The generalization of deep learning-based low-dose computed tomography (CT) reconstruction models to doses unseen in the training data is important and remains challenging. Previous efforts heavily rely on paired data to improve the generalization performance and robustness through collecting either diverse CT data for re-training or a few test data for fine-tuning. Recently, diffusion models have shown promising and generalizable performance in low-dose CT (LDCT) reconstruction, however, they may produce unrealistic structures due to the CT image noise deviating from Gaussian distribution and imprecise prior information from the guidance of noisy LDCT images. In this paper, we propose a noise-inspired diffusion model for generalizable LDCT reconstruction, termed NEED, which tailors diffusion models for noise characteristics of each domain. First, we propose a novel shifted Poisson diffusion model to denoise projection data, which aligns the diffusion process with the noise model in pre-log LDCT projections. Second, we devise a doubly guided diffusion model to refine reconstructed images, which leverages LDCT images and initial reconstructions to more accurately locate prior information and enhance reconstruction fidelity. By cascading these two diffusion models for dual-domain reconstruction, our NEED requires only normal-dose data for training and can be effectively extended to various unseen dose levels during testing via a time step matching strategy. Extensive qualitative, quantitative, and segmentation-based evaluations on two datasets demonstrate that our NEED consistently outperforms state-of-the-art methods in reconstruction and generalization performance. Source code is made available at https://github.com/qgao21/NEED.

Early identification of bowel resection risks is crucial for patients with incarcerated inguinal hernia (IIH). However, the prompt detection of these risks remains a significant challenge. Advancements in radiomic feature extraction and machine learning algorithms have paved the way for innovative diagnostic approaches to assess IIH more effectively. To devise a sophisticated radiomic-clinical model to evaluate bowel resection risks in IIH patients, thereby enhancing clinical decision-making processes. This single-center retrospective study analyzed 214 IIH patients randomized into training (<i>n</i> = 161) and test (<i>n</i> = 53) sets (3:1). Radiologists segmented hernia sac-trapped bowel volumes of interest (VOIs) on computed tomography images. Radiomic features extracted from VOIs generated Rad-scores, which were combined with clinical data to construct a nomogram. The nomogram's performance was evaluated against standalone clinical and radiomic models in both cohorts. A total of 1561 radiomic features were extracted from the VOIs. After dimensionality reduction, 13 radiomic features were used with eight machine learning algorithms to develop the radiomic model. The logistic regression algorithm was ultimately selected for its effectiveness, showing an area under the curve (AUC) of 0.828 [95% confidence interval (CI): 0.753-0.902] in the training set and 0.791 (95%CI: 0.668-0.915) in the test set. The comprehensive nomogram, incorporating clinical indicators showcased strong predictive capabilities for assessing bowel resection risks in IIH patients, with AUCs of 0.864 (95%CI: 0.800-0.929) and 0.800 (95%CI: 0.669-0.931) for the training and test sets, respectively. Decision curve analysis revealed the integrated model's superior performance over standalone clinical and radiomic approaches. This innovative radiomic-clinical nomogram has proven to be effective in predicting bowel resection risks in IIH patients and has substantially aided clinical decision-making.

To train deep learning models to detect hypoxic-ischemic brain injury (HIBI) on early CT scans after pediatric out-of-hospital cardiac arrest (OHCA) and determine if models could identify HIBI that was not visually appreciable to a radiologist. Retrospective study of children who had a CT scan within 24 hours of OHCA compared to age-matched controls. We designed models to detect HIBI by discriminating CT images from OHCA cases and controls, and predict death and unfavorable outcome (PCPC 4-6 at hospital discharge) among cases. Model performance was measured by AUC. We trained a second model to distinguish OHCA cases with radiologist-identified HIBI from controls without OHCA and tested the model on OHCA cases without radiologist-identified HIBI. We compared outcomes between OHCA cases with and without model-categorized HIBI. We analyzed 117 OHCA cases (age 3.1 [0.7-12.2] years); 43% died and 58% had unfavorable outcome. Median time from arrest to CT was 2.1 [1.0,7.2] hours. Deep learning models discriminated OHCA cases from controls with a mean AUC of 0.87±0.05. Among OHCA cases, mean AUCs for predicting death and unfavorable outcome were 0.79±0.06 and 0.69±0.06, respectively. Mean AUC was 0.98±0.01for discriminating between 44 OHCA cases with radiologist-identified HIBI and controls. Among 73 OHCA cases without radiologist-identified HIBI, the model identified 36% as having presumed HIBI; 31% of whom died compared to 17% of cases without HIBI identified radiologically and via the model (p=0.174). Deep learning models can identify HIBI on early CT images after pediatric OHCA and detect some presumed HIBI visually not identified by a radiologist.

Resistance exercise, Taichi exercise, and the hybrid exercise program consisting of the two aforementioned methods have been demonstrated to increase the skeletal muscle mass of older individuals with sarcopenia. However, the exercise sequence has not been comprehensively investigated. Therefore, we designed a self-determined sequence exercise program, incorporating resistance exercises, Taichi, and the hybrid exercise program to overcome the decline of skeletal muscle area and reverse sarcopenia in older individuals. Ninety-one older patients with sarcopenia between the ages of 60 and 75 completed this three-stage randomized controlled trial for 24 weeks, including the self-determined sequence exercise program group (n = 31), the resistance training group (n = 30), and the control group (n = 30). We used quantitative computed tomography to measure the effects of different intervention protocols on skeletal muscle mass in participants. Participants' demographic variables were analyzed using one-way analysis of variance and chi-square tests, and experimental data were examined using repeated-measures analysis of variance. Furthermore, we utilized the Markov model to explain the effectiveness of the exercise programs among the three-stage intervention and explainable artificial intelligence to predict whether intervention programs can reverse sarcopenia. Repeated-measures analysis of variance results indicated that there were statistically significant Group × Time interactions detected in the L3 skeletal muscle density, L3 skeletal muscle area, muscle fat infiltration, handgrip strength, and relative skeletal muscle mass index. The stacking model exhibited the best accuracy (84.5%) and the best F1-score (68.8%) compared to other algorithms. In the self-determined sequence exercise program group, strength training contributed most to the reversal of sarcopenia. One self-determined sequence exercise program can improve skeletal muscle area among sarcopenic older people. Based on our stacking model, we can predict whether sarcopenia in older people can be reversed accurately. The trial was registered in ClinicalTrials.gov. TRN:NCT05694117. Our findings indicate that such tailored exercise interventions can substantially benefit sarcopenic patients, and our stacking model provides an accurate predictive tool for assessing the reversibility of sarcopenia in older adults. This approach not only enhances individual health outcomes but also informs future development of targeted exercise programs to mitigate age-related muscle decline.

Accurate segmentation of hepatic and portal veins is critical for preoperative planning in liver surgery, especially for resection and transplantation procedures. Extensive anatomical variability, pathological alterations, and inherent class imbalance between background and vascular structures challenge this task. Current state-of-the-art deep learning approaches often fail to generalize across patient variability or maintain vascular topology, thus limiting their clinical applicability. To overcome these limitations, we propose the D²-RD-UNet, a dual-stage, dual-class segmentation framework for hepatic and portal vessels. The D²-RD-UNet architecture employs dense and residual connections to improve feature propagation and segmentation accuracy. Our D²-RD-UNet integrates advanced data-driven preprocessing, a dual-path architecture for 3D and 4D data, with the latter concatenating computed tomography (CT) scans with four relevant vesselness filters (Sato, Frangi, OOF, and RORPO). The pipeline is completed by the first developed postprocessing multi-class vessel connectivity correction algorithm based on centerlines. Additionally, we introduce the first radius-based branching algorithm to evaluate the model's predictions locally, providing detailed insights into the accuracy of vascular reconstructions at different scales. In order to make up for the scarcity of well-annotated open datasets for hepatic vessels segmentation, we curated AIMS-HPV-385, a large, pathological, multi-class, and validated dataset on 385 CT scans. We trained different configurations of D²-RD-UNet and state-of-the-art models on 327 CTs of AIMS-HPV-385. Experimental results on the remaining 58 CTs of AIMS-HPV-385 and on the 20 CTs of 3D-IRCADb-01 demonstrate superior performances of the D²-RD-UNet variants over state-of-the-art methods, achieving robust generalization, preserving vascular continuity, and offering a reliable approach for liver vascular reconstructions.

Multimodal Large Language Models (MLLMs) have emerged as a promising way to automate Radiology Report Generation (RRG). In this work, we systematically investigate the design space of 3D MLLMs, including visual input representation, projectors, Large Language Models (LLMs), and fine-tuning techniques for 3D CT report generation. We also introduce two knowledge-based report augmentation methods that improve performance on the GREEN score by up to 10\%, achieving the 2nd place on the MICCAI 2024 AMOS-MM challenge. Our results on the 1,687 cases from the AMOS-MM dataset show that RRG is largely independent of the size of LLM under the same training protocol. We also show that larger volume size does not always improve performance if the original ViT was pre-trained on a smaller volume size. Lastly, we show that using a segmentation mask along with the CT volume improves performance. The code is publicly available at https://github.com/bowang-lab/AMOS-MM-Solution

Spiculation, characterized by irregular, spike-like projections from nodule margins, serves as a crucial radiological biomarker for malignancy assessment and early cancer detection. These distinctive stellate patterns strongly correlate with tumor invasiveness and are vital for accurate diagnosis and treatment planning. Traditional computer-aided diagnosis (CAD) systems are limited in their capability to capture and use these patterns given their subtlety, difficulty in quantifying them, and small datasets available to learn these patterns. To address these challenges, we propose a novel framework leveraging variational autoencoders (VAE) to discover, extract, and vary disentangled latent representations of lung nodule images. By gradually varying the latent representations of non-spiculated nodule images, we generate augmented datasets containing spiculated nodule variations that, we hypothesize, can improve the diagnostic classification of lung nodules. Using the National Institutes of Health/National Cancer Institute Lung Image Database Consortium (LIDC) dataset, our results show that incorporating these spiculated image variations into the classification pipeline significantly improves spiculation detection performance up to 7.53%. Notably, this enhancement in spiculation detection is achieved while preserving the classification performance of non-spiculated cases. This approach effectively addresses class imbalance and enhances overall classification outcomes. The gradual attenuation of spiculation characteristics demonstrates our model's ability to both capture and generate clinically relevant semantic features in an algorithmic manner. These findings suggest that the integration of semantic-based latent representations into CAD models not only enhances diagnostic accuracy but also provides insights into the underlying morphological progression of spiculated nodules, enabling more informed and clinically meaningful AI-driven support systems.

To evaluate the depiction capability of fine lung nodules and airways using high-resolution settings on ultra-high-resolution energy-integrating detector CT (UHR-CT), incorporating large matrix sizes, thin-slice thickness, and iterative reconstruction (IR)/deep-learning reconstruction (DLR), using cadaveric human lungs and corresponding histological images. Images of 20 lungs were acquired using conventional CT (CCT), UHR-CT, and photon-counting detector CT (PCD-CT). CCT images were reconstructed with a 512 matrix and IR (CCT-512-IR). UHR-CT images were reconstructed with four settings by varying the matrix size and the reconstruction method: UHR-512-IR, UHR-1024-IR, UHR-2048-IR, and UHR-1024-DLR. Two imaging settings of PCD-CT were used: PCD-512-IR and PCD-1024-IR. CT images were visually evaluated and compared with histology. Overall, 6769 nodules (median: 1321 µm) and 92 airways (median: 851 µm) were evaluated. For nodules, UHR-2048-IR outperformed CCT-512-IR, UHR-512-IR, and UHR-1024-IR (p < 0.001). UHR-1024-DLR showed no significant difference from UHR-2048-IR in the overall nodule score after Bonferroni correction (uncorrected p = 0.043); however, for nodules > 1000 μm, UHR-2048-IR demonstrated significantly better scores than UHR-1024-DLR (p = 0.003). For airways, UHR-1024-IR and UHR-512-IR showed significant differences (p < 0.001), with no notable differences among UHR-1024-IR, UHR-2048-IR, and UHR-1024-DLR. UHR-2048-IR detected nodules and airways with median diameters of 604 µm and 699 µm, respectively. No significant difference was observed between UHR-512-IR and PCD-512-IR (p > 0.1). PCD-1024-IR outperformed UHR-CTs for nodules > 1000 μm (p ≤ 0.001), while UHR-1024-DLR outperformed PCD-1024-IR for airways > 1000 μm (p = 0.005). UHR-2048-IR demonstrated the highest scores among the evaluated EID-CT images. UHR-CT showed potential for detecting submillimeter nodules and airways. With the 512 matrix, UHR-CT demonstrated performance comparable to PCD-CT. Question There are scarce data evaluating the depiction capabilities of ultra-high-resolution energy-integrating detector CT (UHR-CT) for fine structures, nor any comparisons with photon-counting detector CT (PCD-CT). Findings UHR-CT depicted nodules and airways with median diameters of 604 µm and 699 µm, showing no significant difference from PCD-CT with the 512 matrix. Clinical relevance High-resolution imaging is crucial for lung diagnosis. UHR-CT has the potential to contribute to pulmonary nodule diagnosis and airway disease evaluation by detecting fine opacities and airways.

Post-surgical prediction of recurrence or metastasis for primary gastrointestinal stromal tumors (GISTs) remains challenging. We aim to develop individualized clinical follow-up strategies for primary GIST patients, such as shortening follow-up time or extending drug administration based on the clinical deep learning radiomics model (CDLRM). The clinical information on primary GISTs was collected from two independent centers. Postoperative recurrence or metastasis in GIST patients was defined as the endpoint of the study. A total of nine machine learning models were established based on the selected features. The performance of the models was assessed by calculating the area under the curve (AUC). The CDLRM with the best predictive performance was constructed. Decision curve analysis (DCA) and calibration curves were analyzed separately. Ultimately, our model was applied to the high-potential malignant group vs the low-malignant-potential group. The optimal clinical application scenarios of the model were further explored by comparing the DCA performance of the two subgroups. A total of 526 patients, 260 men and 266 women, with a mean age of 62 years, were enrolled in the study. CDLRM performed excellently with AUC values of 0.999, 0.963, and 0.995 for the training, external validation, and aggregated sets, respectively. The calibration curve indicated that CDLRM was in good agreement between predicted and observed probabilities in the validation cohort. The results of DCA's performance in different subgroups show that it was more clinically valuable in populations with high malignant potential. CDLRM could help the development of personalized treatment and improved follow-up of patients with a high probability of recurrence or metastasis in the future. This model utilizes imaging features extracted from CT scans (including radiomic features and deep features) and clinical data to accurately predict postoperative recurrence and metastasis in patients with primary GISTs, which has a certain auxiliary role in clinical decision-making. We developed and validated a model to predict recurrence or metastasis in patients taking oral imatinib after GIST. We demonstrate that CT image features were associated with recurrence or metastases. The model had good predictive performance and clinical benefit.

Medical imaging datasets often contain heterogeneous biases ranging from erroneous labels to inconsistent labeling styles. Such biases can negatively impact deep segmentation networks performance. Yet, the identification and characterization of such biases is a particularly tedious and challenging task. In this paper, we introduce HyperSORT, a framework using a hyper-network predicting UNets' parameters from latent vectors representing both the image and annotation variability. The hyper-network parameters and the latent vector collection corresponding to each data sample from the training set are jointly learned. Hence, instead of optimizing a single neural network to fit a dataset, HyperSORT learns a complex distribution of UNet parameters where low density areas can capture noise-specific patterns while larger modes robustly segment organs in differentiated but meaningful manners. We validate our method on two 3D abdominal CT public datasets: first a synthetically perturbed version of the AMOS dataset, and TotalSegmentator, a large scale dataset containing real unknown biases and errors. Our experiments show that HyperSORT creates a structured mapping of the dataset allowing the identification of relevant systematic biases and erroneous samples. Latent space clusters yield UNet parameters performing the segmentation task in accordance with the underlying learned systematic bias. The code and our analysis of the TotalSegmentator dataset are made available: https://github.com/ImFusionGmbH/HyperSORT