Multilayer perceptron can be trained with empirical data to estimate general real-valued functions or to be used as a pattern classifier to estimate indicator functions. The typical backpropagation learning algorithm and its variations do not distinguish the training of an MLP as a pattern classifier from that of a general function estimator. In this paper, we present a learning algorithm based on an optimization layer by layer (OLL) procedure. Its main difference from previously reported OLL-type learning algorithms is that the weights between the last hidden layer and the output layer are determined through optimization of a piecewise linear objective function subject to constraints designed specifically for training an MLP to be a pattern classifier. The performance of the proposed learning algorithm is compared with that of the backpropagation algorithm, the modified Newton's method and the improved descending epsilon algorithm over multiple training sessions using both simulated and real data classification problems.
Objective:To analysis the efficacy of external local radiation plus internal.Methods:91 cases of esophagus carcinoma in our hospitol from May 1996 to October 1998,was andomly dividedinto wo groups;45 cases in control group was received the external local radiation(group A) and 46 cases in treatment group received external radiation and internal local high dose rate 192 Ir(group B).Results:The 1 year and 3 year survival rates in the treatment group was 73.9% and 26.1% respectively,and the survival rate in the control group was 48.9% and 17.8% in 1 and 3 year respectively,the survival in the groups was statistically difference(P0.05).Conclusion:The combination of external and internal local radiation treatment modality was better than the external radiation alone,and may get longer survival rate.
Recently, artificial intelligence generated content (AIGC) has been receiving increased attention and is growing exponentially. AIGC is generated based on the intentional information extracted from human-provided instructions by generative artificial intelligence (AI) models. AIGC quickly and automatically generates large amounts of high-quality content. Currently, there is a shortage of medical resources and complex medical procedures in medicine. Due to its characteristics, AIGC can help alleviate these problems. As a result, the application of AIGC in medicine has gained increased attention in recent years. Therefore, this paper provides a comprehensive review on the recent state of studies involving AIGC in medicine. First, we present an overview of AIGC. Furthermore, based on recent studies, the application of AIGC in medicine is reviewed from two aspects: medical image processing and medical text generation. The basic generative AI models, tasks, target organs, datasets and contribution of studies are considered and summarized. Finally, we also discuss the limitations and challenges faced by AIGC and propose possible solutions with relevant studies. We hope this review can help readers understand the potential of AIGC in medicine and obtain some innovative ideas in this field.
Abstract Background Costimulatory molecules play significant roles in mounting anti-tumor immune responses, and antibodies targeting these molecules are recognized as promising adjunctive cancer immunotherapies. Here, we aim to conduct a first full-scale exploration of costimulatory molecules from the B7-CD28 and TNF families in patients with lung adenocarcinoma (LUAD) and generated a costimulatory molecule-based signature (CMS) to predict survival and response to immunotherapy. Methods We enrolled 1549 LUAD cases across 10 different cohorts and included 502 samples from TCGA for discovery. The validation set included 970 cases from eight different Gene Expression Omnibus (GEO) datasets and 77 frozen tumor tissues with qPCR data. The underlying mechanisms and predictive immunotherapy capabilities of the CMS were also explored. Results A five gene-based CMS (CD40LG, TNFRSF6B, TNFSF13, TNFRSF13C, and TNFRSF19) was initially constructed using the bioinformatics method from TCGA that classifies cases as high- vs. low-risk groups per OS. Multivariable Cox regression analysis confirmed that the CMS was an independent prognostic factor. As expected, CMS exhibited prognostic significance in the stratified cohorts and different validation cohorts. Additionally, the prognostic meta-analysis revealed that CMS was superior to the previous signature. Samples in high- and low-risk groups exhibited significantly different tumor-infiltrating leukocytes and inflammatory activities. Importantly, we found that signature high-risk patients were optimal candidates for immunotherapy. Conclusion We conducted the first and most comprehensive costimulatory molecule landscape analysis of patients with LUAD and built a clinically feasible CMS for prognosis and immunotherapy response prediction, which will be helpful for further optimize immunotherapies for cancer.
5573 Background: OVA1 is a panel of biomarkers cleared by FDA currently in clinical use for pre-surgical assessment of adnexal masses for risk of ovarian malignancy. To further improve the specificity of OVA1, we evaluated biomarkers using a designed set of clinical samples enriched with OVA1 false positive benign patients and selected insulin-like growth factor binding protein 2 (IGFBP2), interleukin 6 (IL6), and follicle-stimulating hormone (FSH) to be further evaluated along with the original five biomarkers of OVA1 on a prospectively collected clinical sample set. The inclusion of FSH was to eliminate the need for menopause-specific cutoffs. Methods: Consecutive patients with a documented pelvic mass planned for surgical intervention were prospectively enrolled at 27 sites. Exclusion criteria included a diagnosis of malignancy in the previous 5 years or initial enrollment by a gynecologic oncologist. At the time of analysis, 384 subjects had all biomarker values. Among them 69 were ovarian cancer cases (13 LMPs, 27 stages 1/2, 19 serous, 11 endometrioid , 5 mucinous, and 4 clear cell). Biomarkers were tested by ELISA and reported as continuous values. Using a subset of the samples, the biomarkers were first selected for inclusion in a final panel based on contributions in multivariate models estimated by bootstrap. The selected biomarkers were further assessed for ability to improve specificity of risk stratification at a fixed sensitivity over that of OVA1 using the full data set. This was done by cross-validation of multivariate models with 50/50 split between training and testing. Results: The final panel of biomarkers consisted of CA125II, prealbumin, IGFBP2, IL6, and FSH. At a fixed sensitivity of 90%, the mean and median specificity of models using the new panel in testing were 78.2% (95% CI: 76.7 – 79.8%), and 80.6%, respectively. The mean and median absolute improvements over that of OVA1 were 18.6% (95% CI: 16.4% – 20.9%) and 20.3%, respectively. Conclusions: The new panel demonstrated the potential to significantly improve specificity over that of the first-generation OVA1 algorithm, while maintaining a high sensitivity in pre-surgical assessment of adnexal masses for risk of malignancy.
The effects of different light qualities on the multiplication of the callus which was induced from the seeds of Vitis vinifera‘Pinot Noir’and subcultured for 10 months,and the resveratrol content were studied in this paper.The results showed that the light qualities for the multiphcation were in the order of yellowgreenredbluewhite,but as for resveratrol accumulation and production in the calh were in that of whiteyellowredbluegreen.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)