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.

OBJECTIVE: The skin marking method (SMM) and bow-form-ruler marking method (BFRM) are two commonly used patient marking methods in mainland China. This study aims to evaluate SMM and BFRM by comparing the inter-fraction setup errors from using these two methods together with vacuum cushion immobilization in patients underwent radiotherapy for different treatment sites. MATERIALS AND METHODS: Eighteen patients diagnosed with pelvic, abdominal and thoracic malignant tumors (with 6 patients per treatment site) were enrolled in this prospective study. All patients were immobilized with vacuum cushion. Each patient was marked by both SMM and BFRM before computed tomography (CT) simulation. Target location was verified by cone beam CT images with displacements assessed prior to each sampled treatment session. The localization errors in three translational and three rotational directions were recorded and analyzed. RESULTS: Images from 108 fractions in 18 patients produced 324 translational and 324 rotational comparisons for SMM and BFRM. The setup errors of all treatment sites showed no difference in two marking methods in any directions ( p > 0.05). In subgroups of treatment site analysis, SMM significantly lessened the lateral and yaw setup errors compared to BFRM in the pelvic sites (0.39±1.85 mm vs –1.28±1.13 mm, p < 0.01 and –0.19±0.59° vs –0.61±0.59°, p < 0.05). However, in the abdominal subgroup, BFRM was superior to SMM for reduced vertical errors (0.17±2.73 mm vs 2.28±3.16 mm, p < 0.05). For the underweight or obese patients (with Body Mass Index, BMI < 18.5 or BMI≥24), SMM resulted in less yaw errors compared to BFRM (–0.05±0.38° vs –0.43±0.48°, p < 0.05). No significant difference between SMM and BFRM in setup errors of normal weighted patients (18.5≤BMI < 24) was observed for all three studied treatment sites. CONCLUSIONS: This study shows no significant difference in patient setup errors for various treatment sites between SMM and BFRM in general. SMM may be suitable for the pelvic tumor and patients with BMI < 18.5 or BMI≥24, while BFRM is recommended for the abdominal tumor sites.

Early prediction of lymph node metastasis (LNM) following neoadjuvant therapy (NAT) is crucial for timely treatment optimization in esophageal squamous cell carcinoma (ESCC). This study developed and validated a computed tomography-based radiomic model for predicting pathologically confirmed LNM status at the time of surgery in ESCC patients after NAT. A total of 469 ESCC patients from Sun Yat-sen University Cancer Center were retrospectively enrolled and randomized into a training cohort (n = 328) and a test cohort (n = 141). Three signatures were constructed: the tumor-habitat-based signature (Habitat_Rad), derived from radiomic features of three tumor subregions identified via K-means clustering; the multiple instance learning-based signature (MIL_Rad), combining features from 2.5D deep learning models; and the clinicoradiological signature (Clinic), developed through multivariate logistic regression. A combined radiomic nomogram integrating these signatures outperformed the individual models, achieving areas under the curve (AUCs) of 0.929 (95% CI, 0.901–0.957) and 0.852 (95% CI, 0.778–0.925) in the training and test cohorts, respectively. The decision curve analysis confirmed a high net clinical benefit, highlighting the nomogram’s potential for accurate LNM prediction after NAT and guiding individualized therapy.

In the field of 3D point cloud data, the 3D representation of objects is often affected by factors such as lighting, occlusion, and noise, leading to issues of information loss and incompleteness in the collected point cloud data. Point cloud completion algorithms aim to generate complete object point cloud data using partial or local point cloud data as input. Despite promising results achieved by existing methods, current point cloud completion approaches often lack smooth and structural consistency, resulting in a messy overall structure. To address these shortcomings in point cloud completion, we propose a point cloud generative method based on surface consistency and scale rendering. In addition, to solve the limitation of existing methods that mainly focus on geometric features in 3D point cloud completion and do not make full use of color information, we introduce an object reconstruction method based on texture and geometric features. Extensive experiments demonstrate that our proposed methods exhibit superior performance in terms of local details and overall object structure.

There was an error in the original publication