CT Fusion Research Articles

MRI imaging for planning stereotactic biopsies of the brain lesions

Aim of the study . To evaluate the influence of modern MRI imaging, CT- and MRI fusion in STB planning on effectiveness of morphological verification and risks of intraoperative complications. Materials and methods . The most common indications for brain tumor biopsy is definitive diagnosis of the intracranial lesion and differentiation of the neoplastic and non-neoplastic pathology. During 2019 170 patients (95 men and 75 femmes) 7–69 years old underwent stereotactic biopsy an Integra LifeSciences Corporation Cosman–Roberts–Wells® (CRW®) system. In 80 cases, we used CT and MRI fusion by NeuroSight program to select the target point, entry point and trajectory of the brain biopsy. In 90 cases we use only CT images for stereotactic brain biopsy planning. Results . Among 80 patients with use of combined CT-MRI fusion only one case was not histologically verified. In 90 patients with use of only CT-imaging for STB 3 cases were histologically not verified. There were no hemorrhagic complications among 80 patients in CT-MRI fusion group. In 3 cases intracranial hematomas were identified that required surgical treatment after STB with use of CT-imaging for STB panning. Conclusion . The use of modern MRI and PET-CT imaging for STB planning increases its informativity and reduces the probability of hemorrhagic complications. Stereotactic biopsies remain a safe and reliable method for obtaining histological material. The use of modern imaging methods in biopsy planning increases their accuracy and reduces possible complications.

The impact of needle location on clinical outcome of radiofrequency rhizotomy for trigeminal neuralgia.

Radiofrequency thermocoagulation trigeminal rhizotomy (RT-TR) through the foramen ovale is a minimally invasive treatment for trigeminal neuralgia. Navigation of magnetic resonance imaging (MRI) and CT fusion imaging is a well-established method for cannulation of the Gasserian ganglion. In this study, we use the inline measurements from fusion image to analyze the anatomical parameters between the actual and simulation trajectories and compare the short- and intermediate-term outcomes according to determinable factors. The study included thirty-six idiopathic neuralgia patients who had undergone RT-TR with MRI and CT fusion image as a primary modality or repeated procedures. Among thirty-six treated patients, the inline length of the trigeminal cistern was longer for the simulated trajectory (8.4 ± 2.4 versus 6.5 ± 2.8 mm; p < 0.05), and the predominant structure at risk extrapolated from the inline trajectory was the brainstem, which signified a more medially directed route, in contrast with the equal weighting of temporal lobe and brainstem for the actual trajectory. The preoperative visual analogue scale (VAS) was 9.3 ± 1.0, which decreased to 2.5 ± 2.6 and 2.9 ± 3.1 at first (mean, 3 months) and second (mean, 14 months) postoperative follow-up, respectively. The postoperative VAS scores at the two follow-ups were not statistically significant without a covariate analysis. After adjustment for covariate risk factors, the second follow-up sustained therapeutic benefit was evident in patients with no prior history of related treatment, an ablation temperature greater than 70 °C, and needle location within or adjacent to the trigeminal cistern. This preliminary study demonstrated that the needle location between cistern and ganglion also plays a significant role in better intermediate-term results.

A novel variational optimization model for medical CT and MR image fusion

A novel variational optimization model for medical CT and MR image fusion

CT and MRI Medical Image Fusion Using Noise-Removal and Contrast Enhancement Scheme with Convolutional Neural Network.

Medical image fusion (MIF) has received painstaking attention due to its diverse medical applications in response to accurately diagnosing clinical images. Numerous MIF methods have been proposed to date, but the fused image suffers from poor contrast, non-uniform illumination, noise presence, and improper fusion strategies, resulting in an inadequate sparse representation of significant features. This paper proposes the morphological preprocessing method to address the non-uniform illumination and noise by the bottom-hat–top-hat strategy. Then, grey-principal component analysis (grey-PCA) is used to transform RGB images into gray images that can preserve detailed features. After that, the local shift-invariant shearlet transform (LSIST) method decomposes the images into the low-pass (LP) and high-pass (HP) sub-bands, efficiently restoring all significant characteristics in various scales and directions. The HP sub-bands are fed to two branches of the Siamese convolutional neural network (CNN) by process of feature detection, initial segmentation, and consistency verification to effectively capture smooth edges, and textures. While the LP sub-bands are fused by employing local energy fusion using the averaging and selection mode to restore the energy information. The proposed method is validated by subjective and objective quality assessments. The subjective evaluation is conducted by a user case study in which twelve field specialists verified the superiority of the proposed method based on precise details, image contrast, noise in the fused image, and no loss of information. The supremacy of the proposed method is further justified by obtaining 0.6836 to 0.8794, 0.5234 to 0.6710, and 3.8501 to 8.7937 gain for , CRR, and AG and noise reduction from 0.3397 to 0.1209 over other methods for objective parameters.

The Fusion of MRI and CT in the Planning of Brachytherapy for Cancer of the Uterine Cervix

Introduction: tumors of the uterine cervix are among the most common carcinomas in women. Intracervical brachytherapy is an indispensable part of curative treatment. Although the tumor is significantly more recognizable in MRI than in CT, the practical application of MRI in brachytherapy planning is still difficult. The present study examines the technical possibilities of merging CT and MRI. Materials and Methods: the treatment files and imaging of all 53 patients who had been irradiated by image-guided adaptive brachytherapy (IGABT) between January 2019 and August 2021 at the Department of Radiotherapy of the Hannover Medical School were evaluated, retrospectively. Patients were treated first with an external beam radiotherapy (EBRT) combined with simultaneous chemotherapy. After an average of 4.2 weeks, the preparation for IGABT began. The clinical target volume (CTV) for brachytherapy was contoured first in an MRI acquired before starting EBRT (MRI 1) and once more in a second MRI just before starting IGABT (MRI 2). Then, after inserting the intravaginal applicator, a CT-scan was acquired, and the CTV was contoured in the CT. Finally, the recordings of MRI 1, MRI 2, and the CT were merged, and the congruence of CTVs was quantitatively evaluated. Results: the CTV delineated in MRI 2 was, on average, 28% smaller than that in MRI 1 after an average applied radiation dose of 42 Gy. The CTV delineated in the CT covered an average of no more than 80.8% of the CTV delineated in MRI 2. The congruence of CTVs was not superior in patients with a smit sleeve in the cervical channel, with a 3D-volumetric MRI or with a contrast-enhanced sequence for MRI. Conclusion: the anatomical shape and position of the uterus is significantly changed by introducing a vaginal applicator. Despite the superior delimitability of the tumor in MRI, brachytherapy cannot be reliably planned by the image fusion of an MRI without a vaginal applicator.

Detection of Brain Tumor Using a Fusion of Ct, MRI and Svd

Detection of Brain Tumor Using a Fusion of Ct, MRI and Svd

An Efficient Medical Assistive Diagnostic Algorithm for Visualisation of Structural and Tissue Details in CT and MRI Fusion

An Efficient Medical Assistive Diagnostic Algorithm for Visualisation of Structural and Tissue Details in CT and MRI Fusion

Fusion of 3D lung CT and serum biomarkers for diagnosis of multiple pathological types on pulmonary nodules

Background and Objective: Current researches on pulmonary nodules mainly focused on the binary-classification of benign and malignant pulmonary nodules. However, in clinical applications, it is not enough to judge whether pulmonary nodules are benign or malignant. In this paper, we proposed a fusion model based on the Lung Information Dataset Containing 3D CT Images and Serum Biomarkers (LIDCCISB) we constructed to accurately diagnose the types of pulmonary nodules in squamous cell carcinoma, adenocarcinoma, inflammation and other benign diseases. Methods: Using single modal information of lung 3D CT images and single modal information of Lung Tumor Biomarkers (LTBs) in LIDCCISB, a Multi-resolution 3D Multi-classification deep learning model (Mr-Mc) and a Multi-Layer Perceptron machine learning model (MLP) were constructed for diagnosing multiple pathological types of pulmonary nodules, respectively. To comprehensively use the double modal information of CT images and LTBs, we used transfer learning to fuse Mr-Mc and MLP, and constructed a multimodal information fusion model that could classify multiple pathological types of benign and malignant pulmonary nodules. Results: Experiments showed that the constructed Mr-Mc model can achieve an average accuracy of 0.805 and MLP model can achieve an average accuracy of 0.887. The fusion model was verified on a dataset containing 64 samples, and achieved an average accuracy of 0.906. Conclusions: This is the first study to simultaneously use CT images and LTBs to diagnose multiple pathological types of benign and malignant pulmonary nodules, and experiments showed that our research was more advanced and more suitable for practical clinical applications.

CT and MRI image fusion algorithm based on hybrid [formula omitted] layer decomposing and two-dimensional variation transform

Image fusion of CT and MRI significantly improves the diagnosis of diseases. This study proposes a novel CT and MRI image fusion algorithm based on hybrid ℓ0ℓ1 layer decomposition and two-dimensional variation transforms to reduce the loss of complementary features. From the characteristics of CT and MRI images, the two-dimensional variation transform is used to separate the low-frequency and high-frequency features of the MRI image to preserve all detailed features of MRI. Then, the separated low-frequency feature images and the CT image are transformed by hybrid ℓ0ℓ1 layer decomposing to separate the edge features and contour features. Furthermore, the separated feature images are synthesized separately to obtain the fusion images of different features such as edge and contour feature fusion images. Finally, the fusion images of different features are reconstructed to obtain the final fusion images. Experiments show that the fusion algorithm better fuses the complementary features of different images and reduces the loss of information.

P105. When does the construct need to extend to the thoracic spine in patients undergoing correction for cervical deformity?

<h3>BACKGROUND CONTEXT</h3> Thoracolumbar malalignment is often seen in patients presenting with cervical deformities. For operative cervical deformity (CD) patients, it is unknown when the thoracic spine should be included in the construct. <h3>PURPOSE</h3> To investigate the CD patients in whom fusion to the thoracic spine was warranted. <h3>STUDY DESIGN/SETTING</h3> Retrospective cohort study. <h3>PATIENT SAMPLE</h3> A total of 72 CD patients. <h3>OUTCOME MEASURES</h3> Radiographic parameters; complications, distal junctional kyphosis [DJK], reoperation; Health-Related Quality-of-Life [HRQLs]: NDI, NRS-neck, mJOA. <h3>METHODS</h3> Included: operative CD patients (C2-C7 lordosis < -15°, TS-CL >35°, segmental cervical kyphosis >15° across any 3 vertebra between C2-T1, C2-C7 SVA >4cm, McGregor's slope >20°, or CBVA > 25°) with baseline (BL) and up to 2-year (2Y) data. Patients with UIV at or above C4 and LIV extending beyond C7 into the thoracic spine were isolated (CT fusions). CT fusion patients were further stratified to upper and lower thoracic LIVs: T1-T4 [Short Fusion], beyond T4 [Long Fusion]. CT fusion patients were identified as having an optimal outcome at 2-year postop if they 1) did not have DJF and 2) had Virk et al. good clinical outcome [≥2 of the following: NDI <20 or meeting MCID, mild myelopathy (mJOA ≥14), NRS-Neck ≤5 or improved by ≥2 points from baseline]. Univariate analysis compared patients with long fusion and optimal outcome (L/Success) vs patients with short fusion and treatment failure (S/Fail). Multivariate analysis and ROC curve assessed demographic, surgical, and radiographic predictors of S/ Fail and L/S Success status. Conditional inference tree (CIT) determined cut-off values for the continuous predictors. <h3>RESULTS</h3> Seventy-two cervical deformity patients with CT fusion included (60.3±9.0years, 60% F, 29.4±7.6 kg/m2, levels fused: 7.8±3.2). By approach, 61% posterior-only and 39% combined. Fifty-nine patients (82%) had CT fusions with LIV of T4 or above, while 13 patients (18%) had fusions extending below T4. Thirty-two patients (44.4%) met the optimal outcome criteria, with no difference by fusion length (p=0.171). Eight patients qualified as long fusions with treatment success, while 35 patients were classified as short fusions with treatment failure. Regression analysis identified the predictors of treatment success in patients with fusion construct extending beyond T4: baseline sacral slope ≤33.5° (OR: 15.0), not undergoing high grade (PSO or VCR) osteotomy (OR: 15.0) and being Ames descriptor type C (OR: 13.5); all p<0.05). ROC curve accounting for these factors resulted in an AUC of 82.0%. Regression analysis identified predictors of treatment failure in patients with short fusion construct: levels fused >6 (OR: 4.3), Ames descriptor type CT (OR: 11.5), Ames cSVA modifier grade 1 or 2 at BL (OR: 4.56), and Flatneck Lafage morphotype (OR: 4.5); all p<0.05. Multivariate regression and ROC curve accounting for these factors resulted in an AUC of 84.3%. <h3>CONCLUSIONS</h3> Treatment success in patients with fusion constructs extending into the thoracic spine vs treatment failure in patients with short fusions may be reliably predicted by the location of the deformity apex, measures of surgical invasiveness, and preoperative deformity severity. Specifically, treatment success in longer fusions is related to deformity apex in the cervical spine and having deformity where adequate correction does not necessitate high grade osteotomy. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.

Image Fusion of Multislice Spiral CT with Magnetic Resonance Imaging (MRI) in the Diagnosis and Nursing of Malignant Bone Diseases Using ANOVA

This work aimed to analyze the diagnostic value of dynamic scanning of multislice spiral computed tomography (MSCT) and magnetic resonance imaging (MRI) for benign and malignant bone tumor and nursing intervention. 108 patients with bone tumor were selected as the research objects, all of which underwent MSCT and MRI scans. The accuracy, sensitivity, and specificity of MSCT, MRI, and MSCT + MRI for identifying benign and malignant bone tumors and nursing care were calculated, as well as the diagnostic accuracy of MSCT, MRI, and MSCT + MRI for different bone tumor pathological types. The results showed that the accuracy of MSCT + MRI (97.56%) in distinguishing benign and malignant bone lesions was remarkably higher relative to that of MSCT (85.91%) and MRI (89.85%) ( P &lt; 0.05 ). The sensitivity and specificity of MSCT + MRI (94.85%; 90.52%) in distinguishing benign and malignant bone lesions were obviously greater in contrast to those of MSCT (83.66%; 79.05%) and MRI (86.02%; 81.17%) ( P &lt; 0.05 ). The malignant misdiagnosis rate and malignant missing report rate of MSCT + MRI in distinguishing benign and malignant bone lesions were inferior to those of MSCT and MRI notably ( P &lt; 0.05 ). The accuracy of MSCT + MRI in distinguishing osteosarcoma, giant-cell tumor of bone (GCT), bone cyst, and osteofibrous dysplasia (OFD) was evidently higher versus that of MSCT and MRI ( P &lt; 0.05 ). The accuracy of MSCT + MRI in distinguishing osteofibroma and ganglioneuroma was greatly higher than that of MSCT and MRI ( P &lt; 0.05 ). The accuracy of MSCT + MRI in distinguishing osteofibroma and ganglioneuroma was 68.64% and 71.63%, respectively. In short, in contrast to the single MSCT and MRI examination, MSCT combined with MRI detection can effectively improve the accuracy of judgment for benign and malignant bone tumor lesions and nursing care and had higher sensitivity and specificity. MSCT combined with MRI had better performance in identifying osteosarcoma, GCT, bone cyst, and OFD but poor performance in osteofibroma and ganglioneuroma.

Update in Paravalvular Leak Closure.

Given the low occurrence of clinically important paravalvular leak (PVL), there are no large registries or trials in this space to investigate management strategies. This review integrates newer evidence, particularly in imaging guidance for these complex procedures, novel techniques and approaches that our group has taken, as well as approaches to more complex PVL plugging reported in case reports. Perhaps the largest area of growth in the management of PVL is the use of advanced imaging in both pre-procedure evaluation and intra-procedural guidance with gated cardiac CT, 3D TEE, and fluoroscopy fusion technologies. Outside the USA, a new device, the Occlutech PLD, has become available with early data indicating high success rates. There remains little randomized data to support the efficacy of percutaneous PVL closure. Gated cardiac CT has become key to the pre-procedure evaluation for transcatheter closure as it allows for increased procedural efficiency and more accurate pre-procedure planning, particularly when combined with 3D printing. Intra-procedural TEE-fluoro fusion allows for more rapid crossing of defects by providing a visual target for interventionalists. The advent of purpose-built devices for PVL closure may further increase the efficacy and efficiency of percutaneous closure, but significant barriers remain for approval of these devices in the USA.

MRI and CT Fusion in Stereotactic Electroencephalography: A Literature Review

Epilepsy is a common neurological disease characterized by spontaneous recurrent seizures. Resection of the epileptogenic tissue may be needed in approximately 25% of all cases due to ineffective treatment with anti-epileptic drugs. The surgical intervention depends on the correct detection of epileptogenic zones. The detection relies on invasive diagnostic techniques such as Stereotactic Electroencephalography (SEEG), which uses multi-modal fusion to aid localizing electrodes, using pre-surgical magnetic resonance and intra-surgical computer tomography as the input images. Moreover, it is essential to know how to measure the performance of fusion methods in the presence of external objects, such as electrodes. In this paper, a literature review is presented, applying the methodology proposed by Kitchenham to determine the main techniques of multi-modal brain image fusion, the most relevant performance metrics, and the main fusion tools. The search was conducted using the databases and search engines of Scopus, IEEE, PubMed, Springer, and Google Scholar, resulting in 15 primary source articles. The literature review found that rigid registration was the most used technique when electrode localization in SEEG is required, which was the proposed method in nine of the found articles. However, there is a lack of standard validation metrics, which makes the performance measurement difficult when external objects are presented, caused primarily by the absence of a gold-standard dataset for comparison.

Ten-step method of high-dose LDR 125 I brachytherapy for intermediate-risk prostate cancer.

Dose escalation is key for improved outcomes in intermediate‐risk prostate cancer, including unfavorable intermediate‐risk (UIR) cases. This educational report is designed to provide information about our quality high‐dose 125I seed implantation monotherapy technique in which a biologically effective dose (BED) ≧ 200 Gy is applied for treatment of intermediate‐risk prostate cancer. This protocol is named the “Ten‐step Method,” where the rationale and principle of the method are based on the following four goals: (1) The entire prostate should be covered by the prescription isodose distribution with a sufficient margin from the prostatic capsule, achieving high D90 and V100 values by 125I seed implantation. (2) The high‐dose cloud (240 Gy) should not invade the urethra or rectum. (3) In order to achieve goals (1) and (2), make the high‐dose cloud intentionally along the periphery (bilateral wall to anterior wall) away from the urethra and rectum. (4) In order to achieve goal (3), seeds at the periphery, except those anterior to the rectal wall, should be placed just 1mm inside the capsule. The data obtained from a total of 137 patients with intermediate‐risk prostate cancer treated with low‐dose‐rate (LDR) monotherapy are shown. The dosimetry parameters were monitored at 1 month after seed implantation by using CT and MRI fusion guidance. The data at 1 month after LDR were: Average D90, BED, and V100 of 125I LDR monotherapy were 194.1 Gy, 207.3 Gy, and 99%, respectively. This ten‐step method was reproducible in 137 patients with intermediate‐risk prostate cancer, allowing administration of high‐dose monotherapy with excellent clinical outcomes.

Application of 3D modeling and fusion technology of medical image data in image teaching

BackgroundWe combined anatomy with imaging, transformed the 2D information of various imaging techniques into 3D information, and form the assessment system of real medical imaging cases in order to make up for the deficiencies in the current teaching of the medical imaging technology students.MethodsA total of 460 medical imaging students were selected and randomly divided into two groups. The research group received the teaching of the fusion of the original CT and MR data 3D model and the original image combined with 3D anatomical image. CT and MRI data are imported through load DICOM of 3D slicer. Different tissues and organs are segmented by threshold and watershed algorithm of segment editor module. Models are exported through export / import models and label maps in segmentation. Save the NHDR file of the original data and Obj file of the corresponding model through save the NHDR and corresponding Obj files are loaded into probe 1.0 software. The software can give different colors to the three-dimensional models of different organs or tissues to display the stereo models and related data, and display the hook edges of organ models on coronal, sagittal and axial images. At the same time, annotation can be established in the corresponding anatomical position. Finally, it can be saved as a single file of Hwl, and the teaching can be opened at any time through the program of probe 1.0. Statistical analysis Academic self-efficacy scale and Self-directed learning ability scale was adopted by self-directed learning evaluation scale between two groups.ResultsCompare the theoretical scores and case analysis scores of the two groups. The scores of the study and control groups were significantly higher than those of the control group. Before the experiment, no significant difference was detected in the self-efficacy of learning ability and learning behavior between the two groups, while after the experiment, these differences between the two groups were statistically significan. Moreover, the learning ability self-efficacy and learning behavior of the two groups of students after the experiment was significantly higher than that before the experiment. The self-efficacy of the learning behavior of the control group was higher after the experiment than that before the experiment, albeit the difference was not statistically significant.ConclusionsThe modern, information-based and humanized experimental teaching mode will be constantly improved under the support of PACS system in order to optimize the medical imaging teaching activities for the development of modern medical education.

MRI and dual-energy CT fusion anatomic imaging in Ru-106 ophthalmic brachytherapy

PurposeBrachytherapy with Ru-106 is widely used for the treatment of intraocular tumors, and its efficacy depends on the accuracy of radioactive plaque placement. Ru-106 plaques are MRI incompatible and create severe metal artifacts on conventional CT scans. Dual-energy CT scans (DECT) may be used to suppress such artifacts. This study examines the possibility of creating fusion images from MRI scans (preoperatively) and DECT scans (with the plaque in place) as a tool for confirming the anatomic accuracy of plaque placement. Methods and MaterialsSix patients with intraocular lesions (5 with choroidal melanoma and 1 with a retinal vasoproliferative lesion) were included. Fusion images of preoperative MRI scans and DECT scans with the plaque in place were created with the Demo version of the ImFusion suite (ImFusion GmbH, Munchen Germany). Clearance margins between the tumor and plaque edge in axial, transverse, and coronal planes as well as the elevation of the posterior plaque edge from the sclera were recorded and associated with the location of the lesion. ResultsPlaque-tumor clearance margins for transverse, sagittal, and coronal planes were higher for anteriorly located lesions (5.13 mm ± 0.11 [5.0–5.2], 5.10 mm ± 0.26 [4.9–5.4], and 5.33 mm ± 0.45 [4.9–5.8] respectively) than for posteriorly located lesions (4.16 mm ± 1.44 [2.5–5.1], 4.13 mm ± 1.42 [2.5–5.1], and 4.2 mm ± 1.21 [2.8–5.0], respectively). The elevation of the posterior plaque edge from the sclera was 0.33 mm ± 0.28 [0–0.5] and 0.63 mm ± 0.60 [0.7–1.2] for posterior and anterior lesions, respectively. ConclusionsFusion images between DECT and MRI scans may be used as a tool to confirm the accuracy of Ru-106 plaque placement in relation with the intraocular tumors in ophthalmic brachytherapy.

Inter-observer agreement of computed tomography and magnetic resonance imaging on gross tumor volume delineation of intrahepatic cholangiocarcinoma: an initial study.

Intrahepatic cholangiocarcinoma (ICC) is the second most common primary liver tumor, and local radiotherapy has a positive effect on patients with an unresectable tumor. Accurate delineation of gross tumor volume (GTV) is crucial to improve the efficacy of radiotherapy. The purpose of this article was to evaluate the consistency of CT, diffusion weighted imaging (DWI) and Gadoxetic acid disodium (Gd-EOB-DTPA)-enhanced MRI on GTV delineation of ICC. Fourteen patients with ICC underwent CT (Plain and Portal, CT scans before and 70 s after the injection of Omnipaque, respectively), DWI, and Gd-EOB-DTPA-enhanced MRI (EOB 70 s and EOB 15 min, mDIXON scans at 70 s and 15 min after the injection of Gd-EOB-DTPA, respectively) examinations before radiotherapy. Volumes of GTV delineation on CT and MRI images were recorded. Dice similarity coefficient (DSC) was calculated to evaluate the spatial overlap. Tumor volume on DWI and EOB 15 min were larger than that on EOB 70 s significantly (both P=0.004). DSC of DWI was significantly larger than that of other CT and MRI sequences (all P≤0.002). DSC of EOB 15 min tended to be larger than that of other CT sequences and EOB 70 s, however, without significances (all P>0.005). Significant correlation was found between DSC and tumor volume (R=0.35, P=0.003). DWI had significantly higher agreement on GTV delineation of ICC. GTV delineations of ICC on Gd-EOB-DTPA-enhanced MRI showed excellent inter-observer agreement. Fusion of CT and MRI images should be considered to improve the accuracy of GTV delineation.

A novel mono-modality fusion imaging method based on three-dimensional contrast-enhanced ultrasound for the evaluation of ablation margins after microwave ablation of hepatocellular carcinoma.

To investigate the feasibility and accuracy of using preoperative and postoperative three-dimensional contrast-enhanced ultrasound (3D CEUS) fusion imaging in the evaluation of safety margins after thermal ablation of hepatocellular carcinoma (HCC). A total of the 24 patients with HCC who underwent microwave ablation (MWA) between June 2020 and December 2020 were enrolled in this study. All patients received preoperative and postoperative 3D CEUS. The preoperative and postoperative 3D CEUS images were then fused. The success rate and evaluation time were recorded. The ablation margin and whether or not the safety margin was reached were calculated and recorded. If the ablation safety margin was not reached, the residual ablation volume needed to cover the safety margin was calculated automatically. The agreement between contrast-enhanced computed tomography(CECT) and 3D CEUS fusion imaging in the evaluation of ablation margins was explored using the kappa coefficient. The 3D CEUS fusion success rate was 95.8% (23/24), with a mean fusion time of (4.1±1.8) minutes. Twenty-three tumors were completely ablated, and the safety margin was achieved for 9 tumors. The ablation margin of 14 tumors was <5 mm. The mean uncovered safety margin volume was (2.27±2.11) mL, and the mean proportion of the uncovered safety margin to the whole safety margin was 16.8%. According to the results of preoperative and postoperative CECT fusion imaging, the ablation margin of 13 tumors was <5 mm, and the ablation margin of 10 tumors was >5 mm. The 2 methods showed excellent consistency, with a Kappa value of 0.911 (P=0.000012). This study has presented a novel mono-modality fusion imaging method based on CEUS. We demonstrated that 3D CEUS fusion has a short fusion time and a high success rate, as well as good consistency with enhanced CT fusion. Therefore, 3D CEUS fusion is a feasible and accurate tool for evaluating the immediate efficacy of thermal ablation of HCC.

Medical image fusion method by deep learning

Deep learning technology has been extensively explored in pattern recognition and image processing areas. A multi-mode medical image fusion with deep learning will be proposed, according to the characters of multi-modal medical image, medical diagnostic technology and practical implementation, according to the practical needs for medical diagnosis. It cannot be only made up for the deficiencies of MRI, CT and SPECT image fusion, but also can be implemented in different types of multi-modal medical image fusion problems in batch processing mode, and can be effectively overcome the limitation of only one-page processing. The proposed method can greatly improve the fusion effect, image detail clarity and time efficiency. The experiments on multi-modal medical images are implemented to analyze performance, algorithm stability and so on. The experimental results prove the superiority of our proposed method in terms of visual quality and a variety of quantitative evaluation criteria.

Interactive CT/MRI 3D Fusion for cerebral system analysis and as a preoperative surgical strategy and educational tool

Introduction: The development of systems that merge existing technologies with gathered data may bring some spectacular effects that are usable both in preoperative and educational processes. Augmented reality (AR) is one of the key aspects of the new medical approach. Newly fused data sets draw from it and give users a better overall experience. Aim: The main goal of this study was to enable the interactive presentation of patients’ CT and MRI combined data with the incorporation of AR tools considering the accuracy of the data with an emphasis on vascular structures. Material and methods: The registration method, reconstruction of the vascular system using tubular structures, and error analysis using surface distance measurements results were used in the system to provide accurate combined information about bony structures from CT volume and vascular objects and cerebral vessels from MRI. Results and discussion: The strategies concern a series of CTI volumes that could be used to analyze bony surgical procedures. The methods are preferred, especially to the most complicated and individually modified bony structures of the skull. Removing, replacing, or modifying these bony structures or elements of the skull could be used as an analysis of operating procedures at the particular stages of the operation during neurosurgical or otolaryngological techniques. Conclusions: Presented study regarded to an innovative system consisting of a CT and MRI datasets fusion. The distance analysis of the segmented vascular model and proposed method for stabilization of the human head combined with virtual sculpting technique. In conclusion, it was meaningful in many aspects of the scientific-technological merge.