Multi-attention-based brain tumor classification network: an advanced framework for accurate diagnosis across multiple imaging modalities

Left ventricular (LV) pseudoaneurysm is a rare complication that occurs shortly after myocardial infarction or previous cardiac surgery. We present here a case of an elderly patient with an LV pseudoaneurysm that developed over a long period of time after Dor operation. An 83-year-old man with rapid LV enlargement 18 years after Dor operation presented with dyspnea. We diagnosed his mass as an LV pseudoaneurysm using multiple non-invasive imaging modalities. The suture between the patch and the LV myocardium became fragile and detached over time. However, due to adhesion in the thorax and slow blood flow, the enlargement stopped. Surgery was considered, but due to his age and renal dysfunction, he did not undergo surgery. Conservative treatment with antihypertensive and heart failure therapy was continued. Clinical manifestations of LV pseudoaneurysm vary from asymptomatic to signs of heart failure and even sudden death. Surgery is recommended for LV pseudoaneurysm because of the high risk of rupture, but conservative treatment may be an option that has been developed long after surgery in the era of an increasing number of elderly OMI or post-cardiac surgery patients. Proper diagnosis using multiple imaging modalities may lead to the avoidance of this devastating complication.

X-ray computed tomography (CT) imaging plays an essential role in disease diagnosis due to its noninvasive, painless mode and superior penetration depth. However, the resolution of the soft tissue ...

To elucidate the manifestations and associated complications observed in patients with intraocular silicone oil (SO) emulsification through multiple imaging modalities. This single-center, observational, retrospective study included 116 patients who underwent pars plana vitrectomy (PPV) with SO injection for retinal detachment (RD), followed by subsequent SO removal at the Second Hospital of Hebei Medical University from January 2013 to January 2023. Comprehensive records of ophthalmic examinations utilizing multiple imaging techniques were maintained. The study comprised 56 females and 60 males, with a mean age of 52.75±13.6y. The mean follow-up duration for SO tamponade was 9.04±11.33mo (range: 1-84mo). Among the participants, 59 patients were diagnosed with SO emulsification, while 57 patients were in the SO unemulsified group. Patients with SO emulsification had a significantly prolonged SO tamponade duration (P<0.01). Multiple imaging techniques revealed notable signs of SO emulsification and its complications, such as 4 cases (3.4%) with posterior corneal SO-like keratic precipitates (KP) observed by anterior segment photography, 23 cases (19.8%) exhibiting spherical high-reflection signals in the inter-retina, retinal pigment epithelium, or choroid detected by Spectralis spectral domain optical coherence tomography (SD-OCT), 4 cases (3.4%) showing slow movement of emulsified SO droplets within retinal vessels during fluorescein angiography (FFA), plain and enhanced head magnetic resonance imaging (MRI) images of these four patients did not detect emulsified SO in the lateral ventricles, suprasellar cistern, subarachnoid space, third ventricle, fourth ventricle, or other intracranial locations. Intraocular emulsified SO can lead to damage in both anterior and posterior segment tissues, encompassing corneal degeneration, cataracts, glaucoma, retinal and choroid inflammation. Objective multiple imaging techniques such as anterior segment photography, SD-OCT, FFA, and MRI offer comprehensive evaluation and diagnosis of SO emulsification and its associated complications.

The goal of glioma surgery is maximal safe resection in order to provide optimal tumor control and survival benefit to the patient. There are multiple imaging modalities beyond traditional contrast-enhanced magnetic resonance imaging (MRI) that have been incorporated into the preoperative workup of patients presenting with gliomas. The aim of these imaging modalities is to identify cortical and subcortical areas of eloquence, and their relationship to the lesion. In this article, multiple modalities are described with an emphasis on the underlying technology, clinical utilization, advantages, and disadvantages of each. functional MRI and its role in identifying hemispheric dominance and areas of language and motor are discussed. The nuances of magnetoencephalography and transcranial magnetic stimulation in localization of eloquent cortex are examined, as well as the role of diffusion tensor imaging in defining normal white matter tracts in glioma surgery. Lastly, we highlight the role of stimulated Raman spectroscopy in intraoperative histopathological diagnosis of tissue to guide tumor resection. Tumors may shift the normal arrangement of functional anatomy in the brain; thus, utilization of multiple modalities may be helpful in operative planning and patient counseling for successful surgery.

An image understanding system is discussed that will analyze cardiac dysfunction found by combining information from multiple, noninvasive imaging modality studies. The system is a model for intelligent multimodality image understanding in general, but is applied to a very specific problem: detection and rating of left ventricular (LV) aneurysms. Key features of the system are: 1) its ability to handle and quantify uncertain or partial image-derived information in a concise way using probabilistic evidential reasoning, 2) its ability to fuse pertient relative information from independent diagnostic images of the same patient to achieve an algorithm-assembled, consensus, quantitative opinion of cardiac shape and motion, and finally 3) to arrive at a decision level set of numbers that quantify and localize left ventricular aneurysm formation for each patient. The availability of data such as described in 3) will enable more precise prognostic or diagnostic risk classification for patients, making therapy alternatives more rational. Because of the subjective probabilistic reasoning strategy (based on the principle of maximum entropy) the final quantitative results will carry not only the system’s assessment of a particular patient’s heart, but also the degree of confidence that the automated analysis system has in the result it presents. This confidence is increased when similar LV motion and shape is perceived by the multiple imaging modalities.

Prostate gland segmentation is a critical step in prostate radiotherapy planning, where dose plans are typically formulated on CT. Pretreatment MRI is now beginning to be acquired at several medical centers. Delineation of the prostate on MRI is acknowledged as being significantly simpler to perform, compared to delineation on CT. In this work, the authors present a novel framework for building a linked statistical shape model (LSSM), a statistical shape model (SSM) that links the shape variation of a structure of interest (SOI) across multiple imaging modalities. This framework is particularly relevant in scenarios where accurate boundary delineations of the SOI on one of the modalities may not be readily available, or difficult to obtain, for training a SSM. In this work the authors apply the LSSM in the context of multimodal prostate segmentation for radiotherapy planning, where the prostate is concurrently segmented on MRI and CT. The framework comprises a number of logically connected steps. The first step utilizes multimodal registration of MRI and CT to map 2D boundary delineations of the prostate from MRI onto corresponding CT images, for a set of training studies. Hence, the scheme obviates the need for expert delineations of the gland on CT for explicitly constructing a SSM for prostate segmentation on CT. The delineations of the prostate gland on MRI and CT allows for 3D reconstruction of the prostate shape which facilitates the building of the LSSM. In order to perform concurrent prostate MRI and CT segmentation using the LSSM, the authors employ a region-based level set approach where the authors deform the evolving prostate boundary to simultaneously fit to MRI and CT images in which voxels are classified to be either part of the prostate or outside the prostate. The classification is facilitated by using a combination of MRI-CT probabilistic spatial atlases and a random forest classifier, driven by gradient and Haar features. The authors acquire a total of 20 MRI-CT patient studies and use the leave-one-out strategy to train and evaluate four different LSSMs. First, a fusion-based LSSM (fLSSM) is built using expert ground truth delineations of the prostate on MRI alone, where the ground truth for the gland on CT is obtained via coregistration of the corresponding MRI and CT slices. The authors compare the fLSSM against another LSSM (xLSSM), where expert delineations of the gland on both MRI and CT are employed in the model building; xLSSM representing the idealized LSSM. The authors also compare the fLSSM against an exclusive CT-based SSM (ctSSM), built from expert delineations of the gland on CT alone. In addition, two LSSMs trained using trainee delineations (tLSSM) on CT are compared with the fLSSM. The results indicate that the xLSSM, tLSSMs, and the fLSSM perform equivalently, all of them out-performing the ctSSM. The fLSSM provides an accurate alternative to SSMs that require careful expert delineations of the SOI that may be difficult or laborious to obtain. Additionally, the fLSSM has the added benefit of providing concurrent segmentations of the SOI on multiple imaging modalities.

RADIOGRAPHIC FINDINGS OF CHRONIC THROMBOEMBOLIC PULMONARY HYPERTENSION

Segmentation of fused MR and CT images using DL-CNN with PGK and NLEM filtered AACGK-FCM

Objective: The objective of this narrative review was to summarize and critically appraise current clinical applications of artificial intelligence (AI) in cardiology, with particular emphasis on arrhythmia detection and cardiovascular imaging modalities. Methods: A selective narrative review of peer-reviewed literature published between 2015 and 2025 was performed using PubMed, Scopus, and Web of Science. Studies were selected based on clinical relevance, validation methodology, and reported diagnostic or prognostic performance in human populations. Results: AI-based algorithms demonstrate diagnostic performance comparable to expert interpretation in atrial fibrillation (AF) detection as well as across echocardiography, coronary computed tomography angiography (CCTA), cardiac magnetic resonance imaging (CMR), and intravascular optical coherence tomography (OCT). Beyond diagnostic accuracy, AI has been shown to reduce analysis time and interobserver variability in controlled and retrospective study settings. However, the majority of available studies are retrospective, rely on curated datasets, and lack prospective validation demonstrating a direct impact on clinical outcomes. Conclusions: AI constitutes a valuable decision-support tool in contemporary cardiology, enhancing diagnostic efficiency and risk stratification across multiple imaging and electrophysiological modalities. Nevertheless, broader clinical adoption will require rigorous external validation, improved model interpretability, and prospective outcome-driven studies. This review uniquely integrates evidence across electrophysiology and multiple imaging modalities to identify shared translational challenges and near-term clinical opportunities for AI in cardiology.

Multimodality medical image fusion is the process of combining multiple images from single or multiple modalities of imaging. Medical image fusion methods are adopted to increase the quality of medical images by attaining the salient features in the fusion results. Hence, they raise the clinical applicability of medical images for appraisal and diagnosis problems. This purpose is achieved by capturing the complementary information presented in two or more images of different modalities in the fusion result. Medical image fusion is generally concerned with Magnetic Resonance Imaging (MRI), Magnetic Resonance Angiogram (MRA), Positron Emission Tomography (PET), Structural Positron Emission Tomography (SPET), Computerized Tomography (CT), and Single-Photon Emission Computed Tomography (SPECT) modalities. Each modality has its merits and drawbacks. This induces new fusion methods for merging information from multiple imaging modalities. Researchers have presented several methods for medical image fusion, and these methods achieved good results. However, medical image fusion is a resurgent field that needs to be enhanced to conquer the increasing challenges. This paper presents a comprehensive survey of some existing medical image fusion methods. It is expected that this study will be useful for the researchers scrutinizing medical image fusion. Furthermore, it is expected to establish a concrete foundation for developing more powerful fusion methods for medical applications.

Introduction: The goal of facial transplantation is to restore form and function to patients with devastating or disfiguring injuries or defects of the face. To date, 20 face transplants have been performed around the world. Aesthetic and functional outcomes are difficult to optimize in these three-dimensionally complex procedures. Clearly defining and understanding the complex tissue deficits and defects that accompany devastating facial injuries like electric burns, blast wounds, and accidental trauma is critical for both technical success as well as to objectively analyze return of function after face transplantation. Current state of the art imaging in face transplantation includes multimodality imaging (3D CT, MRI, angio and plain radiography) for surgical planning and 3D plaster or plastic modeling with stereolithography. Conventional imaging modalities, however, are on separate software platforms that are not compatible with real-time user interaction and modification. Our goal was to develop a novel technique of integrating multiple sophisticated modalities ranging from MRI to 3D CT to tractography into a single 3D representation that enables imaging of skeletal, soft tissue, and neurovascular structures and facilitates planning and assessing outcomes of human face transplantation Methods: The craniofacial skeleton is represented as a polygonal model generated from thresholding and “stacking” dicom images from CT scans so it may serve as a framework. A skin model is generated by the same approach. Alternatively, surface scanning with photo-realistic texturing may be employed to generate a skin mesh. Muscles can be extracted from the same dicom dataset as the bone data as follows (if CT fails to capture muscle data in sufficient detail, MRI data sets may be utilized). Key slices are manually segmented by outlining the structures of interest in 2D and lofting between 2D planes to build 3D meshes. If data quality permits, blood vessels relevant to the model can be extracted by thresholding dicoms from a CT angiogram. If the dataset is not of sufficient quality for this approach, key slices can be imported into a 3D package and models manually segmented as above. Nerves are modeled as non-uniform rational basis splines based on tractography data. Results: Intuitive, multiplanar, volume rendered data of 3D relational anatomy (skin, muscle, vessel, nerve and bone) from once-disparate and unwieldy CT, surface scans, CT angiography, MRI and tractography data have been integrated to develop detailed 3D anatomical polygonal meshes compatible with real-time end-user manipulation and modification. Conclusions: For the first time, we have devised a technique that fuses distinct donor and recipient imaging data derived from multiple conventional and cutting edge imaging modalites into a single 3D representation that is compatible with seamless user interaction and visualization. Such 3-D modeling offers critical insight into surface topography and relational anatomy between different facial tissues in donors and recipients. Procedural planning may be enhanced in patients by allowing virtual preoperative interaction with skeletal, soft tissue, and neurovascular anatomy. Combining information from multiple imaging modalities could also optimize patient selection and sequential monitoring of functional outcomes after face transplantation.

Combat-related thoracic trauma (CRTT) is a significant contributor to morbidity and mortality of the casualties from Operation Enduring Freedom (OEF) and Operation Iraqi Freedom (OIF). Penetrating, blunt, and blast injuries are the most common mechanisms of trauma to the chest. Imaging plays a key role in the battlefield management of CRTT casualties. This work discusses the imaging manifestations of thoracic injuries from blunt trauma and blast injury, emphasizing epidemiology and diagnostic clues seen during OEF and OIF. The assessment of radiologic findings in patients who suffer from combat-related blunt thoracic trauma and blast injury is the basis of this work. The imaging modalities for this work include multi-detector computed tomography (MDCT) and chest radiography. Multiple imaging modalities are available to imagers on or near the battlefront, including radiography, fluoroscopy, and MDCT. MDCT with multi-planar reconstructions is the most sensitive imaging modality available in combat hospitals for the evaluation of CRTT. In modern combat, blunt and blast injuries account for a significant portion of CRTT. Individual body armor converts penetrating trauma to blunt trauma, leading to pulmonary contusion that accounted for 50.2% of thoracic injuries during OIF and OEF. Flail chest, a subset of blunt chest injury, is caused by significant blunt force to the chest and occurs four times as frequently in combat casualties when compared with the civilian population. Imaging features of CRTT have significant diagnostic and prognostic value. Pulmonary contusions on chest radiography appear as patchy consolidations in the acute setting with ill-defined and non-segmental borders. MDCT of the chest is a superior imaging modality in diagnosing and evaluating pulmonary contusion. Contusions on MDCT appear as crescentic ground-glass opacities (opacities through which lung interstitium and vasculature are still visible) and areas of consolidation that often do not respect the anatomic boundaries of the affected lobes. Additionally, small pulmonary contusions may exhibit sub-pleural sparing and may distinguish contusion from pneumonia or other lung pathology. Although pulmonary laceration is typically the result of penetrating trauma, laceration may also be caused by displaced rib fractures or significant shearing forces on the lung without penetrating injury. Because of elastic recoil of the normal pulmonary parenchyma surrounding the injury, pulmonary lacerations may present as late as 48-72 h after injury. Pulmonary lacerations may appear similar to pulmonary contusions on chest radiography initially and will require MDCT for definitive diagnosis. Blast injury is a defining injury of modern combat. Blast lung injury is initially diagnosed with chest radiography, where the pattern of lung opacities has previously been described by clinicians as "batwing" or "butterfly" because of its central appearance in the lung. "Peribronchovascular" may be a more accurate description of primary blast lung based on its appearance on MDCT. This pattern may differentiate primary blast lung injury from other causes of thoracic trauma. CRTT continues to be a significant contributor to the morbidity and mortality of those injured during OEF and OIF. The distinct injury patterns and atypical imaging manifestations of blunt trauma and blast lung injury are important to recognize early because of the acuity of this patient population and the influence of accurate diagnosis on clinical management.

Conventional biomedical imaging modalities in wide clinical use, such as ultrasound imaging, X-ray computed tomography, magnetic resonance imaging, and positron emission tomography, can provide morphological, anatomical, and functional information about biological tissues. However, single mode imaging in conventional medicine provides only limited information for definitive diagnoses. Thus, combinational diagnosis using multiple imaging modalities has become increasingly important. Recently, photoacoustic imaging (PAI) has gained significant attention, and several PAI prototypes have been used in clinical trials. At the same time, PAI has been tested in combination with conventional imaging modalities. For all these imaging modalities, various contrast-enhancing agents have been developed for various purposes. In this review article, we will focus on recent progress in developing dual mode contrast agents for PAI in combination with other conventional imaging modalities.

Insights into brain development from neurogenetic syndromes: evidence from fragile X syndrome, Williams syndrome, Turner syndrome and velocardiofacial syndrome

Acquired aorta-pulmonary artery fistula: Diagnosis by multiple imaging modalities