Meta-Radiology最新文献

Schizophrenia is among top disability causes worldwide; however, its pathological mechanism is still unclear. In order to advance the research progress on schizophrenia, the Schizophrenia Imaging Lab has been dedicated to the diagnosis, treatment, and prevention of schizophrenia and conducted systematic studies. One of the recent studies involved the Allen Human Brain Atlas in the imaging study of schizophrenia. The combination of the two methods mitigated the gap between the human brain imaging and genetic mechanism, and enabled us to observe the transcriptome alterations and brain changes at the same time. This review concluded the studies that combined the imaging method and the Allen Human Brain Atlas application in schizophrenia. We aim to provide a holistic view of the brain changes in schizophrenia at the micro level.

Obesity is a highly prevalent and increasing global medical problem. It is expected that most radiologists will come across computed tomography studies of obese patients in their daily work. Obesity has multiple well known effects on the cardiovascular, endocrine, and musculoskeletal systems. Prevalent, but less well described, are the multiple effects that obesity causes in the lungs; this occurs through both direct mechanical and indirect metabolic mechanisms. These result in characteristic imaging features in CT in which this review article will illustrated. Radiologists who interpret chest CT should be aware of these findings and pitfalls in their assessment of obese patients to avoid misdiagnosis. In addition, there are multiple technical challenges to CT scanning of obese patients to achieve diagnostic images. In this review article the pathological mechanisms underlying the imaging findings in the obese lung are presented, as well as technical considerations for optimal scanning and the typical imaging findings. An overall review of the increasing use of AI body morphometry and its use in lung cancer risk and outcome prediction is also explored. We hope this review can provide clinical radiologists and those who have special interests in medical imaging comprehensive summary of the pathophysiology, diagnostic and technical challenges involved in thoracic CT imaging in obesity, as well as the insights into the future outlook with artificial intelligence.

There is a growing interest and adoption of 7 Tesla (T) magnetic resonance imaging (MRI) in the field of medicine and research. In the domain of neuroimaging, 7T MRI shows notable advantages over lower field strength MRI systems by offering improved visualization of anatomical structures, enhanced lesion conspicuity, and better characterization of pathological processes. Cerebrovascular disease, which involves a spectrum of etiologies from large artery abnormalities to small vessel disease, is a leading cause of morbidity and mortality worldwide. Imaging plays an indispensable role in the diagnosis and treatment of cerebrovascular diseases. The excellence in imaging capabilities of 7T MRI can achieve multi-scale, high-precision imaging requirements from large artery disease assessment to small vessel disease assessment, which presents a variety of clinical applications and significant potential for clinical transformation. In this review, we firstly reviewed the literature focusing on technique aspects, comparing 7T with the clinically well-established 3T and 1.5T MRI systems. Then, we reviewed published studies to showcase the state-of-the-art progress in the assessment of cerebrovascular disease at 7T. Additionally, we discussed the challenges and perspectives of 7T techniques.

Objectives

This study aimed to analyze each myocardial segment's ischemic burden, scarring, function, and viability by late gadolinium enhancement (LGE) imaging and stress-MRI using adenosine.

Materials and methods

Semi-quantitative and qualitative parameters of myocardial segments were obtained by stress-MRI. Moreover, segments without perfusion defect were defined as the no ischemic group, segments with a perfusion defect of ≤50% were defined as a low ischemic burden group, and segments with a perfusion defect of >50% were defined as a high ischemic burden group. “Segmental wall thickening (SWT)” was defined as the absolute difference between the end-diastolic and end-systolic wall thickness. Finally, viability was defined by dysfunctional myocardium (<3 ​mm segmental wall thickening [SWT]) and ≤50% late gadolinium enhancement (LGE).

Results

A total of 445 segments in the CTO territory were analyzed, scar tissue was found in the CTO territory, with LGE evident in 18.2% of CTO segments totaling >50%. Among the different ischemic burden groups, there were significant differences in LGE volume (p ​< ​0.01), and the trend of SWT was consistent with the degree of myocardial ischemia. The incidence of ≤50% LGE and viable myocardium was higher in segments of the no ischemia and low ischemic burden groups. However, there was no significant difference in the incidence of dysfunctional myocardial segments among the three groups (P ​> ​0.05).

Conclusions

Stress MRI parameters can accurately and detailly assess myocardial viability and function, so multi-parameter joint assessment of CTO patients by stress MRI may help in treatment decisions.

Objectives

To investigate the diagnostic value of a computed tomography enterography (CTE)-based radiomics model (RM) in the detection of active inflamXGmation in patients with intestinal Crohn's disease (CD).

Methods

CTE images and clinical data of 105 patients with pathologically diagnosed intestinal CD were retrospectively analyzed. Patients were divided into non-mild and moderate-severe activity groups based on histopathology. Among them, 84 cases were randomly divided into the training group (43 positive and 41 negative in an 8:2 ratio) and 21 into the experimental group (11 positive and 10 negative). All lesion areas on the venous-phase CTE image were delineated manually using ITK-Snap, features were extracted, and DARWIN software was used to reduce feature dimensionality. A binary RM using eXtreme Gradient Boosting (XGBOOST) was established to assess the test set, and sensitivity, specificity, and the area under the curve (AUC) were calculated. Finally, the AUC was used to evaluate the diagnostic efficacy and optimal diagnostic threshold of RM for active CD.

Results

In the training set, the AUC for RM to distinguish between non-mild and moderate-severe activity was 0.93, with a sensitivity of 90.2% and specificity of 83.7%. In the validation set, the AUC was 0.86, with a sensitivity of 90% and a specificity of 81.8%.

Conclusion

An imaging-omics model based on CTE can effectively evaluate CD activity.

Fairness of artificial intelligence and machine learning models, often caused by imbalanced datasets, has long been a concern. While many efforts aim to minimize model bias, this study suggests that traditional fairness evaluation methods may be biased, highlighting the need for a proper evaluation scheme with multiple evaluation metrics due to varying results under different criteria. Moreover, the limited data size of minority groups introduces significant data uncertainty, which can undermine the judgement of fairness. This paper introduces an innovative evaluation approach that estimates data uncertainty in minority groups through bootstrapping from majority groups for a more objective statistical assessment. Extensive experiments reveal that traditional evaluation methods might have drawn inaccurate conclusions about model fairness. The proposed method delivers an unbiased fairness assessment by adeptly addressing the inherent complications of model evaluation on imbalanced datasets. The results show that such comprehensive evaluation can provide more confidence when adopting those models.

Objectives

This study aimed to identify age-related changes in the fractal dimensions of the cerebellum and compare the sensitivity of fractal analysis and conventional Euclidean geometry-based morphometry to cerebellar aging.

Material and methods

Two-dimensional T2-weighted magnetic resonance images from the brains of 100 conditionally healthy individuals (44 males and 56 females) aged 18–86 years were examined, with a focus on mid-sagittal sections of the cerebellar vermis. We determined ten parameters derived from Euclidean geometry (perimeter, area, and indices calculated from them), along with seven fractal dimension values derived from fractal geometry (the approximated fractal dimensions of the overall cerebellar tissue, white matter, overall cerebellar cortex and its granular and molecular layers, outer contour, and digital skeleton).

Results

Fractal dimensions demonstrated stronger correlation relationships with age compared to morphometric parameters derived from Euclidean geometry. The most pronounced age-related declines were observed in the approximated fractal dimensions of the cerebellar cortex and its layers, with decreases also noted in the fractal dimensions of the outer contour and digital cerebellar skeleton. Fractal dimension values did not significantly differ between males and females, while several Euclidean geometry-derived parameters showed sexual dimorphism. Although males demonstrated stronger relationships of some studied parameters with age, there was no statistically significant difference in the sex-related dynamics of aging.

Conclusion

The normal aging of the cerebellum involves not only absolute size alterations but also changes in the texture and spatial configuration of cerebellar tissue components, which can be quantitatively and objectively assessed by fractal analysis.

Recent breakthroughs in artificial intelligence (AI) research include advancements in natural language processing (NLP) achieved by large language models (LLMs), and; in particular, generative pre–trained transformer (GPT) architectures. The latest GPT developed by OpenAI, GPT–4, has shown remarkable intelligence across various domains and tasks. It exhibits capabilities in abstraction, comprehension, vision, computer coding, mathematics, and more, suggesting it to be a significant step towards artificial general intelligence (AGI), a level of AI that possesses capabilities similar to human intelligence. This paper explores this AGI, its knowledge diffusive and societal influences, and its governance. In addition to coverage of the major associated topics studied in the literature, and making up for their loopholes, we scrutinize how GPT-4 can facilitate the diffusion of knowledge across different areas of science by promoting their interpretability and explainability (IE) to inexperts. Where applicable, the topics are also accompanied by their specific potential implications on medical imaging.

Diffusion probabilistic models (DPMs) which employ explicit likelihood characterization and a gradual sampling process to synthesize data, have gained increasing research interest. Despite their huge computational burdens due to the large number of steps involved during sampling, DPMs are widely appreciated in various medical imaging tasks for their high-quality and diversity of generation. Magnetic resonance imaging (MRI) is an important medical imaging modality with excellent soft tissue contrast and superb spatial resolution, which possesses unique opportunities for DPMs. Although there is a recent surge of studies exploring DPMs in MRI, a survey paper of DPMs specifically designed for MRI applications is still lacking. This review article aims to help researchers in the MRI community to grasp the advances of DPMs in different applications. We first introduce the theory of two dominant kinds of DPMs, categorized according to whether the diffusion time step is discrete or continuous, and then provide a comprehensive review of emerging DPMs in MRI, including reconstruction, image generation, image translation, segmentation, anomaly detection, and further research topics. Finally, we discuss the general limitations as well as limitations specific to the MRI tasks of DPMs and point out potential areas that are worth further exploration.