Maximum Intensity Projection Images Research Articles

Automated segmentation and deep learning classification of ductopenic parotid salivary glands in sialo cone-beam CT images.

This study addressed the challenge of detecting and classifying the severity of ductopenia in parotid glands, a structural abnormality characterized by a reduced number of salivary ducts, previously shown to be associated with salivary gland impairment. The aim of the study was to develop an automatic algorithm designed to improve diagnostic accuracy and efficiency in analyzing ductopenic parotid glands using sialo cone-beam CT (sialo-CBCT) images. We developed an end-to-end automatic pipeline consisting of three main steps: (1) region of interest (ROI) computation, (2) parotid gland segmentation using the Frangi filter, and (3) ductopenia case classification with a residual neural network (RNN) augmented by multidirectional maximum intensity projection (MIP) images. To explore the impact of the first two steps, the RNN was trained on three datasets: (1) original MIP images, (2) MIP images with predefined ROIs, and (3) MIP images after segmentation. Evaluation was conducted on 126 parotid sialo-CBCT scans of normal, moderate, and severe ductopenic cases, yielding a high performance of 100% for the ROI computation and 89% for the gland segmentation. Improvements in accuracy and F1 score were noted among the original MIP images (accuracy: 0.73, F1 score: 0.53), ROI-predefined images (accuracy: 0.78, F1 score: 0.56), and segmented images (accuracy: 0.95, F1 score: 0.90). Notably, ductopenic detection sensitivity was 0.99 in the segmented dataset, highlighting the capabilities of the algorithm in detecting ductopenic cases. Our method, which combines classical image processing and deep learning techniques, offers a promising solution for automatic detection of parotid glands ductopenia in sialo-CBCT scans. This may be used for further research aimed at understanding the role of presence and severity of ductopenia in salivary gland dysfunction.

Image quality of whole-body diffusion MR images comparing deep-learning accelerated and conventional sequences.

To compare the image quality of deep learning accelerated whole-body (WB) with conventional diffusion sequences. Fifty consecutive patients with bone marrow cancer underwent WB-MRI. Two experts compared axial b900 s/mm2 and the corresponding maximum intensity projections (MIP) of deep resolve boost (DRB) accelerated diffusion-weighted imaging (DWI) sequences (time of acquisition: 6:42 min) against conventional sequences (time of acquisition: 14 min). Readers assessed paired images for noise, artefacts, signal fat suppression, and lesion conspicuity using Likert scales, also expressing their overall subjective preference. Signal-to-noise and contrast-to-noise ratios (SNR and CNR) and the apparent diffusion coefficient (ADC) values of normal tissues and cancer lesions were statistically compared. Overall, radiologists preferred either axial DRB b900 and/or corresponding MIP images in almost 80% of the patients, particularly in patients with a high body-mass index (BMI > 25 kg/m2). In qualitative assessments, axial DRB images were preferred (preferred/strongly preferred) in 56-100% of cases, whereas DRB MIP images were favoured in 52-96% of cases. DRB-SNR/CNR was higher in all normal tissues (p < 0.05). For cancer lesions, the DRB-SNR was higher (p < 0.001), but the CNR was not different. DRB-ADC values were significantly higher for the brain and psoas muscles, but not for cancer lesions (mean difference: + 53 µm2/s). Inter-class correlation coefficient analysis showed good to excellent agreement (95% CI 0.75-0.93). DRB sequences produce higher-quality axial DWI, resulting in improved MIPs and significantly reduced acquisition times. However, differences in the ADC values of normal tissues need to be considered. Deep learning accelerated diffusion sequences produce high-quality axial images and MIP at reduced acquisition times. This advancement could enable the increased adoption of Whole Body-MRI for the evaluation of patients with bone marrow cancer. Deep learning reconstruction enables a more than 50% reduction in acquisition time for WB diffusion sequences. DRB images were preferred by radiologistsin almost 80% of cases due to fewer artefacts, improved background signal suppression, higher signal-to-noise ratio, and increased lesion conspicuity in patients with higher body mass index. Cancer lesion diffusivity from DRB images was not different from conventional sequences.

Vascular feature identification in actinic keratosis grades I-III using dynamic optical coherence tomography with automated, quantitative analysis

Clinical grading of actinic keratosis (AK) is based on skin surface features, while subclinical alterations are not taken into consideration. Dynamic optical coherence tomography (D-OCT) enables quantification of the skin´s vasculature, potentially helpful to improve the link between clinical and subclinical features. We aimed to compare microvascular characteristics across AK grades using D-OCT with automated vascular analysis. This explorative study examined AK and photodamaged skin (PD) on the face or scalp. AKs were clinically graded according to the Olsen Classification scheme before D-OCT assessment. Using an open-source software tool, the OCT angiographic analyzer (OCTAVA), we quantified vascular network features, including total and mean vessel length, mean vessel diameter, vessel area density (VAD), branchpoint density (BD), and mean tortuosity from enface maximum intensity projection images. Additionally, we performed subregional analyses on selected scans to overcome challenges associated with imaging through hyperkeratosis (each lesion group; n = 18). Our study included 45 patients with a total of 205 AKs; 93 grade I lesions, 65 grade II, 47 grade III and 89 areas with PD skin. We found that all AK grades were more extensively vascularized relative to PD, as shown by greater total vessel length and VAD (p ≤ 0.009). Moreover, AKs displayed a disorganized vascular network, with higher BD in AK I-II (p < 0.001), and mean tortuosity in AK II-III (p ≤ 0.001) than in PD. Vascularization also increased with AK grade, showing significantly greater total vessel length in AK III than AK I (p = 0.029). Microvascular quantification of AK unveiled subclinical, quantitative differences among AK grades I-III and PD skin. D-OCT-based microvascular assessment may serve as a supplement to clinical AK grading, potentially raising perspectives to improve management strategies.

Incidental Extracardiac Findings of Technetium-99m Pyrophosphate Scintigraphy: A Pictorial Review.

Technetium-99m pyrophosphate (Tc-99m PYP) cardiac imaging is a simple, widely available, noninvasive method to identify patients with transthyretin-type cardiac amyloidosis (ATTR),and it has remarkably high diagnostic accuracy with very high sensitivity and specificity. Visual scores of 0, 1, 2, and 3 indicate non-myocardial uptake, uptake less than rib, equal to rib, and greater than rib uptake, respectively. Semiquantitative assessment using the heart-to-contralateral lung ratio of more than 1.5 at 1 hour accurately distinguishes ATTR from the cardiac amyloid light chain subtype. However, there are several incidental non-cardiac findings that can be seen in planar images, rotating single-photon emission computed tomography (SPECT) images, maximum intensity projection images, or computed tomography images acquired for attenuation correction. These findings may lead to the early detection of a noncardiac condition that may require additional treatment. The intent of this review is to demonstrate several incidental noncardiac abnormalities that have an impact on patient management and follow-up.

Study on evaluation method of circular small shadow profusion in chest CT reconstruction images of pneumoconiosis

Objective: To select chest CT image patterns for the diagnosis of pneumoconiosis and establish a method for determining the profusion of circular small shadows in chest CT. Methods: In April 2021, 66 cases of occupational pneumoconiosis patients with digital radiography (DR) chest radiographs and chest CT imaging data with circular small shadow as the main manifestations were selected as the study objects. 1.5 mm and 5 mm chest CT axial images, 1 mm and 5 mm chest CT coronal multi-plane recombination (MPR) images, and 5 mm chest CT coronal maximum intensity projection (MIP) images were used to observe the different characteristics of pneumoconiosis patients, and were compared and analyzed with DR chest radiographs to establish the experimental chest CT standards. The consistency of the profusion results between the experimental chest CT standards and GBZ 70-2015 Diagnosis of Occupational Pneumoconiosis was verified. Results: All the 66 objects were male, including 33 cases of stage Ⅰ pneumoconiosis, 17 cases of stage Ⅱ pneumoconiosis and 16 cases of stage Ⅲ pneumoconiosis. By observing five chest CT images of 66 objects, we found that chest CT images of different modes could clearly display and identify abnormal images such as small circular shadow, large shadow, small shadow aggregation, honeycomb glass shadow, flake glass shadow, uniform low-profusion glass shadow, mesh glass shadow, cable shadow, linear shadow, subpleural spinous shadow, subpleural nodules, various kinds of emphysema and lung texture distortion and fracture. Small shadow aggregation was usually accompanied by the appearance of large shadow. The vascular shadows in 5 mm CT images had good ductility, and small nodules were easy to distinguish. The coronal MIP image of 5 mm chest CT used edge enhancement technology, which was prone to small shadow fusion and fibrotic shadow fusion. The coronal MPR image of 5 mm chest CT was highly consistent with the DR chest radiographs in terms of the integrity of film reading. GBZ 70-2015 standard was used to compare the profusion of DR chest radiographs and 5 mm chest CT coronal MPR images of 66 objects, and the consistency test Kappa=0.64. GBZ 70-2015 standard and experimental chest CT standard were used to compare the profusion results of DR chest radiographs and 5 mm chest CT coronal MPR images of 66 objects, respectively, and the consistency test Kappa=0.80, with high consistency. Conclusion: 5 mm coronal MPR image is suitable for chest CT imaging in the diagnosis of pneumoconiosis. Following the selection path and method of GBZ 70-2015 profusion criterion, the established experimental chest CT standard in determining the profusion of small circular shadows in 5 mm coronal MPR images of chest CT with pneumoconiosis has a high consistency with GBZ 70-2015 standard.

Evaluation of the Contrast Enhancement Performance of Gadopiclenol for Magnetic Resonance Angiography in Healthy Rabbits and Pigs.

Unexpected accumulations of gadolinium in various organs were reported after the administration of gadolinium-based contrast agents, making desirable to reduce the dose while maintaining equivalent diagnostic performance. The aim of this study was to evaluate the contrast enhancement performance of high relaxivity gadopiclenol compared with gadoterate meglumine in abdominal contrast-enhanced magnetic resonance angiography (CE-MRA). In a first study in healthy rabbits, axial 3D gradient echo sequences were applied at 4.7 T to study arterial enhancement as a function of gadopiclenol dose (0.025, 0.05, 0.075, and 0.1 mmol Gd/kg) or gadoterate meglumine at 0.1 mmol Gd/kg (n = 5-6/group). The increase in signal-to-noise ratio (ΔSNR) in the aorta at the first pass was measured and compared. In a second, crossover study in 6 healthy pigs, abdominal CE-MRA sequences were acquired at 3 T with gadopiclenol at 0.05 mmol Gd/kg or gadoterate meglumine at 0.1 mmol Gd/kg at a 1-week interval. Quantitatively on the maximum intensity projection (MIP) images, the mean MIP SNR within the aorta of both groups was compared. Qualitatively, a blinded comparison of the angiograms was performed by an experienced radiologist to determine the preferred contrast agent. In the rabbit, ∆SNR is linearly correlated with the gadopiclenol dose (P = 0.0010). Compared with gadoterate meglumine 0.1 mmol Gd/kg, an increase in the ∆SNR is observed after 0.05, 0.075, and 0.1 mmol Gd/kg of gadopiclenol (+63% P = 0.0731, +78% P = 0.0081, and +72% P = 0.0773, respectively), whereas at 0.025 mmol Gd/kg, ∆SNR is in the same range as with gadoterate meglumine 0.1 mmol Gd/kg (+15% P > 0.9999). In pigs, contrast enhancement after gadopiclenol at 0.05 mmol/kg is +22% superior to MIP SNR after gadoterate meglumine at 0.1 mmol Gd/kg (P = 0.3095). Qualitatively, a preference was shown for gadopiclenol images (3/6) over the gadoterate meglumine examinations (1/6), with no preference being shown for the remainder (2/6). First-pass CE-MRA is feasible with gadopiclenol at 0.05 mmol Gd/kg with at least the same arterial signal enhancement and image quality as gadoterate meglumine at 0.1 mmol Gd/kg.

Computerized assessment of background parenchymal enhancement on breast dynamic contrast-enhanced-MRI including electronic lesion removal.

Current clinical assessment qualitatively describes background parenchymal enhancement (BPE) as minimal, mild, moderate, or marked based on the visually perceived volume and intensity of enhancement in normal fibroglandular breast tissue in dynamic contrast-enhanced (DCE)-MRI. Tumor enhancement may be included within the visual assessment of BPE, thus inflating BPE estimation due to angiogenesis within the tumor. Using a dataset of 426 MRIs, we developed an automated method to segment breasts, electronically remove lesions, and calculate scores to estimate BPE levels. A U-Net was trained for breast segmentation from DCE-MRI maximum intensity projection (MIP) images. Fuzzy -means clustering was used to segment lesions; the lesion volume was removed prior to creating projections. U-Net outputs were applied to create projection images of both, affected, and unaffected breasts before and after lesion removal. BPE scores were calculated from various projection images, including MIPs or average intensity projections of first- or second postcontrast subtraction MRIs, to evaluate the effect of varying image parameters on automatic BPE assessment. Receiver operating characteristic analysis was performed to determine the predictive value of computed scores in BPE level classification tasks relative to radiologist ratings. Statistically significant trends were found between radiologist BPE ratings and calculated BPE scores for all breast regions (Kendall correlation, ). Scores from all breast regions performed significantly better than guessing ( from the -test). Results failed to show a statistically significant difference in performance with and without lesion removal. BPE scores of the affected breast in the second postcontrast subtraction MIP after lesion removal performed statistically greater than random guessing across various viewing projections and DCE time points. Results demonstrate the potential for automatic BPE scoring to serve as a quantitative value for objective BPE level classification from breast DCE-MR without the influence of lesion enhancement.

Treatment planning and delivery practice of lung SBRT: Results of the 2022 ESTRO physics survey

Background and purposeThe use of Stereotactic Body Radiation Therapy (SBRT) in lung cancer is increasing. However, there is no consensus on the most appropriate treatment planning and delivery practice for lung SBRT. To gauge the range of practice, quantify its variability and identify where consensus might be achieved, ESTRO surveyed the medical physics community. Materials and methodsAn online survey was distributed to ESTRO’s physicist membership in 2022, covering experience, dose and fractionation, target delineation, dose calculation and planning practice, imaging protocols, and quality assurance. ResultsTwo-hundred and forty-four unique answers were collected after data cleaning. Most respondents were from Europe the majority of which had more than 5 years’ experience in SBRT. The large majority of respondents deliver lung SBRT with the VMAT technique on C-arm Linear Accelerators (Linacs) employing daily pre-treatment CBCT imaging. A broad spectrum of fractionation schemes were reported, alongside an equally wide range of dose prescription protocols. A clear preference was noted for prescribing to 95% or greater of the PTV. Several issues emerged regarding the dose calculation algorithm: 22% did not state it while 24% neglected to specify the conditions under which the dose was calculated. Contouring was usually performed on Maximum or Average Intensity Projection images while dose was mainly computed on the latter. No clear indications emerged for plan homogeneity, complexity, and conformity assessment. Approximately 40% of the responders participated in inter-centre credentialing of SBRT in the last five years. Substantial differences emerged between high and low experience centres, with the latter employing less accurate algorithms and older equipment. ConclusionThe survey revealed an evident heterogeneity in numerous aspects of the clinical implementation of lung SBRT treatments. International guidelines and codes of practice might promote harmonisation.

Automatic detection method of intracranial aneurysms on maximum intensity projection images based on SE-CaraNet

Conventional maximum intensity projection (MIP) images tend to ignore some morphological features in the detection of intracranial aneurysms, resulting in missed detection and misdetection. To solve this problem, a new method for intracranial aneurysm detection based on omni-directional MIP image is proposed in this paper. Firstly, the three-dimensional magnetic resonance angiography (MRA) images were projected with the maximum density in all directions to obtain the MIP images. Then, the region of intracranial aneurysm was prepositioned by matching filter. Finally, the Squeeze and Excitation (SE) module was used to improve the CaraNet model. Excitation and the improved model were used to detect the predetermined location in the omni-directional MIP image to determine whether there was intracranial aneurysm. In this paper, 245 cases of images were collected to test the proposed method. The results showed that the accuracy and specificity of the proposed method could reach 93.75% and 93.86%, respectively, significantly improved the detection performance of intracranial aneurysms in MIP images.

Lesion-conditioning of synthetic MRI-derived subtraction-MIPs of the breast using a latent diffusion model

The purpose of this feasibility study is to investigate if latent diffusion models (LDMs) are capable to generate contrast enhanced (CE) MRI-derived subtraction maximum intensity projections (MIPs) of the breast, which are conditioned by lesions. We trained an LDM with n = 2832 CE-MIPs of breast MRI examinations of n = 1966 patients (median age: 50 years) acquired between the years 2015 and 2020. The LDM was subsequently conditioned with n = 756 segmented lesions from n = 407 examinations, indicating their location and BI-RADS scores. By applying the LDM, synthetic images were generated from the segmentations of an independent validation dataset. Lesions, anatomical correctness, and realistic impression of synthetic and real MIP images were further assessed in a multi-rater study with five independent raters, each evaluating n = 204 MIPs (50% real/50% synthetic images). The detection of synthetic MIPs by the raters was akin to random guessing with an AUC of 0.58. Interrater reliability of the lesion assessment was high both for real (Kendall’s W = 0.77) and synthetic images (W = 0.85). A higher AUC was observed for the detection of suspicious lesions (BI-RADS ≥\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ge $$\\end{document} 4) in synthetic MIPs (0.88 vs. 0.77; p = 0.051). Our results show that LDMs can generate lesion-conditioned MRI-derived CE subtraction MIPs of the breast, however, they also indicate that the LDM tended to generate rather typical or ‘textbook representations’ of lesions.

Diagnosis of a Dural Arteriovenous Fistula Involving the Superior Sagittal Sinus: The MMA Sign on Routine 3D-TOF MR Angiography

The main feeding artery of a dural arteriovenous fistula (DAVF) involving the superior sagittal sinus (SSS) is the middle meningeal artery (MMA). MMA extends continuously from the proximal part to SSS in DAVF involving SSS. It is possible to diagnose DAVF involving SSS by evaluating the proximal part of MMA on routine three-dimensional time-of-flight MR angiography (MRA). We termed the finding in an anteroposterior maximum intensity projection (MIP) image of routine MRA in which MMA was continuously visualized at the top of the imaging slab (the upper corpus callosum level) without tapering as the MMA sign. The purpose of the present study was to examine the frequency of the MMA sign on routine MRA in patients with DAVF involving SSS and control patients. Subjects comprised 18 patients with DAVF involving SSS confirmed by angiography and 871 control subjects who underwent routine MRA. The presence of the MMA sign was retrospectively evaluated using an anteroposterior MIP image of routine MRA in patients with DAVF involving SSS and control patients. The MMA sign was observed in 17 of the 18 (94%) patients with DAVF involving SSS. In one patient with DAVF involving the posterior part of SSS without the MMA sign, the main feeding artery was the occipital artery, not MMA. The MMA sign was observed in 13 of the 871 (1.5%) control patients. The MMA sign on routine MRA is useful, suggesting DAVF involving SSS.

Adult age estimation from the sternum using maximum intensity projection images of CT and data mining in a Chinese population.

This study aimed to explore and develop data mining models for adult age estimation based on CT reconstruction images from the sternum. Maximum intensity projection (MIP) images of chest CT were retrospectively collected from a modern Chinese population, and data from 2700 patients (1349 males and 1351 females) aged 20 to 70years were obtained. A staging technique within four indicators was applied. Several data mining models were established, and mean absolute error (MAE) was the primary comparison parameter. The intraobserver and interobserver agreement levels were good. Within internal validation, the optimal data mining model obtained the lowest MAE of 9.08 in males and 10.41 in females. For the external validation (N = 200), MAEs were 7.09 in males and 7.15 in females. In conclusion, the accuracy of our model for adult age estimation was among similar studies. MIP images of the sternum could be a potential age indicator. However, it should be combined with other indicators since the accuracy level is still unsatisfactory.

Evaluation of renal arteries in hypertensive patients by multi-detector computed tomography

Introduction: Among various causes of hypertension, only 1 to 2% is renovascular hypertension which may be characterized by the various reasons in renal vascular supply. Multi-detector computed tomography is used to know about the details of the vascular structures in patients. This study aims to evaluate the renal arteries in hypertensive patients. Methods: A prospective observational study was conducted among 93 hypertensive patients. Contrastenhanced computed tomography of the abdomen and pelvis was conducted, and measurements were obtained from the axial section of the maximum intensity projection image. Data were analyzed using SPSS version 23.0. Independent t-test and Pearson’s Correlation were used. The level of statistical significance was set at p&lt;0.05. Results: The study found mean lengths for the right and left main renal arteries to be 40.33±10.26 mm and 32.36±9.55 mm, and diameters were 4.31±0.86 mm and 4.16±0.81 mm. No significant sex-based differences were observed. However, a significant age-related difference was noted in the length (p=0.012) and diameter (p=0.036) of the right main artery within the 20 to 29 and 80 to 89 years age groups. Weak correlations were observed between left renal artery length and age (r=0.221, p=0.33) and mean right renal artery diameter and age (r=-0.218, p=0.036). Prevalence of accessory renal arteries and early branches was 19.35% and 16.10% respectively. Conclusions: This study found that there were no statistically significant differences in renal arteries dimensions of hypertensive patients with different age groups and genders. Although the proportion of early renal artery branching and accessory arteries are found to be similar to previous studies there may be other associated factors for all hypertensive patients who don’t have early renal branches or accessory arteries.

Radiological evaluation of middle meningeal artery by 3D rotational angiography.

This study aimed to evaluate the topographic features and branches of the middle meningeal artery (MMA) by three-dimensional rotational angiography (3DRA) and describe its variations' radiological classification based on previous clinical examples of cadaver studies and literature. We consecutively evaluated 87 cerebral hemispheres, including 32 females and 49 right sides between May 2020 and December 2021 who had undergone unilateral or bilateral cerebral and carotid artery 3D rotational angiography imaging by CCA injection and between 18 and 76-years-old. Maximum intensity projection images with 10-30 mm slice thickness were used for the morphological evaluations and measurements of MMA and its foramen, canal, and branches. The diameters of the common carotid artery, internal carotid artery, and external carotid artery were significantly smaller in females than in men (p = 0.021, 0.021, and <0.001, respectively). According to the branching pattern, the most common pattern of the MMA was Type Ia (49.4%). The ophthalmic artery completely originated from MMA in the 1 (1.1%) cerebral hemisphere. The MMA arises from the ophthalmic artery in 2 cerebral hemispheres (2.3%), and the ophthalmic artery and maxillary artery in 2 cerebral hemispheres (2.3%). It was observed that the branching pattern may show differences when compared to the cadaver studies with a radiologic evaluation with 3D-RA.

Evaluation of Venous Structures that Are Involved in Transsylvian Approach Using 3D Rotational Venography.

In the transsylvian (TS) approach, as characterized by clipping surgery, the presurgical visualization of the superficial middle cerebral vein (SMCV) can help change the surgical approach to ensure safe microsurgery. Nevertheless, identifying preoperatively the venous structures that are involved in this approach is difficult. In this study, we investigated the venous structures that are involved in the TS approach using three-dimensional (3D) rotational venography (3D-RV) and evaluated the effectiveness of this method for presurgical simulation. Patients who underwent 3D-RV between August 2018 and June 2020 were involved in this retrospective study. The 3D-RV and partial maximum intensity projection images with a thickness of 5 mm were computationally reconstructed. The venous structures were subdivided into the following three portions according to the anatomic location: superficial, intermediate, and basal portions. In the superficial portion, predominant frontosylvian veins were observed on 31 (41%) sides, predominant temporosylvian veins on seven (9%) sides, and equivalent fronto- and temporosylvian veins on 28 (37%) sides. The veins in the intermediate (deep middle cerebral and uncal veins) and basal portions (frontobasal bridging veins) emptied into the SMCV on 57 (75%) and 34 (45%) sides, respectively. The 3D-RV images were highly representative of the venous structures observed during microsurgery. In this study, 3D-RV was utilized to capture the details of the venous structures from the superficial to the deep portions. Presurgical simulation of the venous structures that are involved in the TS approach using 3D-RV may increase the safety of microsurgical approaches.

The guiding value of the cinematic volume rendering technique in the preoperative diagnosis of brachial plexus schwannoma.

This study aimed to explore and compare the guiding value of Maximum Intensity Projection (MIP) and Cinematic Volume Rendering Technique (cVRT) in the preoperative diagnosis of brachial plexus schwannomas. We retrospectively analyzed the clinical and imaging data of 45 patients diagnosed with brachial plexus schwannomas at the First Affiliated Hospital of Zhengzhou University between January 2020 and December 2022. The enhanced three-dimensional short recovery time inversion-recovery fast spin-echo imaging (3D-STIR-SPACE) sequence served as source data for the reconstruction of MIP and cVRT. Two independent observers scored the image quality and evaluated the location of the tumor and the relationship between the tumor and the brachial plexus. The image quality scores of the two reconstruction methods were compared using the nonparametric Wilcoxon signed-rank test, and the consistency between the image and surgical results was assessed using the weighted kappa. Compared to MIP images, cVRT images had a better performance of overall image quality (p < 0.001), nerve and lump visualization (p < 0.001), spatial positional relationship conspicuity (p < 0.001), and diagnostic confidence (p < 0.001). Additionally, the consistency between the cVRT image results and surgical results (kappa =0.913, P<0.001) was higher than that of the MIP images (kappa =0.829, P<0.001). cVRT provides a high guiding value in the preoperative diagnosis of brachial plexus schwannomas and is an important basis for formulating surgical plans.

MEDS-Net: Multi-encoder based self-distilled network with bidirectional maximum intensity projections fusion for lung nodule detection

Early detection of lung cancer increases cure rates, making precise detection of lung nodules in computed tomography (CT) images crucial. However, creating an accurate computer-aided detection (CADe) system is challenging due to lung structure complexity and nodule heterogeneity. We present a lung nodule detection scheme, inspired by radiologist workflow, fusing CT scan sub-volumes with bidirectional maximum intensity projection (MIP) images within a single architecture. We have developed a novel multi-encoder-based network (MEDS-Net) using self-distillation to learn from three types of inputs: 3D sub-volume, forward, and backward MIP images. MEDS-Net utilizes three encoders to extract insights, passed to the decoder via attention units to suppress redundant features. The decoder employs a self-distillation mechanism, connecting to a distillation block with four auxiliary lung nodule detectors, accelerating convergence and enhancing learning ability. The auxiliary detectors are used during inference to reduce false positives. Our scheme, tested on 888 LIDC-IDRI dataset scans, achieved a 93.6% competition performance metric score, illustrating that fusing bidirectional MIP images with raw CT slices and auxiliary detectors empowers MEDS-Net to accurately detect lung nodules while minimizing false positives.

Improved small vessel visibility in diabetic foot arteriography using dual-energy CT

To test the feasibility and performance of dual-energy computed tomography (DECT) in foot arteriography of diabetic patients, where contrast medium is largely reduced within the small vessels. A total of 50 diabetic patients were enrolled prospectively, where DECT was acquired immediately after the CT angiography (CTA, group A) of the lower extremity. Two images were derived from the DECT data, one optimal virtual monochromatic image (VMI, group B) and one fusion image (group C), both of which were compared against the CTA image for visualising the foot arteries. The contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) were evaluated. The arterial course and contrast were graded each using a five-point scale. The clarity of small vessel depiction was quantified by comparing the number of plantar metatarsal arteries found in the maximum intensity projection image. The median CNRs and SNRs obtained in group B were approximately 45% and 20% higher than those in groups A and C, respectively (p<0.05). Group B also received higher subjective scores on the posterior tibial artery and the foot arteries (all >3) than groups A and C. The number of visible branches of the plantar metatarsal arteries was found to be substantially higher (p<0.05) in group B (median=6) than in groups A (median=2) and C (median=4). DECT was found to be superior to conventional CTA in foot arteriography, and beyond the lower extremity, it might be a general favourable solution for imaging regions with small vessels and reduced contrast medium.

Facilitating cell segmentation with the projection-enhancement network

Contemporary approaches to instance segmentation in cell science use 2D or 3D convolutional networks depending on the experiment and data structures. However, limitations in microscopy systems or efforts to prevent phototoxicity commonly require recording sub-optimally sampled data that greatly reduces the utility of such 3D data, especially in crowded sample space with significant axial overlap between objects. In such regimes, 2D segmentations are both more reliable for cell morphology and easier to annotate. In this work, we propose the projection enhancement network (PEN), a novel convolutional module which processes the sub-sampled 3D data and produces a 2D RGB semantic compression, and is trained in conjunction with an instance segmentation network of choice to produce 2D segmentations. Our approach combines augmentation to increase cell density using a low-density cell image dataset to train PEN, and curated datasets to evaluate PEN. We show that with PEN, the learned semantic representation in CellPose encodes depth and greatly improves segmentation performance in comparison to maximum intensity projection images as input, but does not similarly aid segmentation in region-based networks like Mask-RCNN. Finally, we dissect the segmentation strength against cell density of PEN with CellPose on disseminated cells from side-by-side spheroids. We present PEN as a data-driven solution to form compressed representations of 3D data that improve 2D segmentations from instance segmentation networks.

New Data-Driven Gated (DDG) PET/CT for Radiation Treatment Planning of NSCLC

PET/CT for RT simulation of NSCLC is typically conducted in two separate imaging sessions of gated 4D-CT in radiation oncology and fusion with non-gated PET/CT in diagnostic imaging, making treatment response assessment (TRA) with PET/CT and radiation treatment planning (RTP) challenging. To remedy this short-coming, we designed a new data-driven gated (DDG) PET/CT, based on the self-gated PET and self-gated CT data to derive the gated PET/CT data free of tumor blurring and misregistration artifacts without any respiratory monitoring device (ease of use and cost reduction). The new DDG PET/CT also provides the important 4D-CT information of end-inspiratory (EI), end-expiratory (EE), maximum-intensity-projection (MIP) and average-intensity-projection (AIP) CT's. The total imaging time is < 15 min, which makes DDG PET/CT simulation in a single session without any respiratory monitoring device possible. Acquisition is standard PET/CT imaging protocol followed by a cine CT imaging of the tumor area. We utilized the commercial DDG PET from GE Discovery PET/CT and designed a new DDG CT for attenuation correction of the DDG PET for reduction of tumor motion and mis-registration. The AIP and MIP images were from the average and maximum pixel values of the cine CT images, respectively. The cine CT images with the largest and smallest average CT number in the lung region were identified as the EE and EI CT images, respectively. For the image slices without any lung region present, the largest and smallest expansions of the body outline contour were selected to be the EE and EI CT phases, respectively. The AIP and EE CT images were used for attenuation correction of the PET and DDG PET data, respectively. Both DDG PET/CT and 4D-CT of (EI, EE, AIP and MIP) are for RT simulation of NSCLC. 38 4D-CT patient data sets were compared at the EE and EI phases between our DDG CT and 4D-CT to demonstrate the applicability of the newly designed DDG CT. A prototype software on a Dell PC of i5-6500 CPU has been successfully developed to enable DDG PET/CT and 4D-CT on seven GE Discovery PET/CT scanners. The operating system was Ubuntu and the computer language was Python. In the EE phase, the images selected by DDG CT and 4D CT were identical 62.5±21.6% of the time. In the EI phase, the images selected by DDG CT and 4D CT were identical 68.2±18.9% of the time. Inspection of the EE and EI phases of DDG CT and 50% and 0% of 4D-CT demonstrated both data sets of DDG CT and 4D CT were almost identical. A new DDG PET/CT and 4D-CT in a single imaging session without any hardware gating has been developed for TRA and RTP of NSCLC. The utility is available on the network and it enables all PET/CT scanners on the network for TRA and RTP. The new DDG CT provides the benefits of 4D-CT without any requirement of external hardware gating and make possible DDG PET/CT free of tumor motion and misregistration artifacts.