Stimuli-responsive nano-immunochemotherapy modulates the tumor microenvironment to enhance antitumor immunity.

Medical image report generation (MeIRG) aims at generating associated diagnosis descriptions with natural language sentences from medical images, which is essential in the computer-aided diagnosis system. Nevertheless, this task remains challenging in that medical images and linguistic expressions should be understood jointly which however show great discrepancies in the modality. To fill this visual-to-semantic gap, we propose a novel framework that follows the encoder-decoder pipeline. Our framework is characterized by encoding both deep visual and semantic embeddings through a triple-branch network (TriNet) during the encoding phase. The visual attention branch captures attended visual embeddings from medical images with the soft-attention mechanism. The medical report (MeRP) embedding branch predicts semantic report embeddings. The embedding branch of medical subject headings (MeSH) obtains semantic embeddings of related medical tags as complementary information. Then, outputs of these branches are fused and fed into a decoder for the report generation. Experimental results on two benchmark datasets have demonstrated the excellent performance of our method.

In recent years, the growing demand for medical imaging diagnosis has placed a significant burden on radiologists. As a solution, Medical Vision-Language Pre-training (Med-VLP) methods have been proposed to learn universal representations from medical images and reports, benefiting downstream tasks without requiring fine-grained annotations. However, existing methods have overlooked the importance of cross-modal alignment in joint image-text reconstruction, resulting in insufficient cross-modal interaction. To address this limitation, we propose a unified Med-VLP framework based on Multi-task Paired Masking with Alignment (MPMA) to integrate the cross-modal alignment task into the joint imagetext reconstruction framework to achieve more comprehensive cross-modal interaction, while a Global and Local Alignment (GLA) module is designed to assist self-supervised paradigm in obtaining semantic representations with rich domain knowledge. Furthermore, we introduce a Memory-Augmented Cross-Modal Fusion (MA-CMF) module to fully integrate visual information to assist report reconstruction and fuse the multi-modal representations adequately. Experimental results demonstrate that the proposed unified approach outperforms previous methods in all downstream tasks, including uni-modal, cross-modal, and multimodal tasks.

Medical image reporting focused on automatically generating the diagnostic reports from medical images has garnered growing research attention. In this task, learning cross-modal alignment between images and reports is crucial. However, the exposure bias problem in autoregressive text generation poses a notable challenge, as the model is optimized by a word-level loss function using the teacher-forcing strategy. To this end, we propose a novel Token-Mixer framework that learns to bind image and text in one embedding space for medical image reporting. Concretely, Token-Mixer enhances the cross-modal alignment by matching image-to-text generation with text-to-text generation that suffers less from exposure bias. The framework contains an image encoder, a text encoder and a text decoder. In training, images and paired reports are first encoded into image tokens and text tokens, and these tokens are randomly mixed to form the mixed tokens. Then, the text decoder accepts image tokens, text tokens or mixed tokens as prompt tokens and conducts text generation for network optimization. Furthermore, we introduce a tailored text decoder and an alternative training strategy that well integrate with our Token-Mixer framework. Extensive experiments across three publicly available datasets demonstrate Token-Mixer successfully enhances the image-text alignment and thereby attains a state-of-the-art performance. Related codes are available at https://github.com/yangyan22/Token-Mixer.

Background: Lung adenocarcinoma (LUAD) is a major cause of cancer death. Therefore, identifying potential prognostic risk factors is critical to improve the survival of patients with LUAD. Methods: Here, relevant datasets were downloaded from TCGA and GEO databases to screen the differentially expressed genes (DEGs). Univariate Cox analysis, LASSO regression analysis and multivariate Cox analysis were conducted on the DEGs combined with TCGA clinical data, and finally a risk assessment model based on 10 feature genes was constructed. Results: The prognosis of patients was evaluated after the patients were grouped based on the median risk score and the results showed that the survival time of patients in the high-risk group was significantly shorter than that in the low-risk group. ROC analysis showed that the AUC values of the 1, 3, 5-year survival were 0.753, 0.724, and 0.73, respectively, indicating that the model was precise in predicting the prognosis, which was also verified in the external dataset GSE72094. In addition, a significant correlation was found between the risk score and the clinical stages of LUAD, that is, a later stage always corresponded to a higher risk score. Then, we performed survival analysis on the 10 feature genes independently in the TCGA-LUAD dataset through the GEPIA database, finding that the high expression of 6 genes (COL5A2, PLEK2, BAIAP2L2, S100P, ZIC2, SFXN1) was associated with the poor prognosis of LUAD patients. Conclusion: To sum, this study established a 10-gene risk assessment model and further evaluated its value in predicting LUAD prognosis, which provided a new method for the prognosis prediction of LUAD.

Purpose: Inflammatory myofibroblastic tumors (IMT) was a rare kind of tumor defined by WHO since 2012. Little was known about this disease. There were controversies about IMT's behavior, predilection site, age distribution, and the best treatment methods. Here we provided a systematic overview on tumor demographical, clinical, biological features as well as treatment efficacy based on real cases from Surveillance, Epidemiology, and End Results (SEER) database. Methods: 92 patients diagnosed with IMT by histopathology were drawn from SEER database between 2002 and 2014. Patient demographics, clinical features and treatment information were analyzed. Results: The mean age of onset was 47.4 ± 22.4 years (0 to 83y) and the ages prone to this disease are middle-aged (from 41y to 64y), accounting for 1/3 of all patients. Three peak ages of onsets were 0–4y, 36–40y and more than 50y. 42% of the tumors were located in the soft tissues of limbs, hip, shoulder, head, face and neck. The average tumor sizes were 6.5 ± 5.3cm (1cm to 25cm). Survival in the group of tumor size smaller than 6.5cm was better compared to group of tumor size larger than 6.5cm (P < 0.05). Most of the tumors were malignant or malignant potential (89%), though local and distant metastasis rate were low (5%). Surgery was the most common treatment. However, the survival benefit was still uncertain compared to adjuvant chemotherapy or radiotherapy. Multivariate regression analysis demonstrated that young patients had better survival than old ones. Conclusions: IMT was a malignant tumor with low risk of local and distant metastasis. The peak ages were 0-4y, 36-40y and more than 50y. The prone sites were the soft tissues of the limbs, hip, shoulder, head, face and neck. Tumor sizes and ages were the factors correlated with survival time.