Baoming Luo's research while affiliated with Sun Yat-Sen University and other places

... These innovative technologies are now being applied across the spectrum of HCC management, from enhancing screening and early diagnosis to improving tumor characterization, prognostication, and personalized therapy selection as depicted in Table 1 [10], [11], [12]. [13], [14], [15], [16], [17], [18], [19], [20], [21], [22] ML analysis of ultrasound images Convolutional neural networks (CNNs) identify subtle imaging traits of early HCCs Diagnosis and Staging [23], [24] Deep learning on computed tomography (CT)/magnetic resonance imaging (MRI) scans CNNs differentiate HCC from other liver lesions Prognostication [25], [26] ML models with clinical, imaging, genomic data ...

... Another regulatory axis, circESRP1/miR-3942/CTCF, was revealed by Gong et al., and they validated that CTCF can precisely promote circESRP1 transcript expression, thus creating a positive-feedback loop in ccRCC progression [72]. Apart from the already mentioned positive feedback loop, the circNIPBL/miR-16-2-3p/Wnt5a/ZEB1 axis in bladder cancer [73], circMGA/HNRNPL complex as a pivotal tumor suppressor by recruiting CD8 + T cells in bladder cancer [74], and loop among circCDK13/ miR-212-5p/449a/E2F5 in prostate carcinogenesis [75] were also shown to participate in tumor progression. Our study identified circSOD2 with significant tumor-promoting effects in RCC based on bioinformatics and preliminary validation of clinical samples. ...

... Apart from NC318, which is currently undergoing clinical trials, the utilization of nanotechnology for the targeted inhibition of Siglec-15 has been identified as a significant recent research initiative. There is a high level of anticipation within the scientific community for the data that is beginning to emerge, as well as for the potential therapeutic applications that this novel strategy may yield [142,143]. We explore the expression profiles of specific Siglecs in various cancer types and their interplay with immune checkpoints like PD-L1. These insights offer valuable perspectives for pioneering novel cancer immunotherapies, particularly for patients resistant to PD-1/PD-L1 treatments [61]. ...

... However, when we added the complex to the depleted conditioned media, repopulation was restored. Several previous studies have been shown that exogenous HMGB1 promotes pro-survival autophagy and stimulates the tumor growth of gliomas [53] and esophageal squamous carcinoma [29], in addition to enhancing the chemoresistance of tumor cells in leukemic and bladder cancers [50]. Hence, we checked the expression of autophagy markers in A549 cells treated with either full Eto-CM or with Eto-CM depleted by specific anti-Hsp70 antibodies or with HMGB1-binding HBHP peptide. ...

... The YOLOV1-V4 target detection technique is used to pinpoint the precise location of lesions. The lesion region of interest was identified in an image using 3D-CNN, and it was subsequently classified and shown using the visualization interface (13,14). It provided comprehensive data processing and picture analysis facilities, while also being reviewed in a multi-planar reconstruction display. ...

... Studies have shown that imaging modalities, including US (7), CT (4,8), PET/CT (9), and MRI (10-16), can help accurately diagnose MTM, in which arterial phase (AP) hypoenhancing components of more than 20% or 50%, the presence of substantial necrosis, and severe ischemia have been identified as independent predictors of MTM but have either unsatisfactory sensitivity (7.4%-57%) or unsatisfactory specificity (31.3%-66%) (7,11,13,15). Unfortunately, substantial necrosis and severe ischemia rarely appear in small HCCs (eg, ≤3 cm diameter), making it challenging for radiologists to identify such lesions as the MTM subtype. ...

... During the autonomous US scanning process, the magnitude and variation of contact force significantly impact the quality and stability of US images, as observed by Ipsen et al. (2021). Researchers have proposed various control algorithms to maintain a constant contact force, including impedance control (Suligoj et al., 2021), admittance control (Tan et al., 2022;Wang Z. et al., 2023), and PID control (Ma et al., 2021;Zhang et al., 2021;. Moreover, due to the significant fluctuations in window quality during cardiac US scanning, researchers have introduced several US image servo control algorithms to ensure stable, high-quality imaging. ...

... Additionally, these biselective probes complement existing antibody drug conjugates and create opportunities to target disease-related immune cell subsets in the tumor microenvironment. Xu et al. [57] introduced a mannose-modified N-azidoacetylgalactosamine (Ac4GalNAz, AAG) liposome, which participated in a two-step click chemistry procedure for detecting TAMs. After entering TAMs, AAG is converted to N-azidoacetylgalactosamine (GalNAz) and tagged mucin-type O-linked glycans. ...

... In order to develop an autonomous robotic US system (RUSS), Huang et al. computed a multiple-line trajectory based on an external RGBD camera [7]. To en-hance the representation of 3D objects, Tan et al. planned a path on fused surface point clouds captured from multiple cameras [8], [9]. However, these methods only consider the outer surface, whereas the planned path cannot guarantee the visibility of the thoracic organs, such as the liver. ...

... Perek et al. [148] arrived at a similar conclusion when trialling MoCo with a private dataset [11]. Proposing a video-specific pretext task instead, Lin et al. [149] pretrained an encoder-decoder architecture to restore a US video after randomly masking out entire frames and patches in the remaining frames. Upon performing semisupervised fine-tuning for benign versus malignant lesion classification on a private dataset, masked video pretraining yielded 1% greater accuracy compared to random initialization. ...