Medical Imaging Research Articles

Accelerated muscle mass estimation from CT images through transfer learning

BackgroundThe cost of labeling to collect training data sets using deep learning is especially high in medical applications compared to other fields. Furthermore, due to variances in images depending on the computed tomography (CT) devices, a deep learning based segmentation model trained with a certain device often does not work with images from a different device.MethodsIn this study, we propose an efficient learning strategy for deep learning models in medical image segmentation. We aim to overcome the difficulties of segmentation in CT images by training a VNet segmentation model which enables rapid labeling of organs in CT images with the model obtained by transfer learning using a small number of manually labeled images, called SEED images. We established a process for generating SEED images and conducting transfer learning a model. We evaluate the performance of various segmentation models such as vanilla UNet, UNETR, Swin-UNETR and VNet. Furthermore, assuming a scenario that a model is repeatedly trained with CT images collected from multiple devices, in which is catastrophic forgetting often occurs, we examine if the performance of our model degrades.ResultsWe show that transfer learning can train a model that does a good job of segmenting muscles with a small number of images. In addition, it was confirmed that VNet shows better performance when comparing the performance of existing semi-automated segmentation tools and other deep learning networks to muscle and liver segmentation tasks. Additionally, we confirmed that VNet is the most robust model to deal with catastrophic forgetting problems.ConclusionIn the 2D CT image segmentation task, we confirmed that the CNN-based network shows better performance than the existing semi-automatic segmentation tool or latest transformer-based networks.

Accurate pneumoconiosis staging via deep texture encoding and discriminative representation learning.

Accurate pneumoconiosis staging is key to early intervention and treatment planning for pneumoconiosis patients. The staging process relies on assessing the profusion level of small opacities, which are dispersed throughout the entire lung field and manifest as fine textures. While conventional convolutional neural networks (CNNs) have achieved significant success in tasks such as image classification and object recognition, they are less effective for classifying fine-grained medical images due to the need for global, orderless feature representation. This limitation often results in inaccurate staging outcomes for pneumoconiosis. In this study, we propose a deep texture encoding scheme with a suppression strategy designed to capture the global, orderless characteristics of pneumoconiosis lesions while suppressing prominent regions such as the ribs and clavicles within the lung field. To further enhance staging accuracy, we incorporate an ordinal label distribution to capture the ordinal information among profusion levels of opacities. Additionally, we employ supervised contrastive learning to develop a more discriminative feature space for downstream classification tasks. Finally, in accordance with standard practices, we evaluate the profusion levels of opacities in each subregion of the lung, rather than relying on the entire chest X-ray image. Experimental results on the pneumoconiosis dataset demonstrate the superior performance of the proposed method confirming its effectiveness for accurate pneumoconiosis staging.

3D Printing Materials Mimicking Human Tissues after Uptake of Iodinated Contrast Agents for Anthropomorphic Radiology Phantoms.

(1) Background: 3D printable materials with accurately defined iodine content enable the development and production of radiological phantoms that simulate human tissues, including lesions after contrast administration in medical imaging with X-rays. These phantoms provide accurate, stable and reproducible models with defined iodine concentrations, and 3D printing allows maximum flexibility and minimal development and production time, allowing the simulation of anatomically correct anthropomorphic replication of lesions and the production of calibration and QA standards in a typical medical research facility. (2) Methods: Standard printing resins were doped with an iodine contrast agent and printed using a consumer 3D printer, both (resins and printer) available from major online marketplaces, to produce printed specimens with iodine contents ranging from 0 to 3.0% by weight, equivalent to 0 to 3.85% elemental iodine per volume, covering the typical levels found in patients. The printed samples were scanned in a micro-CT scanner to measure the properties of the materials in the range of the iodine concentrations used. (3) Results: Both mass density and attenuation show a linear dependence on iodine concentration (R2 = 1.00), allowing highly accurate, stable, and predictable results. (4) Conclusions: Standard 3D printing resins can be doped with liquids, avoiding the problem of sedimentation, resulting in perfectly homogeneous prints with accurate dopant content. Iodine contrast agents are perfectly suited to dope resins with appropriate iodine concentrations to radiologically mimic tissues after iodine uptake. In combination with computer-aided design, this can be used to produce printed objects with precisely defined iodine concentrations in the range of up to a few percent of elemental iodine, with high precision and anthropomorphic shapes. Applications include radiographic phantoms for detectability studies and calibration standards in projective X-ray imaging modalities, such as contrast-enhanced dual energy mammography (abbreviated CEDEM, CEDM, TICEM, or CESM depending on the equipment manufacturer), and 3-dimensional modalities like CT, including spectral and dual energy CT (DECT), and breast tomosynthesis.

Medical Image Segmentation : A Comparative Study and Survey

Image segmentation involves dividing an image into distinct segments. It is important to distinguish this from image enhancement, which focuses on improving the image’s visual attributes, such as brightness, contrast, and texture. In segmentation, specific parts of the image are emphasized based on the problem being addressed. Clinical imaging techniques are continuously evolving, aiming to improve the quality of services in the healthcare industry. Methods such as interpolation, image registration, compression, and diagnosis need to be updated to meet the increasing demands of the field, especially with advancements in mobile and cloud computing technologies. The integration of medical devices with wearable technologies presents a promising area for further exploration. This paper provides valuable insights into the domain of medical imaging systems and aims to outline the future potential of work in this field.

Retraction Note: Application of optical medical equipment imaging detection and fire needle therapy for intestinal mucosal barrier function in patients with intestinal diseases

Retraction Note: Application of optical medical equipment imaging detection and fire needle therapy for intestinal mucosal barrier function in patients with intestinal diseases

Innovative Scientific Discoveries: The Role of Intelligent Computing in the Fifth Paradigm Shift

This article explores the key role of intelligent computing in driving the paradigm shift of scientific discovery. The article first outlines the five paradigms of scientific discovery, from empirical observation to theoretical models, then to computational simulation and data intensive science, and finally introduces intelligent computing as the core of the fifth paradigm. Intelligent computing enhances the ability to understand, predict, and automate scientific discoveries of complex systems through technologies such as deep learning and machine learning. The article further analyzes the applications of intelligent computing in fields such as bioinformatics, astronomy, climate science, materials science, and medical image analysis, demonstrating its practical utility in solving scientific problems and promoting knowledge development. Finally, the article predicts that intelligent computing will play a more critical role in future scientific research, promoting interdisciplinary integration, open science, and collaboration, providing new solutions for solving complex problems.

Tracing proliferating potential of lung tumours by [11C]Nicotinamide

Many tumours rely on aerobic glycolysis, known as the Warburg effect, to satisfy energy needs and sustain rapid proliferation [1]. High demands of glucose in tumours can be measured by [18F[fluorodeoxyglucose, [18F]FDG, positron emission tomography (PET). Nicotinamide adenine dinucleotide (NAD) is a coenzyme involved in redox reactions in many metabolic pathways, including glycolysis. To maintain the high rates of glycolysis and proliferation, tumours need increased NAD+ levels and use nicotinamide for the upregulated NAD+ salvage pathway that is the main source of NAD+ in proliferative cancer cells [2]. Herein, we show an improved synthesis method of [11C]Nicotinamide, [11C]NAM, and studied the uptake of [11C]NAM and [18F]FDG to measure proliferating capacity in a lung cancer model. [Carboxyl-11C]nicotinamide was synthesized by a one-pot radiolabeling method [3] on the SYNTHRA synthesizer (Synthra GmbH, Germany) (see Figure 1A). We used a KrasG12D/+ genetically engineered mouse model of lung cancer and evaluated the uptake of [11C]nicotinamide and [18F]FDG with dual-tracer sequential imaging protocol (see Figure 1B) using a NanoScan PET/MRI scanner (Mediso Medical Imaging Systems, Hungary). We obtained [carboxyl-11C]nicotinamide in a radiochemical yield of 15 ± 5 %, volumic activity of 50 ± 10 MBq/ml, and radiochemical purity was > 99 %. We optimized the imaging protocol and found the uptake of [11C]nicotinamide co-localized with the uptake of [18F]FDG in the KrasG12D/+ model of lung cancer as shown in Figure 1C suggesting the higher NAM uptake was associated with high NAD+ production in highly proliferating tumours. We demonstrated that metabolic PET/MRI imaging with [11C]NAM can complement [18F]FDG PET and be useful for probing proliferative potential in lung cancer. This method could be translated and applied in clinical practice to enable visualization of NAD+ metabolism in lung tumours and targeted therapeutically. Please click on the 'PDF' for the full abstract!

Radiomic Applications in Skull Base Pathology: A Systematic Review of Potential Clinical Uses

Objectives Radiomics involves the extraction and analysis of numerous quantitative features of medical imaging which can add more information from radiological images often beyond initial comprehension of a clinician. Unlike deep learning, radiomics allows some understanding of identified quantitative features for clinical prediction. We sought to explore the current state of radiomics applications in the skull base literature. Methods A systematic review of studies evaluating radiomics in skull base was performed, including those with and without machine-learning approaches. Studies were summarized into thematic elements as well as specific pathologies. Results A total of 102 studies with 26,280 radiographic images were included. The earliest radiomic study was published in 2017 with exponential growth in research since then. Most studies focused on tumor diagnosis (40.8%), followed by tumor prognosis (31.1%), automated segmentation (16.5%), other applications (7.8%), and lastly prediction of intraoperative features (3.9%). Pituitary adenomas (41.7%) and vestibular schwannomas (18.4%) represented the most commonly evaluated pathologies; however, radiomics could be applied to a heterogeneous collection of skull base pathologies. The average study included 258677 cases (range 4; 6755). Conclusion Radiomics offers many functions in treating skull base pathology and will likely be an essential component of future clinical care. Larger sample sizes, validation of predictive models, and clinical application are needed. Further investigation into the strengths and weaknesses of radiomic applications in skull base treatments is warranted.

Automatic detection and segmentation of lesions in 18 F-FDG PET/CT imaging of patients with Hodgkin lymphoma using 3D dense U-Net.

The accuracy of automatic tumor segmentation in PET/computed tomography (PET/CT) images is crucial for the effective treatment and monitoring of Hodgkin lymphoma. This study aims to address the challenges faced by certain segmentation algorithms in accurately differentiating lymphoma from normal organ uptakes due to PET image resolution and tumor heterogeneity. Variants of the encoder-decoder architectures are state-of-the-art models for image segmentation. Among these kinds of architectures, U-Net is one of the most famous and predominant for medical image segmentation. In this study, we propose a fully automatic approach for Hodgkin lymphoma segmentation that combines U-Net and DenseNet architectures to reduce network loss for very small lesions, which is trained using the Tversky loss function. The hypothesis is that the fusion of these two deep learning models can improve the accuracy and robustness of Hodgkin lymphoma segmentation. A dataset with 141 samples was used to train our proposed network. Also, to test and evaluate the proposed network, we allocated two separate datasets of 20 samples. We achieved 0.759 as the mean Dice similarity coefficient with a median value of 0.767, and interquartile range (0.647-0.837). A good agreement was observed between the ground truth of test images against the predicted volume with precision and recall scores of 0.798 and 0.763, respectively. This study demonstrates that the integration of U-Net and DenseNet architectures, along with the Tversky loss function, can significantly enhance the accuracy of Hodgkin lymphoma segmentation in PET/CT images compared to similar studies.

GestaltMatcher Database - A global reference for facial phenotypic variability in rare human diseases.

The most important factor that complicates the work of dysmorphologists is the significant phenotypic variability of the human face. Next-Generation Phenotyping (NGP) tools that assist clinicians with recognizing characteristic syndromic patterns are particularly challenged when confronted with patients from populations different from their training data. To that end, we systematically analyzed the impact of genetic ancestry on facial dysmorphism. For that purpose, we established the GestaltMatcher Database (GMDB) as a reference dataset for medical images of patients with rare genetic disorders from around the world. We collected 10,980 frontal facial images - more than a quarter previously unpublished - from 8,346 patients, representing 581 rare disorders. Although the predominant ancestry is still European (67%), data from underrepresented populations have been increased considerably via global collaborations (19% Asian and 7% African). This includes previously unpublished reports for more than 40% of the African patients. The NGP analysis on this diverse dataset revealed characteristic performance differences depending on the composition of training and test sets corresponding to genetic relatedness. For clinical use of NGP, incorporating non-European patients resulted in a profound enhancement of GestaltMatcher performance. The top-5 accuracy rate increased by +11.29%. Importantly, this improvement in delineating the correct disorder from a facial portrait was achieved without decreasing the performance on European patients. By design, GMDB complies with the FAIR principles by rendering the curated medical data findable, accessible, interoperable, and reusable. This means GMDB can also serve as data for training and benchmarking. In summary, our study on facial dysmorphism on a global sample revealed a considerable cross ancestral phenotypic variability confounding NGP that should be counteracted by international efforts for increasing data diversity. GMDB will serve as a vital reference database for clinicians and a transparent training set for advancing NGP technology.

Medical image segmentation method based on confidence learning and collaborative training

Confidence learning plays an important role in the training of low-quality annotated data of medical images. However, the current applications of confidence learning are based on the average teacher model, and the possibility of confidence learning on other networks has not been explored. To address this problem, a segmentation model based on confidence learning and co-training is proposed. The model uses two different networks to encourage the output results of the two networks to be consistent. Then, the output of one of the networks is compared with the original low-quality annotations using confidence learning, and the low-quality annotated data is corrected to provide an effective training reference. The proposed model is compared on three medical image datasets of different modalities. Experimental results show that the segmentation effect of the proposed model is better than that of the existing confidence learning model.

The association between patient age, abscess size, and white blood cell count on duration of catheter stay for percutaneous abscess drainage of abdominal abscesses

ObjectiveKnowing factors that impact catheter stay duration is important since removing drainage catheters too early or late can have significant consequences. We present a single center retrospective study that analyzes multiple variables, including abscess size, white blood cell count, and patient age, to see if a correlation between them and duration of catheter stay exists. The inclusion criteria were patients that had abdominal abscesses treated with percutaneous abscess drainage using a pigtail catheter, ≥ 18 years of age, and had available medical images and records. 44 patients were included.ResultsAmong white blood cell count, patient age, and abscess volume, the only significant relationship with duration of catheter stay was abscess size (R = 0.42, p-value = 0.0049).

Empowering Brain Tumor Diagnosis through Explainable Deep Learning

Brain tumors are among the most lethal diseases, and early detection is crucial for improving patient outcomes. Currently, magnetic resonance imaging (MRI) is the most effective method for early brain tumor detection due to its superior imaging quality for soft tissues. However, manual analysis of brain MRI scans is prone to errors, largely influenced by the radiologists’ experience and fatigue. To address these challenges, computer-aided diagnosis (CAD) systems are more significant. These advanced computer vision techniques such as deep learning provide accurate predictions based on medical images, enhancing diagnostic precision and reliability. This paper presents a novel CAD framework for multi-class brain tumor classification. The framework employs six pre-trained deep learning models as the base and incorporates comprehensive data preprocessing and augmentation strategies to enhance computational efficiency. To address issues related to transparency and interpretability in deep learning models, Gradient-weighted Class Activation Mapping (Grad-CAM) is utilized to visualize the decision-making processes involved in tumor classification from MRI scans. Additionally, a user-friendly Brain Tumor Detection System has been developed using Streamlit, demonstrating its practical applicability in real-world settings and providing a valuable tool for clinicians. All simulation results are derived from a public benchmark dataset, showing that the proposed framework achieves state-of-the-art performance, with accuracy approaching 99% in ResNet-50, Xception, and InceptionV3 models.

Deep Conformal Supervision: Leveraging Intermediate Features for Robust Uncertainty Quantification.

Trustworthiness is crucial for artificial intelligence (AI) models in clinical settings, and a fundamental aspect of trustworthy AI is uncertainty quantification (UQ). Conformal prediction as a robust uncertainty quantification (UQ) framework has been receiving increasing attention as a valuable tool in improving model trustworthiness. An area of active research is the method of non-conformity score calculation for conformal prediction. We propose deep conformal supervision (DCS), which leverages the intermediate outputs of deep supervision for non-conformity score calculation, via weighted averaging based on the inverse of mean calibration error for each stage. We benchmarked our method on two publicly available datasets focused on medical image classification: a pneumonia chest radiography dataset and a preprocessed version of the 2019 RSNA Intracranial Hemorrhage dataset. Our method achieved mean coverage errors of 16e-4 (CI: 1e-4, 41e-4) and 5e-4 (CI: 1e-4, 10e-4) compared to baseline mean coverage errors of 28e-4 (CI: 2e-4, 64e-4) and 21e-4 (CI: 8e-4, 3e-4) on the two datasets, respectively (p < 0.001 on both datasets). Based on our findings, the baseline results of conformal prediction already exhibit small coverage errors. However, our method shows a significant improvement on coverage error, particularly noticeable in scenarios involving smaller datasets or when considering smaller acceptable error levels, which are crucial in developing UQ frameworks for healthcare AI applications.

-New frontiers in domain-inspired radiomics and radiogenomics: increasing role of molecular diagnostics in CNS tumor classification and grading following WHO CNS-5 updates.

Gliomas and Glioblastomas represent a significant portion of central nervous system (CNS) tumors associated with high mortality rates and variable prognosis. In 2021, the World Health Organization (WHO) updated its Glioma classification criteria, most notably incorporating molecular markers including CDKN2A/B homozygous deletion, TERT promoter mutation, EGFR amplification, + 7/-10 chromosome copy number changes, and others into the grading and classification of adult and pediatric Gliomas. The inclusion of these markers and the corresponding introduction of new Glioma subtypes has allowed for more specific tailoring of clinical interventions and has inspired a new wave of Radiogenomic studies seeking to leverage medical imaging information to explore the diagnostic and prognostic implications of these new biomarkers. Radiomics, deep learning, and combined approaches have enabled the development of powerful computational tools for MRI analysis correlating imaging characteristics with various molecular biomarkers integrated into the updated WHO CNS-5 guidelines. Recent studies have leveraged these methods to accurately classify Gliomas in accordance with these updated molecular-based criteria based solely on non-invasive MRI, demonstrating the great promise of Radiogenomic tools. In this review, we explore the relative benefits and drawbacks of these computational frameworks and highlight the technical and clinical innovations presented by recent studies in the landscape of fast evolving molecular-based Glioma subtyping. Furthermore, the potential benefits and challenges of incorporating these tools into routine radiological workflows, aiming to enhance patient care and optimize clinical outcomes in the evolving field of CNS tumor management, have been highlighted.

Transformative insights: Image-based breast cancer detection and severity assessment through advanced AI techniques

Abstract In the realm of image-based breast cancer detection and severity assessment, this study delves into the revolutionary potential of sophisticated artificial intelligence (AI) techniques. By investigating image processing, machine learning (ML), and deep learning (DL), the research illuminates their combined impact on transforming breast cancer diagnosis. This integration offers insights into early identification and precise characterization of cancers. With a foundation in 125 research articles, this article presents a comprehensive overview of the current state of image-based breast cancer detection. Synthesizing the transformative role of AI, including image processing, ML, and DL, the review explores how these technologies collectively reshape the landscape of breast cancer diagnosis and severity assessment. An essential aspect highlighted is the synergy between advanced image processing methods and ML algorithms. This combination facilitates the automated examination of medical images, which is crucial for detecting minute anomalies indicative of breast cancer. The utilization of complex neural networks for feature extraction and pattern recognition in DL models further enhances diagnostic precision. Beyond diagnostic improvements, the abstract underscores the substantial influence of AI-driven methods on breast cancer treatment. The integration of AI not only increases diagnostic precision but also opens avenues for individualized treatment planning, marking a paradigm shift toward personalized medicine in breast cancer care. However, challenges persist, with issues related to data quality and interpretability requiring continued research efforts. Looking forward, the abstract envisions future directions for breast cancer identification and diagnosis, emphasizing the adoption of explainable AI techniques and global collaboration for data sharing. These initiatives promise to propel the field into a new era characterized by enhanced efficiency and precision in breast cancer care.

ANALYZING THE RELATIONSHIP AMONG CLINICAL SYMPTOMS, X-RAY RESULTS, AND CARTILAGE WEAR PATTERNS IN KNEE OSTEOARTHRITIS USING T2 MAPPING MRI

Objectives: Osteoarthritis (OA) is a condition that causes joint pain, varying degrees of functional limitations, and a decrease in overall quality of life. The purpose of this study was to examine the progression of cartilage degeneration in knees affected by symptomatic OA and determine its relationship with X-ray and T2 map MRI findings. Methods: This study was conducted at a Tertiary Care Teaching Institute in India and involved 40 patients. It was an observational type of study. For our assessment, we utilized the Western Ontario and McMaster University (WOMAC) osteoarthrosis index to evaluate function, quality of life, and joint pain. Medical imaging technique A standing AP view of the affected knee joint was taken using plain radiographs. Once the X-rays were obtained, they were graded using the Kellgren–Lawrence Classification, which ranges from 0 to 4 grades. Results: Out of the total patients, 6.0% were &lt;40 years old, 21.2% were between 41 and 50 years old, 27.2% were between 51 and 60 years old, 39.3% were between 61 and 70 years old, and 6.0% were above 70 years old. In this study, there were 21 females (63.6%) and 12 males (36.4%). In this study, the researchers found that there was no distinction between the sides of the limb affected. Both the left and right sides were involved in 13 cases each, accounting for 39.39% of the total cases. In addition, 7 cases, or 21.21%, reported experiencing bilateral knee pain. Merchant grade reported that out of the knees examined, 17.5% had Grade 0, 45% had Grade 1, 27.5% had Grade 2, 10% had Grade 3, and none had Grade 4. Conclusion: The findings of this study suggest that T2 map MRI has the ability to accurately measure abnormal cartilage changes in conditions that affect the mechanical properties of the knee joint. It is crucial to have non-invasive methods that can assess early cartilage matrix changes. These methods are important for initiating early treatment, monitoring disease progression, planning operative procedures, and following up on operative cartilage repair.

A review of deep learning approaches for multimodal image segmentation of liver cancer.

This review examines the recent developments in deep learning (DL) techniques applied to multimodal fusion image segmentation for liver cancer. Hepatocellular carcinoma is a highly dangerous malignant tumor that requires accurate image segmentation for effective treatment and disease monitoring. Multimodal image fusion has the potential to offer more comprehensive information and more precise segmentation, and DL techniques have achieved remarkable progress in this domain. This paper starts with an introduction to liver cancer, then explains the preprocessing and fusion methods for multimodal images, then explores the application of DL methods in this area. Various DL architectures such as convolutional neural networks (CNN) and U-Net are discussed and their benefits in multimodal image fusion segmentation. Furthermore, various evaluation metrics and datasets currently used to measure the performance of segmentation models are reviewed. While reviewing the progress, the challenges of current research, such as data imbalance, model generalization, and model interpretability, are emphasized and future research directions are suggested. The application of DL in multimodal image segmentation for liver cancer is transforming the field of medical imaging and is expected to further enhance the accuracy and efficiency of clinical decision making. This review provides useful insights and guidance for medical practitioners.

A Hybrid Approach for Robust Object Detection: Integrating Template Matching and Faster R-CNN

Object detection is a critical task in computer vision, with applications ranging from autonomous vehicles to medical imaging. Traditional methods like template matching offer precise localization but struggle with variations in object appearance, while deep learning approaches such as Faster R-CNN excel in handling diverse and complex datasets but often require extensive computational resources and large amounts of labeled data. This paper proposes a hybrid approach that integrates template matching with Faster R-CNN to leverage the strengths of both techniques. By combining the accuracy of template matching with the robustness and generalization capabilities of Faster R-CNN, our method achieves superior performance in challenging scenarios, including objects with occlusions, varying scales, and complex backgrounds. Extensive experiments demonstrate that the hybrid model not only enhances detection accuracy but also reduces computational load, making it a practical solution for real-world applications.

Explainable AI for Transparent MRI Segmentation: Deep Learning and Visual Attribution in Clinical Decision Support

For medical diagnosis and therapy planning, the importance of accurate MRI segmentation cannot be overemphasized. Conversely, the inscrutability of deep learning models remains obstacles to their application in therapeutic contexts. In this article, an interpretability artificial intelligence framework is introduced. It combines an MRI segmentation model based on deep learning, visual attribution algorithms and natural language explanations. EXPERIMENT The dataset is consisting of plenty of different types of brain MRI scans, and used to test the architecture. The average of Dice score of our method is 88.7% and 92.3% for segmentation of tumor and categorization of tissues, respectively. Both are pretty epic scores. The insights extracted from both the visual attributions and textual explations improve our understanding of how the model arrives at its decisions, thereby increasing the transparency and interpretability of the model. believe this approach to explainable artificial intelligence can help to close the gap between state-of-the-art performance in MRI segmentation and clinical interpretability, by increasing the transparency of complex models and facilitating their implementation into a clinical workflow. Conclusion Our approach may have implications in the transparent and reliable development of AI-based decision support systems for medical imaging