spatial resolution: down to 1.18 mm full width at half maximum (FWHM) detector spatial resolution and 0.75 mm mean absolute error (MAE) in the planar-monolithic direction, and 2.14 mm FWHM and 1.03 mm MAE for DOI at an energy window of (435-585) keV. Correct slab interaction identification in planar-segmented direction exceeded 80%, alongside an energy resolution of 12.7% and a CTR of 450 ps FWHM. CONCLUSIONS: The introduced finely segmented, high-resolution slab detector demonstrates appealing performance characteristics suitable for high-resolution PET applications. The current benchtop-based detector calibration routine allows these detectors to be used in PET systems.

Pathophysiologic Mapping of Chronic Liver Diseases With Longitudinal Multiparametric MRI in Animal Models.

OBJECTIVES: Chronic liver diseases (CLDs) have diverse etiologies. To better classify CLDs, we explored the ability of longitudinal multiparametric MRI (magnetic resonance imaging) in depicting alterations in liver morphology, inflammation, and hepatocyte and macrophage activity in murine high-fat diet (HFD)- and carbon tetrachloride (CCl4)-induced CLD models. MATERIALS AND METHODS: Mice were either untreated, fed an HFD for 24 weeks, or injected with CCl4 for 8 weeks. Longitudinal multiparametric MRI was performed every 4 weeks using a 7 T MRI scanner, including T1/T2 relaxometry, morphological T1/T2-weighted imaging, and fat-selective imaging. Diffusion-weighted imaging was applied to assess fibrotic remodeling and T1-weighted and T2*-weighted dynamic contrast-enhanced MRI and dynamic susceptibility contrast MRI using gadoxetic acid and ferucarbotran to target hepatocytes and the mononuclear phagocyte system, respectively. Imaging data were associated with histopathological and serological analyses. Principal component analysis and clustering were used to reveal underlying disease patterns. RESULTS: The MRI parameters significantly correlated with histologically confirmed steatosis, fibrosis, and liver damage, with varying importance. No single MRI parameter exclusively correlated with 1 pathophysiological feature, underscoring the necessity for using parameter patterns. Clustering revealed early-stage, model-specific patterns. Although the HFD model exhibited pronounced liver fat content and fibrosis, the CCl4 model indicated reduced liver fat content and impaired hepatocyte and macrophage function. In both models, MRI biomarkers of inflammation were elevated. CONCLUSIONS: Multiparametric MRI patterns can be assigned to pathophysiological processes and used for murine CLD classification and progression tracking. These MRI biomarker patterns can directly be explored clinically to improve early CLD detection and differentiation and to refine treatments.

Overcoming data scarcity in biomedical imaging with a foundational multi-task model.

Foundational models, pretrained on a large scale, have demonstrated substantial success across non-medical domains. However, training these models typically requires large, comprehensive datasets, which contrasts with the smaller and more specialized datasets common in biomedical imaging. Here we propose a multi-task learning strategy that decouples the number of training tasks from memory requirements. We trained a universal biomedical pretrained model (UMedPT) on a multi-task database including tomographic, microscopic and X-ray images, with various labeling strategies such as classification, segmentation and object detection. The UMedPT foundational model outperformed ImageNet pretraining and previous state-of-the-art models. For classification tasks related to the pretraining database, it maintained its performance with only 1% of the original training data and without fine-tuning. For out-of-domain tasks it required only 50% of the original training data. In an external independent validation, imaging features extracted using UMedPT proved to set a new standard for cross-center transferability.

In vitro and in vivo evaluation of biohybrid tissue-engineered vascular grafts with transformative 1H/19F MRI traceable scaffolds.

Biohybrid tissue-engineered vascular grafts (TEVGs) promise long-term durability due to their ability to adapt to hosts' needs. However, the latter calls for sensitive non-invasive imaging approaches to longitudinally monitor their functionality, integrity, and positioning. Here, we present an imaging approach comprising the labeling of non-degradable and degradable TEVGs' components for their in vitro and in vivo monitoring by hybrid 1H/19F MRI. TEVGs (inner diameter 1.5 mm) consisted of biodegradable poly(lactic-co-glycolic acid) (PLGA) fibers passively incorporating superparamagnetic iron oxide nanoparticles (SPIONs), non-degradable polyvinylidene fluoride scaffolds labeled with highly fluorinated thermoplastic polyurethane (19F-TPU) fibers, a smooth muscle cells containing fibrin blend, and endothelial cells. 1H/19F MRI of TEVGs in bioreactors, and after subcutaneous and infrarenal implantation in rats, revealed that PLGA degradation could be faithfully monitored by the decreasing SPIONs signal. The 19F signal of 19F-TPU remained constant over weeks. PLGA degradation was compensated by cells' collagen and α-smooth-muscle-actin deposition. Interestingly, only TEVGs implanted on the abdominal aorta contained elastin. XTT and histology proved that our imaging markers did not influence extracellular matrix deposition and host immune reaction. This concept of non-invasive longitudinal assessment of cardiovascular implants using 1H/19F MRI might be applicable to various biohybrid tissue-engineered implants, facilitating their clinical translation.

Histopathological biomarkers for predicting the tumour accumulation of nanomedicines.

The clinical prospects of cancer nanomedicines depend on effective patient stratification. Here we report the identification of predictive biomarkers of the accumulation of nanomedicines in tumour tissue. By using supervised machine learning on data of the accumulation of nanomedicines in tumour models in mice, we identified the densities of blood vessels and of tumour-associated macrophages as key predictive features. On the basis of these two features, we derived a biomarker score correlating with the concentration of liposomal doxorubicin in tumours and validated it in three syngeneic tumour models in immunocompetent mice and in four cell-line-derived and six patient-derived tumour xenografts in mice. The score effectively discriminated tumours according to the accumulation of nanomedicines (high versus low), with an area under the receiver operating characteristic curve of 0.91. Histopathological assessment of 30 tumour specimens from patients and of 28 corresponding primary tumour biopsies confirmed the score's effectiveness in predicting the tumour accumulation of liposomal doxorubicin. Biomarkers of the tumour accumulation of nanomedicines may aid the stratification of patients in clinical trials of cancer nanomedicines.

Challenges and current methods for attenuation correction in PET/MR.

Quantitative PET imaging requires an attenuation map to correct for attenuation. In stand-alone PET or PET/CT, the attenuation map is usually derived from a transmission scan or CT image, respectively. In PET/MR, these methods will most likely not be used. Therefore, attenuation correction has long been regarded as one of the major challenges in the development of PET/MR. In the past few years, much progress has been made in this field. In this review, the challenges faced in attenuation correction for PET/MR are discussed. Different methods have been proposed to overcome these challenges. An overview of the MR-based (template-based and voxel-based), transmission-based and emission-based methods and the results that have been obtained is provided. Although several methods show promising results, no single method fulfils all of the requirements for the ideal attenuation correction method for PET/MR. Therefore, more work is still necessary in this field. To allow implementation in routine clinical practice, extensive evaluation of the proposed methods is necessary to demonstrate robustness and automation.