A semi-automatic segmentation method for meningioma developed using a variational approach model.

Recently, the incidence of skin cancer has increased considerably and is seriously threatening human health. Automatic detection of this disease, where early detection is critical to human life, is quite challenging. Factors such as undesirable residues (hair, ruler markers), indistinct boundaries, variable contrast, shape differences, and color differences in the skin lesion images make automatic analysis quite difficult. To overcome these challenges, a highly effective segmentation method based on a fully convolutional network (FCN) is presented in this paper. The proposed improved FCN (iFCN) architecture is used for the segmentation of full-resolution skin lesion images without any pre- or post-processing. It is to support the residual structure of the FCN architecture with spatial information. This situation, which creates a more advanced residual system, enables more precise detection of details on the edges of the lesion, and an analysis independent of skin color can be performed. It offers two contributions: determining the center of the lesion and clarifying the edge details despite the undesirable effects. Two publicly available datasets, the IEEE International Symposium on Biomedical Imaging (ISBI) 2017 Challenge and PH2 datasets, are used to evaluate the performance of the iFCN method. The mean Jaccard index is 78.34%, the mean Dice score is 88.64%, and the mean accuracy value is 95.30% for the proposed method for the ISBI 2017 test dataset. Furthermore, the mean Jaccard index is 87.1%, the mean Dice score is 93.02%, and the mean accuracy value is 96.92% for the proposed method for the PH2 test dataset.

Biological diversity is decreasing at a rate of 100-1000 times pre-human rates [1] [2], and tropical rainforests are among the most vulnerable ecosystems. To avoid species extinction, we need to understand factors influencing the occurrence of species. Fast, reliable computer-assisted tools can help to describe the habitat and thus to understand species habitat associations. This understanding is of utmost importance for more targeted species conservation efforts. Due to logistical challenges and time-consuming manual processing of field data, months up to years are often needed to progress from data collection to data interpretation. Deep learning can be used to significantly shorten the time while keeping a similar level of accuracy. Here, we propose Habitat-Net: a novel Convolutional Neural Network (CNN) based method to segment habitat images of rainforests. Habitat-Net takes color images as input and after multiple layers of convolution and deconvolution produces a binary segmentation of an image. The primary contribution of Habitat-Net is the translation of medical imaging knowledge (inspired by U-Net [3]) to ecological problems. The entire Habitat-Net pipeline works automatically without any user interaction. Our only assumption is the availability of annotated images, from which Habitat-Net learns the most distinguishing features automatically. In our experiments, we use two habitat datasets: (1) canopy and (2) understory vegetation. We train the model with 800 canopy images and 700 understory images separately. Our testing dataset has 150 canopy and 170 understory images. We use the Dice coefficient and Jaccard Index to quantify the overlap between ground-truthed segmentation images and those obtained by Habitat-Net model. This results in a mean Dice Score (mean Jaccard Index) for the segmentation of canopy and understory images of 0.89 (0.81) and 0.79 (0.69), respectively. Compared to manual segmentation, Habitat-Net prediction is approximately 3K – 150K times faster. For a typical canopy dataset of 335 images, Habitat-Net reduces total processing time to 5 seconds (15 milliseconds/ image) from 4 hours (45 seconds/ image). In this study, we show that it is possible to speed up the data pipeline using deep learning in the ecological domain. In the future, we plan to create a freely available mobile app based on Habitat-Net technology to characterize the habitat directly and automated in the field. In combination with ecological models our tools will help to understand the ecology of some poorly known, but often highly threatened, species and thus contribute to more timely conservation interventions.
\nREFERENCES: 
\n1. Sachs et al. "Biodiversity conservation and the millennium development goals." Science 325.5947 (2009): 1502-1503.
\n2. Chapin Iii, F. Stuart, et al. "Consequences of changing biodiversity." Nature 405.6783 (2000): 234.
\n3. Ronneberger, Olaf, Philipp Fischer, and Thomas Brox. "U-net: Convolutional networks for biomedical image segmentation." International Conference on Medical image computing and computer-assisted intervention. Springer, Cham, 2015.

P-018: Automatic analysis of magnetic resonance imaging in multiple myeloma patients: deep-learning based pelvic bone marrow segmentation and radiomics analysis for prediction of plasma cell infiltration

Spinocerebellar ataxia type 2 (SCA2) is a neurodegenerative disorder marked by progressive brainstem and cerebellar atrophy, leading to gait ataxia. Quantifying this atrophy in magnetic resonance imaging (MRI) is critical for tracking disease progression in both symptomatic patients and preclinical subjects. However, manual segmentation of brainstem subregions (mesencephalon, pons, and medulla) is time-consuming and prone to human error. This work presents an automated deep learning framework to assess brainstem atrophy in SCA2. Using T1-weighted MRI scans from patients, preclinical carriers, and healthy controls, a U-shaped convolutional neural network (CNN) was trained to segment brainstem subregions and quantify volume loss. The model achieved strong agreement with manual segmentations, significantly outperforming four U-Net-based benchmarks (mean Dice scores: whole brainstem 0.96 vs. 0.93-0.95, pons 0.96 vs. 0.91-0.94, mesencephalon 0.96 vs. 0.89-0.93, and medulla 0.95 vs. 0.91-0.93). Results revealed severe atrophy in preclinical and symptomatic cohorts, with pons volumes reduced by nearly 50% compared to controls (p < 0.001). The mesencephalon and medulla showed milder degeneration, underscoring regional vulnerability differences. This automated approach enables rapid, precise assessment of brainstem atrophy, advancing early diagnosis and monitoring in SCA2.

Purpose To implement and evaluate two-dimensional spiral hyperpolarized xenon 129 (129Xe) ventilation MRI with flip-angle (FA) correction, as compared with conventional N4ITK (N4) correction, in healthy individuals and those with cystic fibrosis (CF). Materials and Methods In this prospective study, participants with mild CF and age-matched healthy control participants underwent 129Xe ventilation MRI using both rapid spiral (approximately 3 seconds) and conventional Cartesian (approximately 10 seconds) acquisitions. Images were corrected using N4 bias correction, and ventilation defect percentage (VDP) was calculated using median-anchored generalized linear binning (mGLB). Separately, B1 inhomogeneities in spiral images were FA-corrected and analyzed using mGLB. Gravitational gradients in ventilation were quantified from uncorrected and N4- and FA-corrected images in healthy participants. VDP from N4-corrected (VDPN4) and FA-corrected (VDPFA) images were compared between participant groups and against reader-segmented VDP (VDPRS). Statistical analyses included Wilcoxon signed rank test, Pearson correlation, and Bland-Altman analysis. Results The final analysis included 38 participants with CF (mean age, 16 years ± 6 [SD]; 20 female) and 25 healthy controls (mean age, 18 years ± 7; 13 male). Qualitatively, Cartesian and spiral acquisitions produced similar regional ventilation images. There was no evidence of a difference in VDPN4 between acquisition types (Cartesian = 9.1% ± 8.1; spiral = 9.3% ± 8.7; P = .97) with strong correlation (r2 = 0.95; P < .001) and no systemic bias (mean difference, -0.2%; 95% CI: 3.6, -3.9). FA correction removed coil-related inhomogeneities while preserving physiologic heterogeneity, including gravitational gradients that were removed by N4 correction (mean slope in healthy participants: FA-corrected = 0.026 SNorm/cm ± 0.013; N4-corrected = 0.002 SNorm/cm ± 0.001; P < .001). VDPN4 and VDPFA were strongly correlated with VDPRS (r2 = 0.94 and 0.95, respectively; P < .001 for both). Defect masks from FA-corrected images showed better agreement with reader segmentations compared with N4-corrected image-based defect masks (17% higher Dice score from FA-corrected images; mean Dice score: N4-corrected, 0.41 ± 0.31; FA-corrected, 0.48 ± 0.29; P =.001) and better depicted regional hypo- and hyperventilation. Conclusion Two-dimensional spiral acquisition combined with FA correction and mGLB analysis enabled rapid 129Xe ventilation MRI, effectively mitigating inhomogeneities while preserving physiologic heterogeneity. This approach provided accurate and efficient quantification of ventilation abnormalities in both healthy individuals and individuals with CF. Keywords: MRI, Pulmonary, Lung, Xenon, Ventilation Supplemental material is available for this article. © RSNA, 2025.

Intra-articular Injection of Bone Marrow Concentrate for Treatment of Patellofemoral Osteoarthritis: Preliminary Results Utilizing an Ultrasound-Guided Marrow Harvesting Technique

Purpose To develop and validate a deep learning model for automatic segmentation of primary central nervous system lymphoma (PCNSL) at postcontrast T1-weighted MRI. Materials and Methods Data were retrospectively collected from patients with pathologically proven immunocompetent PCNSL between September 2010 and February 2022. Postcontrast T1-weighted MRI scans were used to train and validate a deep learning model based on the nnU-Net framework. Manual segmentation by neuroradiologists served as the reference standard. The model was trained using an internal dataset from a single center and tested on both internal and external test sets from seven additional centers. Performance was assessed using Dice score, mean average surface distance, and F1 score. Statistical comparisons were performed using Mann-Whitney U test and bootstrap resampling for CIs. Results The study included 135 patients (68 female, 66 male, and one of unspecified sex; internal dataset: mean age ±SD, 67.0 years ± 12.0; external dataset: mean age, 75.5 years ± 13.6). The model achieved a mean Dice score of 0.84 (95% CI: 0.79, 0.88) on the internal test set (n = 44) and 0.88 (95% CI: 0.84, 0.91) on the external test set (n = 48), with no evidence of a difference between test sets (P = .59). Performance varied by lesion type; accuracy was highest in homogeneous discrete lesions, and performance was slightly decreased when numerous poorly defined infracentimetric lesions occurred. Strong volumetric correlation was observed between automatic and manual segmentations (internal: r = 0.99, P < .001; external: r = 0.98, P < .001). Conclusion A deep learning model achieved accurate and robust automatic segmentation of PCNSL across multiple clinical centers with different MRI acquisition parameters. Keywords: Brain Lymphoma, Brain Tumor, Automatic Segmentation, Artificial Intelligence, Deep Learning Supplemental material is available for this article. © RSNA, 2025.

Comparison Between Visual Assessment of MTA and Hippocampal Volumes in a Non-Demented Elderly Population

Based on a 3D pre-treatment magnetic resonance (MR) scan, we developed DREME-MR to jointly reconstruct the reference patient anatomy and solve a data-driven, patient-specific cardiorespiratory motion model. Via a motion encoder simultaneously learned during the reconstruction, DREME-MR further enables real-time volumetric MR imaging and cardiorespiratory motion tracking with minimal intra-treatment k-space data.&#xD;&#xD;Approach: DREME-MR integrates dynamic MRI reconstruction and real-time MR imaging into a unified, dual-task learning framework. From a 3D radial-spoke-based pre-treatment MR scan, DREME-MR uses spatiotemporal implicit-neural-representation (INR) to reconstruct pre-treatment dynamic volumetric MR images (learning task 1). The INR-based reconstruction takes a joint image reconstruction and deformable registration approach, yielding a reference anatomy and a corresponding cardiorespiratory motion model. The motion model adopts a low-rank, multi-resolution representation to approximate motion fields as products of motion coefficients and motion basis components (MBCs). Via a progressive, frequency-guided strategy, DREME-MR decouples cardiac MBCs from respiratory MBCs to resolve the two distinct motion modes. Simultaneously with the pre-treatment dynamic MRI reconstruction, DREME-MR also trains a multilayer perceptron (MLP)-based motion encoder to infer cardiorespiratory motion coefficients directly from the raw k-space data (learning task 2), allowing real-time, intra-treatment volumetric MR imaging and motion tracking with minimal k-space data (20-30 spokes) acquired after the pre-treatment MRI scan.&#xD;&#xD;Main results: Evaluated using data from a digital phantom (XCAT) and a human scan, DREME-MR solves real-time 3D cardiorespiratory motion with a latency of < 165 ms (= 150-ms data acquisition + 15-ms inference time), fulfilling the temporal constraint of real-time imaging. The XCAT study achieves mean(±S.D.) center-of-mass tracking errors of 0.73±0.38mm for a lung tumor and 1.69±1.12mm for the left ventricle. The human study shows good motion correlations (liver: 0.96; left ventricle: 0.65) between DREME-MR-solved motion and extracted surrogate signals.&#xD;&#xD;Significance: DREME-MR allows real-time 3D MRI and cardiorespiratory motion tracking with low latency, advancing intra-treatment MR-guided adaptive radiotherapy, including real-time multileaf collimator (MLC) tracking.

ABSTRACTObjective: The present study was an attempt to investigate tongue/mandible volume ratio in children, using volumetric magnetic resonance imaging (MRI) for early screening and to aid in treatment planning. Methods: Volumetric evaluation of tongue volume/mandible volume ratio (TV/MV ratio) in children with obstructive sleep apnea (OSA) using MRI was carried out retrospectively on available DICOM MR images of children in the age group of 10-14 years. MRI image records of patients diagnosed with OSA were obtained from interventional radiology department records, at Sharad Pawar Dental College and Hospital (Datta Meghe Institute of Medical Sciences, Nagpur/India). The age, gender, height and weight of the subjects were retrieved from patient database and registered. For the control group, available MRI images of healthy subjects without OSA were retrieved. Body mass index (BMI) was also calculated using the height and the weight present in the records. Measurements from MR images were made using DICOM image processing software. Soft tissue and bony structure segmentation was performed by manual tracing. The tongue volume and mandible volume were directly computed using the software. The tongue volume/mandible volume ratio (TV/MV) was generated using the above values and expressed as a percentage for both groups. Results: The difference between OSA group and control group with respect to TV/MV ratio was found to be highly significant at 0.05 level of significance. There was no significant correlation between BMI and TV/MV ratio in OSA group (p= 0.451) as well as in control group (p= 0.094). Conclusion: TV/MV ratio may be an appropriate variable to evaluate the risk of OSA, representing the balance between skeletal morphology and soft tissue morphology in craniofacial complex.

While deep learning has catalyzed breakthroughs across numerous domains, its broader adoption in clinical settings is inhibited by the costly and time-intensive nature of data acquisition and annotation. To further facilitate medical machine learning, we present an ultrasound dataset of 10,223 brightness-mode (B-mode) images consisting of sagittal slices of porcine spinal cords (N = 25) before and after a contusion injury. We additionally benchmark the performance metrics of several state-of-the-art object detection algorithms to localize the site of injury and semantic segmentation models to label the anatomy for comparison and creation of task-specific architectures. Finally, we evaluate the zero-shot generalization capabilities of the segmentation models on human ultrasound spinal cord images to determine whether training on our porcine dataset is sufficient for accurately interpreting human data. Our results show that the YOLOv8 detection model outperforms all evaluated models for injury localization, achieving a mean Average Precision (mAP50-95) score of 0.606. Segmentation metrics indicate that the DeepLabv3 segmentation model achieves the highest accuracy on unseen porcine anatomy, with a Mean Dice score of 0.587, while SAMed achieves the highest mean Dice score generalizing to human anatomy (0.445). To the best of our knowledge, this is the largest annotated dataset of spinal cord ultrasound images made publicly available to researchers and medical professionals, as well as the first public report of object detection and segmentation architectures to assess anatomical markers in the spinal cord for methodology development and clinical applications.

Several hundreds of thousand humans are diagnosed with brain cancer every year, and the majority dies within the next two years. The chances of survival could be easiest improved by early diagnosis. This is why there is a strong need for reliable algorithms that can detect the presence of gliomas in their early stage. While an automatic tumor detection algorithm can support a mass screening system, the precise segmentation of the tumor can assist medical staff at therapy planning and patient monitoring. This paper presents a random forest based procedure trained to segment gliomas in multispectral volumetric MRI records. Beside the four observed features, the proposed solution uses 100 further features extracted via morphological operations and Gabor wavelet filtering. A neighborhood-based post-processing was designed to regularize and improve the output of the classifier. The proposed algorithm was trained and tested separately with the 54 low-grade and 220 high-grade tumor volumes of the MICCAI BRATS 2016 training database. For both data sets, the achieved accuracy is characterized by an overall mean Dice score &gt; 83%, sensitivity &gt; 85%, and specificity &gt; 98%. The proposed method is likely to detect all gliomas larger than 10 mL.

Vision transformers (ViTs) have revolutionized computer vision by employing self-attention instead of convolutional neural networks and demonstrated success due to their ability to capture global dependencies and remove spatial biases of locality. In medical imaging, where input data may differ in size and resolution, existing architectures require resampling or resizing during pre-processing, leading to potential spatial resolution loss and information degradation. This study proposes a co-ordinate-based embedding that encodes the geometry of medical images, capturing physical co-ordinate and resolution information without the need for resampling or resizing. The effectiveness of the proposed embedding is demonstrated through experiments with UNETR and SwinUNETR models for infarct segmentation on MRI dataset with AxTrace and AxADC contrasts. The dataset consists of 1142 training, 133 validation and 143 test subjects. Both models with the addition of co-ordinate based positional embedding achieved substantial improvements in mean Dice score by 6.5% and 7.6%. The proposed embedding showcased a statistically significant advantage p-value< 0.0001 over alternative approaches. In conclusion, the proposed co-ordinate-based pixel-wise positional embedding method offers a promising solution for Transformer-based models in medical image analysis. It effectively leverages physical co-ordinate information to enhance performance without compromising spatial resolution and provides a foundation for future advancements in positional embedding techniques for medical applications.

The amount of epicardial fat around the heart has a significant impact on cardiovascular function and requires precise measurement for timely treatment. In this work, an improved U-Net architecture is proposed for accurate segmentation of epicardial fat in computer tomography (CT) images. The proposed method integrates a modified squeeze-and-excitation (MSE) block and a multi-scale dense (MS-D) convolutional neural network (CNN) to improve feature extraction. In addition, a metaheuristic optimization algorithm from falcon optimization algorithm (FOA) is used for efficient feature selection. The selected features are then classified using the XGBoost algorithm to determine the fat severity. Experimental evaluations on a CT image dataset show the superior segmentation performance of the proposed U-Net compared to existing architectures. It achieves a mean intersection over union (Mean IOU) of 89.5 %, a mean Dice score (MDS) of 94.3 %, and a Pearson correlation coefficient (PCC) of 0.973. FOA-guided feature selection further increases the accuracy of severity classification. The overall classification accuracy of the model is 98 %. These results highlight the technological advancements and measurement accuracy of the proposed U-Net architecture. They also demonstrate the suitability of the model to improve cardiovascular risk assessment and management.

Quantitative Assessment of Fluorescein Angiography Leakage via Deep Learning in Pediatric Uveitis: Correlation with Clinical Parameters