Search

Abstract Although previous studies conducted on the segmentation of hemorrhage images were based on the U-Net model, which comprises an encoder-decoder architecture, these models exhibit low parameter passing efficiency between the encoder and decoder, large model size, and slow speed. Therefore, to overcome these drawbacks, this study proposes TransHarDNet, an image segmentation model for the diagnosis of intracerebral hemorrhage in CT scan images of the brain. In this model, the HarDNet block is applied to the U-Net architecture, and the encoder and decoder are connected using a transformer block. As a result, the network complexity was reduced and the inference speed improved while maintaining the high performance compared to conventional models. Furthermore, the superiority of the proposed model was verified by using 82,636 CT scan images showing five different types of hemorrhages to train and test the model. Experimental results showed that the proposed model exhibited a Dice coefficient and IoU of 0.712 and 0.597, respectively, in a test set comprising 1200 images of hemorrhage, indicating better performance compared to typical segmentation models such as U-Net, U-Net++, SegNet, PSPNet, and HarDNet. Moreover, the inference time was 30.78 frames per second (FPS), which was faster than all en-coder-decoder-based models except HarDNet.

Abstract This paper aims to consider the dual Brunn–Minkowski inequality for log-volume of star bodies, and the equivalent Minkowski inequality for mixed log-volume.

BackgroundEndoscopic forceps biopsy (EFB) lacks precision in diagnosing indeterminate tumors. When the presence of early gastric cancer (EGC) is macroscopically suspected, but biopsy pathology fails to give a diagnosis of neoplasia, it causes problems in clinical management. The purpose of this study was to discuss the outcome of gastric indeterminate tumors and the clinical factors associated with predicting EGC.MethodsThe medical records of 209 patients diagnosed with gastric indeterminate neoplasia by biopsy forceps were retrospectively studied. Initial endoscopic findings were analyzed and predictors of EGC were evaluated.ResultsThe final pathological diagnosis in 209 patients included adenocarcinoma (n = 7), high-grade intraepithelial neoplasia (n = 11), low-grade intraepithelial neoplasia (n = 21), and non-neoplastic lesion (n = 170). Multivariate analysis showed that older age (OR = 1.78; 95% CI = 1.17–2.71; p = 0.008), patients undergoing narrow band imaging (NBI) (OR = 3.40; 95% CI = 1.37–8.43; p = 0.008), and surface erosion (OR = 3.53; 95% CI = 1.41–8.84; p = 0.007) were associated with the upgraded group, and were significantly associated with risk. Univariate logistic regression analysis showed that among patients with NBI, the presence of demarcation line (DL) (OR = 24.00; 95% CI = 4.99–115.36; p < 0.0001), microvascular (MV) pattern irregularity (OR = 9.129; 95% CI = 2.36–35.34; p = 0.001), and the presence of white opaque substance (WOS) (OR = 10.77; 95% CI = 1.14–101.72; p = 0.038) were significant risk factors.ConclusionsFor gastric indeterminate tumors, older patient age, lesion surface with erosion, clear DL visible under NBI observation, presence of WOS, and irregular MV pattern are suggestive of the high possibility of neoplasia and need to be focused on and may benefit more from endoscopic resection treatment as opposed to simple endoscopic follow-up.

BackgroundEarly colorectal cancer (ECRC) refers to any size of colorectal cancer (CRC) whose depth of invasion is limited to the mucosa and submucosa. About 10% of patients with ECRC die from cancer after surgery. KRAS, NRAS, and BRAF mutations and microsatellite instability (MSI) are considered diagnostic and prognostic markers in CRC. However, their characteristics in ECRC and whether postoperative chemotherapy based on them will benefit ECRC patients or not remain unknown.Patients and MethodsPatients with ECRC and 298 patients with advanced colorectal cancer (ACRC) were collected in our hospital from January 2013 to December 2015. The Amplification Refractory Mutation System (ARMS)-PCR was used to perform the KRAS, NRAS, and BRAF mutant tests.ResultsIn ECRC patients, 43 cases of KRAS mutation were found, accounting for 69.35%. Interestingly, among KRAS mutations, there were 10 KRAS multi-site mutation patients (16.13% in 62 ECRC patients). Moreover, the NRAS mutation rate was 3.23% but no BRAF mutation was found and only 1 case of MSI-High was detected. KRAS mutation was only related to the depth of tumor invasion whereas KRAS multi-site mutations were related to mucus components and tumor size. As far as NRAS is concerned, mutations were associated with elevated CEA, mucus components, and the depth of tumor invasion. Notably, compared with 2.35% KRAS multi-site mutation in ACRC, the rate of KRAS multi-site mutation in ECRC was much higher. Furthermore, Cox regression analysis revealed that KRAS mutation could be an independent prognostic factor of ECRC in patients who have undergone endoscopic resection or surgery.ConclusionPatients with ECRC might benefit from KRAS mutation testing but not from postoperative chemotherapy.