Automated Segmentation Tools Research Articles

Essential parameters needed for a U-Net-based segmentation of individual bones on planning CT images in the head and neck region using limited datasets for radiotherapy application.

Objective.The field of radiotherapy is highly marked by the lack of datasets even with the availability of public datasets. Our study uses a very limited dataset to provide insights on essential parameters needed to automatically and accurately segment individual bones on planning CT images of head and neck cancer patients.Approach.The study was conducted using 30 planning CT images of real patients acquired from 5 different cohorts. 15 cases from 4 cohorts were randomly selected as training and validation datasets while the remaining were used as test datasets. Four experimental sets were formulated to explore parameters such as background patch reduction, class-dependent augmentation and incorporation of a weight map on the loss function.Main results.Our best experimental scenario resulted in a mean Dice score of 0.93 ± 0.06 for other bones (skull, mandible, scapulae, clavicles, humeri and hyoid), 0.93 ± 0.02 for ribs and 0.88 ± 0.03 for vertebrae on 7 test cases from the same cohorts as the training datasets. We compared our proposed solution approach to a retrained nnU-Net and obtained comparable results for vertebral bones while outperforming in the correct identification of the left and right instances of ribs, scapulae, humeri and clavicles. Furthermore, we evaluated the generalization capability of our proposed model on a new cohort and the mean Dice score yielded 0.96 ± 0.10 for other bones, 0.95 ± 0.07 for ribs and 0.81 ± 0.19 for vertebrae on 8 test cases.Significance.With these insights, we are challenging the utilization of an automatic and accurate bone segmentation tool into the clinical routine of radiotherapy despite the limited training datasets.

P818 Imaging-based preoperative body composition is associated with the risk of postoperative complications and postoperative endoscopic recurrence in patients with Crohn’s disease

Abstract Background Alterations in body composition are common in Crohn’s disease (CD) patients and influence disease outcomes. However, the impact of preoperative body composition on postoperative outcomes is not completely understood. Therefore, this study aimed to investigate the association of preoperative body composition with postoperative complications and endoscopic recurrence in CD patients undergoing ileocolic (re-)resection (ICR). Methods CD patients (≥16 years) scheduled for an ICR with a computed tomography (CT) scan available (<12 months prior to ICR) were identified from an ongoing prospective, multicenter cohort study. Skeletal muscle mass (SM), visceral (VAT) and subcutaneous adipose tissue (SAT), and their radiation attenuation (RA) were assessed on a single CT image at L3 using a validated deep-learning automatic segmentation tool (Mosamatic). Low RA is associated with increased tissue triglyceride content (e.g. myosteatosis in case of SM). Cut-offs, based on sex-specific tertiles, were set at the lower and upper tertile for low and high values. The primary outcome was the development of overall postoperative complications within 90 days. Secondary outcomes were postoperative endoscopic recurrence (Rutgeerts’ score ≥ i2b) at 6 months, infectious complications, and moderate to severe complications (Clavien-Dindo ≥ II). Multivariable logistic regression was performed to assess the association of body composition variables with postoperative outcomes, including adjustment for confounders. Results 121 patients were included (median age 37.1 years (IQR 27.1–53.2), 59.5% female). Median disease duration at surgery was 3.6 years (IQR 0.4–12.7), and 57.0% had prior exposure to ≥ 1 biological. Overall postoperative complications were reported in 51.2%, endoscopic recurrence in 33.3%, infectious complications in 26.4%, and moderate to severe complications in 41.3%. High SM-RA (aOR 0.67; 95%CI 0.16–0.85) and low VAT (aOR 0.38; 95%CI 0.17–0.87) were associated with decreased overall postoperative complications. Low VAT-RA was associated with decreased endoscopic recurrence (aOR 0.21; 95%CI 0.07–0.67), and high SAT with increased infectious complications (aOR 3.05; 95%CI 1.28–7.28). Conclusion Low visceral adipose tissue and high radiation attenuation of skeletal muscle (i.e. no myosteatosis) were protective of overall postoperative complications in CD patients undergoing ICR. Low radiation attenuation of visceral adipose tissue (i.e. high triglyceride content) was protective of endoscopic recurrence. High subcutaneous adipose tissue increased the risk of postoperative infectious complications. Further research in larger populations is warranted to validate these findings to improve preoperative stratification and strategies for CD.

Manual Segmentation of the Human Choroid Plexus Using Brain MRI.

The choroid plexus has been implicated in neurodevelopment and a range of brain disorders. Evidence demonstrates that the choroid plexus is critical for brain maturation, immune/ inflammatory regulation, and behavioral/cognitive functioning. However, current automated neuroimaging segmentation tools are poor at accurately and reliably segmenting the lateral ventricle choroid plexus. Furthermore, there is no existing tool that segments the choroid plexus located in the third and fourth ventricles of the brain. Thus, a protocol delineating how to segment the choroid plexus in the lateral, third, and fourth ventricle is needed to increase the reliability and replicability of studies examining the choroid plexus in neurodevelopmental and brain disorders. This protocol provides detailed steps to create separately labeled files in 3D Slicer for the choroid plexus based on DICOM or NIFTI images. The choroid plexus will be manually segmented using the axial, sagittal, and coronal planes of T1w images making sure to exclude voxels from gray or white matter structures bordering the ventricles. Windowing will be adjusted to assist in the localization of the choroid plexus and its anatomical boundaries. Methods for assessing accuracy and reliability will be demonstrated as part of this protocol. Gold standard segmentation of the choroid plexus using manual delineations can be used to develop better and more reliable automated segmentation tools that can be openly shared to elucidate changes in the choroid plexus across the lifespan and within various brain disorders.

AI-assisted quantification of hypothalamic atrophy in amyotrophic lateral sclerosis by convolutional neural network-based automatic segmentation

The hypothalamus is a small structure of the brain with an essential role in metabolic homeostasis, sleep regulation, and body temperature control. Some neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS) and dementia syndromes are reported to be related to hypothalamic volume alterations. Despite its crucial role in human body regulation, neuroimaging studies of this structure are rather scarce due to work-intensive operator-dependent manual delineations from MRI and lack of automated segmentation tools. In this study we present a fully automatic approach based on deep convolutional neural networks (CNN) for hypothalamic segmentation and volume quantification. We applied CNN of U-Net architecture with EfficientNetB0 backbone to allow for accurate automatic hypothalamic segmentation in seconds on a GPU. We further applied our approach for the quantification of the normalized hypothalamic volumes to a large neuroimaging dataset of 432 ALS patients and 112 healthy controls (without the ground truth labels). Using the automated volumetric analysis, we could reproduce hypothalamic atrophy findings associated with ALS by detecting significant volume differences between ALS patients and controls at the group level. In conclusion, a fast and unbiased AI-assisted hypothalamic quantification method is introduced in this study (whose acceptance rate based on the outlier removal strategy was estimated to be above 95%) and made publicly available for researchers interested in the conduction of hypothalamus studies at a large scale.

Early detection of white matter hyperintensities using SHIVA-WMH detector.

White matter hyperintensities (WMHs) are well-established markers of cerebral small vessel disease, and are associated with an increased risk of stroke, dementia, and mortality. Although their prevalence increases with age, small and punctate WMHs have been reported with surprisingly high frequency even in young, neurologically asymptomatic adults. However, most automated methods to segment WMH published to date are not optimized for detecting small and sparse WMH. Here we present the SHIVA-WMH tool, a deep-learning (DL)-based automatic WMH segmentation tool that has been trained with manual segmentations of WMH in a wide range of WMH severity. We show that it is able to detect WMH with high efficiency in subjects with only small punctate WMH as well as in subjects with large WMHs (i.e., with confluency) in evaluation datasets from three distinct databases: magnetic resonance imaging-Share consisting of young university students, MICCAI 2017 WMH challenge dataset consisting of older patients from memory clinics, and UK Biobank with community-dwelling middle-aged and older adults. Across these three cohorts with a wide-ranging WMH load, our tool achieved voxel-level and individual lesion cluster-level Dice scores of 0.66 and 0.71, respectively, which were higher than for three reference tools tested: the lesion prediction algorithm implemented in the lesion segmentation toolbox (LPA: Schmidt), PGS tool, a DL-based algorithm and the current winner of the MICCAI 2017 WMH challenge (Park et al.), and HyperMapper tool (Mojiri Forooshani et al.), another DL-based method with high reported performance in subjects with mild WMH burden. Our tool is publicly and openly available to the research community to facilitate investigations of WMH across a wide range of severity in other cohorts, and to contribute to our understanding of the emergence and progression of WMH.

Separating Alzheimer’s disease from Dementia with Lewy bodies using a volumetric MRI classifier

AbstractBackgroundSeparating Alzheimer’s dementia (AD) from dementia with Lewy bodies (DLB) can be challenging, especially in DLB patients with concomitant AD pathology. It is imperative to distinguish the diseases for optimal patient management. To date, no single MRI feature can accurately separate the two diseases. This study examines whether a linear combination of regional atrophy measures can differentiate cognitively normal (CN), AD, and DLB patients.MethodWe computed volumes of 70 brain regions using an automatic segmentation tool. For each region, we derived age, sex, and head size‐dependent z‐scores. A linear classifier with all brain regions was used to predict if a patient belonged to CN, AD, or DLB. Diagnoses were based on international consensus criteria and CSF‐biomarkers. The classifier was tested in 327 CN (mean age 67±10y, 48%F), 501 AD (70±9y, 49%F), and 99 DLB (69±7y, 15%F) patients from three cohorts (ADNI, PredictND, ADC). The ADC cohort was assessed separately, as almost all DLB patients were from the ADC. In addition, we assessed classification performance in a subgroup of DLB patients with AD‐like CSF (DLB+, n = 57). Performance was evaluated using 10‐fold cross‐validation.ResultPerformance measures for the classifier are summarized in Table 1. The classifier accuracy was assessed in the three‐class classification task and in two‐class classifications where one of the groups was excluded from the analysis. The balanced accuracy in the three‐class classification was 67% for the DLB group and the DLB+ subgroup. For AD vs. DLB, the AUC was 0.77(95% CI:0.70‐0.82), for AD vs. DLB+ AUC was 0.72(0.64‐0.80). The classifier correctly classified CN and AD patients more often than DLB patients (Table 2). Most DLB+ patients were misclassified as AD. The hippocampus was the best single region for distinguishing AD vs. DLB, with an AUC of 0.70(0.64‐0.76).ConclusionThe joint classifier achieved high accuracy in separating AD vs. DLB patients. Although the accuracy was lower, the classifier could distinguish DLB patients with concomitant AD pathology from AD. Next, we will combine different brain regions to discern disease‐specific patterns and add clinical data to the classifier to improve differential diagnostic accuracy for DLB and DLB with concomitant AD pathology.

A longitudinal analysis of brain volume changes in myelin oligodendrocyte glycoprotein antibody-associated disease.

Myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) is a relapsing demyelinating condition. There are several cross-sectional studies showing evidence of brain atrophy in people with MOGAD (pwMOGAD), but longitudinal brain volumetric assessment is still an unmet need. Current recommendations do not include monitoring with MRI and assume distinct attacks. Evidence of ongoing axon loss will have diagnostic and therapeutic implications. In this study, we assessed brain volume changes in pwMOGAD over a mean follow-up period of 2years and compared this to changes in people with multiple sclerosis (pwMS). This is a retrospective single-center study over a 7-year period from 2014 to 2021. MRI brain scans at the time of diagnosis and follow-up in remission were collected from 14 Caucasian pwMOGAD, confirmed by serum myelin oligodendrocyte glycoprotein immunoglobulin G antibody presence, detected by live cell-based assays. Total brain volume (TBV), white matter (WM), gray matter (GM), and demyelinating lesion volumes were assessed automatically using the Statistical Parametric Mapping and FMRIB automated segmentation tools. MRI brain scans at diagnosis and follow-up on remission were collected from 32-matched pwMS for comparison. Statistical analysis was done using analysis of variance. There is evidence of TBV loss, affecting particularly GM, over an approximately 2-year follow-up period in pwMOGAD (p<.05), comparable to pwMS. WM and lesion volume change over the same period were not statistically significant (p>.1). We found evidence of loss of GM and TBV over time in pwMOGAD, similar to pwMS, although the WM and lesion volumes were unchanged.

FastSurfer-HypVINN: Automated sub-segmentation of the hypothalamus and adjacent structures on high-resolutional brain MRI

Abstract The hypothalamus plays a crucial role in the regulation of a broad range of physiological, behavioral, and cognitive functions. However, despite its importance, only a few small-scale neuroimaging studies have investigated its substructures, likely due to the lack of fully automated segmentation tools to address scalability and reproducibility issues of manual segmentation. While the only previous attempt to automatically sub-segment the hypothalamus with a neural network showed promise for 1.0 mm isotropic T1-weighted (T1w) magnetic resonance imaging (MRI), there is a need for an automated tool to sub-segment also high-resolutional (HiRes) MR scans, as they are becoming widely available, and include structural detail also from multi-modal MRI. We, therefore, introduce a novel, fast, and fully automated deep-learning method named HypVINN for sub-segmentation of the hypothalamus and adjacent structures on 0.8 mm isotropic T1w and T2w brain MR images that is robust to missing modalities. We extensively validate our model with respect to segmentation accuracy, generalizability, in-session test-retest reliability, and sensitivity to replicate hypothalamic volume effects (e.g., sex differences). The proposed method exhibits high segmentation performance both for standalone T1w images as well as for T1w/T2w image pairs. Even with the additional capability to accept flexible inputs, our model matches or exceeds the performance of state-of-the-art methods with fixed inputs. We, further, demonstrate the generalizability of our method in experiments with 1.0 mm MR scans from both the Rhineland Study and the UK Biobank—an independent dataset never encountered during training with different acquisition parameters and demographics. Finally, HypVINN can perform the segmentation in less than a minute (graphical processing unit [GPU]) and will be available in the open source FastSurfer neuroimaging software suite, offering a validated, efficient, and scalable solution for evaluating imaging-derived phenotypes of the hypothalamus.

The utility of automatic segmentation of kidney MRI in chronic kidney disease using a 3D convolutional neural network

We developed a 3D convolutional neural network (CNN)-based automatic kidney segmentation method for patients with chronic kidney disease (CKD) using MRI Dixon-based T1-weighted in-phase (IP)/opposed-phase (OP)/water-only (WO) images. The dataset comprised 100 participants with renal dysfunction (RD; eGFR < 45 mL/min/1.73 m2) and 70 without (non-RD; eGFR ≥ 45 mL/min/1.73 m2). The model was applied to the right, left, and both kidneys; it was first evaluated on the non-RD group data and subsequently on the combined data of the RD and non-RD groups. For bilateral kidney segmentation of the non-RD group, the best performance was obtained when using IP image, with a Dice score of 0.902 ± 0.034, average surface distance of 1.46 ± 0.75 mm, and a difference of − 27 ± 21 mL between ground-truth and automatically computed volume. Slightly worse results were obtained for the combined data of the RD and non-RD groups and for unilateral kidney segmentation, particularly when segmenting the right kidney from the OP images. Our 3D CNN-assisted automatic segmentation tools can be utilized in future studies on total kidney volume measurements and various image analyses of a large number of patients with CKD.

Raidionics: an open software for pre- and postoperative central nervous system tumor segmentation and standardized reporting

For patients suffering from central nervous system tumors, prognosis estimation, treatment decisions, and postoperative assessments are made from the analysis of a set of magnetic resonance (MR) scans. Currently, the lack of open tools for standardized and automatic tumor segmentation and generation of clinical reports, incorporating relevant tumor characteristics, leads to potential risks from inherent decisions’ subjectivity. To tackle this problem, the proposed Raidionics open-source software has been developed, offering both a user-friendly graphical user interface and stable processing backend. The software includes preoperative segmentation models for each of the most common tumor types (i.e., glioblastomas, lower grade gliomas, meningiomas, and metastases), together with one early postoperative glioblastoma segmentation model. Preoperative segmentation performances were quite homogeneous across the four different brain tumor types, with an average Dice around 85% and patient-wise recall and precision around 95%. Postoperatively, performances were lower with an average Dice of 41%. Overall, the generation of a standardized clinical report, including the tumor segmentation and features computation, requires about ten minutes on a regular laptop. The proposed Raidionics software is the first open solution enabling an easy use of state-of-the-art segmentation models for all major tumor types, including preoperative and postsurgical standardized reports.

AI-Accelerated Digitisation of Insect Collections: The next generation of Angled Label Image Capture Equipment (ALICE)

The digitisation of natural science specimens is a shared ambition of many of the largest collections, but the scale of these collections, estimated at at least 1.1 billion specimens (Johnson et al. 2023), continues to challenge even the most resource-rich organisations. The Natural History Museum, London (NHM) has been pioneering work to accelerate the digitisation of its 80 million specimens. Since the inception of the NHM Digital Collection Programme in 2014, more than 5.5 million specimen records have been made digitally accessible. This has enabled the museum to deliver a tenfold increase in digitisation, compared to when rates were first measured by the NHM in 2008. Even with this investment, it will take circa 150 years to digitise its remaining collections, leading the museum to pursue technology-led solutions alongside increased funding to deliver the next increase in digitisation rate. Insects comprise approximately half of all described species and, at the NHM, represent more than one-third (c. 30 million specimens) of the NHM’s overall collection. Their most common preservation method, attached to a pin alongside a series of labels with metadata, makes insect specimens challenging to digitise. Early Artificial Intelligence (AI)-led innovations (Price et al. 2018) resulted in the development of ALICE, the museum's Angled Label Image Capture Equipment, in which a pinned specimen is placed inside a multi-camera setup, which captures a series of partial views of a specimen and its labels. Centred around the pin, these images can be digitally combined and reconstructed, using the accompanying ALICE software, to provide a clean image of each label. To do this, a Convolutional Neural Network (CNN) model is incorporated, to locate all labels within the images. This is followed by various image processing tools to transform the labels into a two-dimensional viewpoint, align the associated label images together, and merge them into one label. This allows users to manually, or computationally (e.g., using Optical Character Recognition [OCR] tools) extract label data from the processed label images (Salili-James et al. 2022). With the ALICE setup, a user might average imaging 800 digitised specimens per day, and exceptionally, up to 1,300. This compares with an average of 250 specimens or fewer daily, using more traditional methods involving separating the labels and photographing them off of the pin. Despite this, our original version of ALICE was only suited to a small subset of the collection. In situations when the specimen is very large, there are too many labels, or these labels are too close together, ALICE fails (Dupont and Price 2019). Using a combination of updated AI processing tools, we hereby present ALICE version 2. This new version of ALICE provides faster rates, improved software accuracy, and a more streamlined pipeline. It includes the following updates: Hardware: after conducting various tests, we have optimised the camera setup. Further hardware updates include a Light-Emitting Diode (LED) ring light, as well as modifications to the camera mounting. Software: our latest software incorporates machine learning and other computer vision tools to segment labels from ALICE images and stitch them together more quickly and with a higher level of accuracy, significantly reducing the image processing failure rate. These processed label images can be combined with the latest OCR tools for automatic transcription and data segmentation. Buildkit: we aim to provide a toolkit that any individual or institution can incorporate into their digitisation pipeline. This includes hardware instructions, an extensive guide detailing the pipeline, and new software code accessible via Github. Hardware: after conducting various tests, we have optimised the camera setup. Further hardware updates include a Light-Emitting Diode (LED) ring light, as well as modifications to the camera mounting. Software: our latest software incorporates machine learning and other computer vision tools to segment labels from ALICE images and stitch them together more quickly and with a higher level of accuracy, significantly reducing the image processing failure rate. These processed label images can be combined with the latest OCR tools for automatic transcription and data segmentation. Buildkit: we aim to provide a toolkit that any individual or institution can incorporate into their digitisation pipeline. This includes hardware instructions, an extensive guide detailing the pipeline, and new software code accessible via Github. We provide test data and workflows to demonstrate the potential of ALICE version 2 as an effective, accessible, and cost-saving solution to digitising pinned insect specimens. We also describe potential modifications, enabling it to work with other types of specimens.

Evaluating automated longitudinal tumor measurements for glioblastoma response assessment.

Automated tumor segmentation tools for glioblastoma show promising performance. To apply these tools for automated response assessment, longitudinal segmentation, and tumor measurement, consistency is critical. This study aimed to determine whether BraTumIA and HD-GLIO are suited for this task. We evaluated two segmentation tools with respect to automated response assessment on the single-center retrospective LUMIERE dataset with 80 patients and a total of 502 post-operative time points. Volumetry and automated bi-dimensional measurements were compared with expert measurements following the Response Assessment in Neuro-Oncology (RANO) guidelines. The longitudinal trend agreement between the expert and methods was evaluated, and the RANO progression thresholds were tested against the expert-derived time-to-progression (TTP). The TTP and overall survival (OS) correlation was used to check the progression thresholds. We evaluated the automated detection and influence of non-measurable lesions. The tumor volume trend agreement calculated between segmentation volumes and the expert bi-dimensional measurements was high (HD-GLIO: 81.1%, BraTumIA: 79.7%). BraTumIA achieved the closest match to the expert TTP using the recommended RANO progression threshold. HD-GLIO-derived tumor volumes reached the highest correlation between TTP and OS (0.55). Both tools failed at an accurate lesion count across time. Manual false-positive removal and restricting to a maximum number of measurable lesions had no beneficial effect. Expert supervision and manual corrections are still necessary when applying the tested automated segmentation tools for automated response assessment. The longitudinal consistency of current segmentation tools needs further improvement. Validation of volumetric and bi-dimensional progression thresholds with multi-center studies is required to move toward volumetry-based response assessment.

Abnormalities of the Amygdala in schizophrenia: a real world study

BackgroundAmygdala plays an important role in schizophrenia (SC), but its mechanisms are still unclear. Therefore, we investigated the relationship between the resting-state magnetic resonance imaging (rsMRI) signals of the amygdala and cognitive functions, providing references for future research in this area.MethodsWe collected 40 drug-naïve SC patients and 33 healthy controls (HC) from the Third People’s Hospital of Foshan. We used rsMRI and the automatic segmentation tool to extract the structural volume and local neural activity values of the amygdala and conducted Pearson correlation analysis with the Positive and Negative Syndrome Scale (PANSS) and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) scores. Finally, we compared the clinical data, as well as the volume and functional changes of the amygdala in SC patients before and after treatment.ResultsCompared with HC, SC had widespread cognitive impairments, significant abnormalities in left amygdala function, while the reduction in volume of SC was not significant. Further Pearson correlation analysis with Bonferroni correction showed that only Immediate memory (learning) was significantly negatively correlated with fractional amplitude of low-frequency fluctuation (FALFF, r = -0.343, p = 0.001, p’ = 0.014 (Bonferroni correction)). When compared and analyzed the data difference of SC before and after treatment, we found that immediate memory and delayed memory of SC showed varying degrees of recovery after treatment (tlearning = -2.641, plearning = 0.011; tstory memory = -3.349, pstory memory = 0.001; tlist recall = -2.071, plist recall = 0.043; tstory recall = -2.424, pstory recall = 0.018). But the brain structure and function did not recover.ConclusionThere was significant dysfunction in the amygdala in SC, and after conventional treatment, the function of the amygdala did not improve with the improvement of clinical symptoms and cognitive function.

Investigation of the porosity of L/LL4 ordinary chondrite Bjurböle using synchrotron radiation microtomography and scanning electron microscopy: Implications for parent body evolution

Porosity is an essential property of chondritic meteorites and is closely related to the genesis, thermal evolution, metamorphism, and thermal properties of the meteorite parent bodies. We study porosity, its texture, and shapes at sub-micron resolution in 3D and 2D within a 0.35 cm3 sample of the L/LL4 ordinary chondrite (OC) Bjurböle using two techniques, synchrotron radiation microtomography (SRμCT) and scanning electron microscopy (SEM). We employ automated segmentation tools that can be applied to both SRμCT and SEM data. Successful segmentation results can be achieved by combining visual qualitative examination and machine learning algorithms.We report novel measurement results of three-dimensional porosity properties of Bjurböle, such as aspect ratio and connectivity of void spaces, and compare the results of 2D and 3D porosity analysis. The Bjurböle sample in this study is a complex, highly porous, and friable medium and the dominant type of porosity is intergranular, continuous porosity, which contains almost all porosity volume. The shapes of the void volumes have an important effect on the connectivity of the porosity and thermal transport properties. Positive correlations between void diameter and aspect ratio as well as void volume and connectivity are present in Bjurböle, which indicate that smaller voids have lower aspect ratios and lower connectivity. In Bjurböle, small, near-spherical voids with few connections have the highest relative frequency, whereas larger void spaces with higher aspect ratios and connectivity are significantly fewer. Completely isolated pores, i.e., voids surrounded by solid material, also have a high relative frequency, and they exist within the chondrules and the matrix. However, the volume percentage of these pores is negligible compared to that of the continuous porosity.Our results support the previously measured high porosities of Bjurböle. The volume percentage of intergranular void spaces, in particular in the matrix, and the measured high porosities are not in line with the results of thermal evolution and sintering models of chondritic parent bodies regarding petrologic type 4, which implies that Bjurböle originates from a parent body with an initial onion shell structure that fragmented during or after its metamorphic peak and quickly reaccreted into a rubble pile.

Automated localization and segmentation of cervical lymph nodes on contrast-enhanced CT using a 3D foveal fully convolutional neural network

BackgroundIn the management of cancer patients, determination of TNM status is essential for treatment decision-making and therefore closely linked to clinical outcome and survival. Here, we developed a tool for automatic three-dimensional (3D) localization and segmentation of cervical lymph nodes (LNs) on contrast-enhanced computed tomography (CECT) examinations.MethodsIn this IRB-approved retrospective single-center study, 187 CECT examinations of the head and neck region from patients with various primary diseases were collected from our local database, and 3656 LNs (19.5 ± 14.9 LNs/CECT, mean ± standard deviation) with a short-axis diameter (SAD) ≥ 5 mm were segmented manually by expert physicians. With these data, we trained an independent fully convolutional neural network based on 3D foveal patches. Testing was performed on 30 independent CECTs with 925 segmented LNs with an SAD ≥ 5 mm.ResultsIn total, 4,581 LNs were segmented in 217 CECTs. The model achieved an average localization rate (LR), i.e., percentage of localized LNs/CECT, of 78.0% in the validation dataset. In the test dataset, average LR was 81.1% with a mean Dice coefficient of 0.71. For enlarged LNs with a SAD ≥ 10 mm, LR was 96.2%. In the test dataset, the false-positive rate was 2.4 LNs/CECT.ConclusionsOur trained AI model demonstrated a good overall performance in the consistent automatic localization and 3D segmentation of physiological and metastatic cervical LNs with a SAD ≥ 5 mm on CECTs. This could aid clinical localization and automatic 3D segmentation, which can benefit clinical care and radiomics research.Relevance statementOur AI model is a time-saving tool for 3D segmentation of cervical lymph nodes on contrast-enhanced CT scans and serves as a solid base for N staging in clinical practice and further radiomics research.Key points• Determination of N status in TNM staging is essential for therapy planning in oncology.• Segmenting cervical lymph nodes manually is highly time-consuming in clinical practice.• Our model provides a robust, automated 3D segmentation of cervical lymph nodes.• It achieves a high accuracy for localization especially of enlarged lymph nodes.• These segmentations should assist clinical care and radiomics research.Graphical

Computer-assisted pre-operative automatic segmentation and registration tool for malunited radius osteotomy: A proof-of-concept study

Corrective osteotomy is a standard treatment for distal radius fractures in malunited radius cases. In order to increase the efficiency of the osteotomy pre-operative plan, in this study, a proof-of-concept framework of automatic computer-assisted segmentation and registration tool was developed for the purpose of malunited radius osteotomy pre-operative planning. The program consisted of the functions of segmentation, virtual cutting, automatic alignment and registration. One computed tomography (CT) scanning dataset of a patient's bilateral forearm was employed as an illustration example in this study. Three templates of 3D models including the healthy radius, and the pre- and post-correction injured radius were output as STL geometries for pre-operative plan purposes.

Multi-Task Learning U-Net for Functional Shoulder Sub-Task Segmentation.

The assessment of a frozen shoulder (FS) is critical for evaluating outcomes and medical treatment. Analysis of functional shoulder sub-tasks provides more crucial information, but current manual labeling methods are time-consuming and prone to errors. To address this challenge, we propose a deep multi-task learning (MTL) U-Net to provide an automatic and reliable functional shoulder sub-task segmentation (STS) tool for clinical evaluation in FS. The proposed approach contains the main task of STS and the auxiliary task of transition point detection (TPD). For the main STS task, a U-Net architecture including an encoder-decoder with skip connection is presented to perform shoulder sub-task classification for each time point. The auxiliary TPD task uses lightweight convolutional neural networks architecture to detect the boundary between shoulder sub-tasks. A shared structure is implemented between two tasks and their objective functions of them are optimized jointly. The fine-grained transition-related information from the auxiliary TPD task is expected to help the main STS task better detect boundaries between functional shoulder sub-tasks. We conduct the experiments using wearable inertial measurement units to record 815 shoulder task sequences collected from 20 healthy subjects and 43 patients with FS. The experimental results present that the deep MTL U-Net can achieve superior performance compared to using single-task models. It shows the effectiveness of the proposed method for functional shoulder STS. The code has been made publicly available at https://github.com/RobinChu9890/MTL-U-Net-for-Functional-Shoulder-STS.Clinical Relevance- This work provides an automatic and reliable functional shoulder sub-task segmentation tool for clinical evaluation in frozen shoulder.

Deep learning for the rapid automatic segmentation of forearm muscle boundaries from ultrasound datasets.

Ultrasound (US) is widely used in the clinical diagnosis and treatment of musculoskeletal diseases. However, the low efficiency and non-uniformity of artificial recognition hinder the application and popularization of US for this purpose. Herein, we developed an automatic muscle boundary segmentation tool for US image recognition and tested its accuracy and clinical applicability. Our dataset was constructed from a total of 465 US images of the flexor digitorum superficialis (FDS) from 19 participants (10 men and 9 women, age 27.4 ± 6.3years). We used the U-net model for US image segmentation. The U-net output often includes several disconnected regions. Anatomically, the target muscle usually only has one connected region. Based on this principle, we designed an algorithm written in C++ to eliminate redundantly connected regions of outputs. The muscle boundary images generated by the tool were compared with those obtained by professionals and junior physicians to analyze their accuracy and clinical applicability. The dataset was divided into five groups for experimentation, and the average Dice coefficient, recall, and accuracy, as well as the intersection over union (IoU) of the prediction set in each group were all about 90%. Furthermore, we propose a new standard to judge the segmentation results. Under this standard, 99% of the total 150 predicted images by U-net are excellent, which is very close to the segmentation result obtained by professional doctors. In this study, we developed an automatic muscle segmentation tool for US-guided muscle injections. The accuracy of the recognition of the muscle boundary was similar to that of manual labeling by a specialist sonographer, providing a reliable auxiliary tool for clinicians to shorten the US learning cycle, reduce the clinical workload, and improve injection safety.

Computed Tomographic Assessment of Pituitary Gland Dimensions in Domestic Short-Haired Cats.

The detection of subtle changes in the pituitary dimensions has relevant clinical implications. In cats, a few studies have established the cut-off values of the pituitary gland's dimensions using small and inhomogeneous samples. The aims of this study were: to determine by computed tomography (CT) the pituitary linear dimensions and the pituitary-to-brain (P:B) ratio in a sample of domestic short-haired (DSH) cats; to assess the effects of sex, age, and weight on pituitary dimensions; and to evaluate the inter- and intra-observer agreement for such measurements. All skull CTs of DSH cats performed over four years using a multidetector CT and a standardized protocol were retrospectively reviewed. The exclusion criteria were: clinical, laboratory, or CT alterations of the pituitary gland, brain diseases, fractures of the neurocranium, and diabetes. The pituitary dimensions and brain area were assessed by two different observers using multiplanar reconstructions and automated segmentation tools. Fifty-one cats were included in the final sample. The intraclass correlation coefficients for intra- and inter-observer reliability were good/excellent, and moderate/good, respectively. No differences between sexes were detected, and negligible correlations were found between age and weight. According to this study, a pituitary gland with a height > 4 mm or a P:B ratio > 0.49 mm should be considered enlarged.

Biomarkers for the detection of progressive early Alzheimer’s Disease

AbstractBackgroundWe investigated the ability of MRI, plasma, and cognitive biomarkers in detecting preclinical or prodromal Alzheimer’s Disease (AD) who will progress to the next syndrome stage of the cognitive continuum.Method589 subjects with longitudinal cognitive data were recruited from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), which included 227 cognitive unimpaired (CU) and 362 mild cognitive impairment (MCI) subjects. Preclinical or prodromal AD was defined by a low amyloid beta (Aβ)42 (A+) and high phosphorylated ‐tau (p‐tau) (T+) on cerebrospinal fluid (CSF) assessment at baseline. The following baseline MRI biomarkers were derived by an automatic segmentation tool (AccuBrain®): AD‐resemblance atrophy index (AD‐RAI), quantitative medial temporal lobe atrophy (QMTA) and hippocampus volume (HV). Plasma biomarkers included p‐tau 181, neurofilament (NFL) and APOE ε4. Montreal Cognitive Assessment (MoCA) score was collected for all subjects at baseline. Conversion (C+) was defined as subjects progressed from CU to MCI and from MCI to dementia within 4 years.ResultsOf the 589 subjects (mean [SD] age, 72.2 [6.9] years; 314 men [53.3%]), 96 (16.3%) were A+T+C+ and 180 (30.6%) were A+T+C‐. In the ROC analysis, AD‐RAI achieved the best detection ability than other individual biomarker with a sensitivity of 83.4%, a specificity of 69.4% and an AUC of 87.8%. A combination of AD‐RAI, plasma p‐tau 181, APOE ε4 and MoCA score achieved the best detection ability with AUC of 89.8%, sensitivity of 85.1% and specificity of 80.2%. In the subgroup analysis, the combination of AD‐RAI, plasma p‐tau 181, APOE ε4 and MoCA score also showed good accuracy in identifying A+T+C+ subjects in the CU group (AUC = 87.6%) and in the MCI group (AUC = 89.3%).ConclusionA panel of MRI, plasma and cognitive biomarkers might help to detect progressive early AD subjects.