CIRF: Coupled Image Reconstruction and Fusion Strategy for Deep Learning Based Multi-Modal Image Fusion.

Multi-modal medical image fusion (MMIF) is crucial for disease diagnosis and treatment because the images reconstructed from signals collected by different sensors can provide complementary information. In recent years, deep learning (DL) based methods have been widely used in MMIF. However, these methods often adopt a serial fusion strategy without feature decomposition, causing error accumulation and confusion of characteristics across different scales. To address these issues, we have proposed the Coupled Image Reconstruction and Fusion (CIRF) strategy. Our method parallels the image fusion and reconstruction branches which are linked by a common encoder. Firstly, CIRF uses the lightweight encoder to extract base and detail features, respectively, through the Vision Transformer (ViT) and the Convolutional Neural Network (CNN) branches, where the two branches interact to supplement information. Then, two types of features are fused separately via different blocks and finally decoded into fusion results. In the loss function, both the supervised loss from the reconstruction branch and the unsupervised loss from the fusion branch are included. As a whole, CIRF increases its expressivity by adding multi-task learning and feature decomposition. Additionally, we have also explored the impact of image masking on the network's feature extraction ability and validated the generalization capability of the model. Through experiments on three datasets, it has been demonstrated both subjectively and objectively, that the images fused by CIRF exhibit appropriate brightness and smooth edge transition with more competitive evaluation metrics than those fused by several other traditional and DL-based methods.

Clinical Value of Magnetic Resonance Imaging Combined With Serum Prostate-specific Antigen, Epithelial Cadherin and Early Prostate Cancer Antigen 2 In Diagnosis of Prostate Cancer.

AIM: To explore the clinical value of magnetic resonance imaging (MRI) combined with serum prostate specific antigen (PSA), epithelial cadherin (sE-cadherin) and early prostate cancer antigen-2 (EPCA-2) in prostate cancer (PC) diagnosis. PATIENTS AND METHODS: Fifty patients with PC and 50 with benign prostatic hyperplasia (BPH) confirmed by pathology from January 2020 to July 2021 were studied retrospectively. All patients underwent MRI and measurement of the serum levels of PSA, EPCA-2, and sE-cadherin. The diagnostic accuracy and efficacy of these methods was compared between the groups. RESULTS: In MRI diagnosis of PC, lesions were mainly located in the peripheral zone; T2-weighted imaging of this zone showed low signal intensity, with different degrees of prostate enlargement. BPH had a clear boundary, complete capsule and central zone hyperplasia and uneven signal nodules. PC and BPH had different degrees of prostate enlargement. Serum levels of PSA, sE-cadherin and EPCA-2 in the cancer group were significantly higher than those in the BPH group (p<0.05). The diagnostic concordance of combined assessment of MRI, PSA, sE-cadherin, and EPCA-2 in differentiating PC from BPH was 93%, which was significantly higher than these approaches used alone (84%, 79%, 81% and 82%, respectively; p<0.05). The area under the receiver operating characteristics curve for the combined approach in PC diagnosis was 0.900, which was significantly higher than those for the individual methods (0.840, 0.730, 0.760 and 0.810, respectively; Z=2.343, p=0.004). CONCLUSION: MRI combined with PSA, sE-cadherin and EPCA-2 can improve the sensitivity and accuracy of PC diagnosis and has potential as a guiding scheme for early diagnosis of PC.

Maize Golden2-like transcription factors boost rice chloroplast development, photosynthesis, and grain yield.

Chloroplasts are the sites for photosynthesis, and two Golden2-like factors act as transcriptional activators of chloroplast development in rice (Oryza sativa L.) and maize (Zea mays L.). Rice OsGLK1 and OsGLK2 are orthologous to maize ZmGLK1 (ZmG1) and ZmGLK2 (ZmG2), respectively. However, while rice OsGLK1 and OsGLK2 act redundantly to regulate chloroplast development in mesophyll cells, maize ZmG1 and ZmG2 are functionally specialized and expressed in different cell-specific manners. To boost rice chloroplast development and photosynthesis, we generated transgenic rice plants overexpressing ZmG1 and ZmG2, individually or simultaneously, with constitutive promoters (pZmUbi::ZmG1 and p35S::ZmG2) or maize promoters (pZmG1::ZmG1, pZmG2::ZmG2, and pZmG1::ZmG1/pZmG2::ZmG2). Both ZmG1 and ZmG2 genes were highly expressed in transgenic rice leaves. Moreover, ZmG1 and ZmG2 showed coordinated expression in pZmG1::ZmG1/pZmG2::ZmG2 plants. All Golden2-like (GLK) transgenic plants had higher chlorophyll and protein contents, Rubisco activities and photosynthetic rates per unit leaf area in flag leaves. However, the highest grain yields occurred when maize promoters were used; pZmG1::ZmG1, pZmG2::ZmG2, and pZmG1::ZmG1/pZmG2::ZmG2 transgenic plants showed increases in grain yield by 51%, 47%, and 70%, respectively. In contrast, the pZmUbi::ZmG1 plant produced smaller seeds without yield increases. Transcriptome analysis indicated that maize GLKs act as master regulators promoting the expression of both photosynthesis-related and stress-responsive regulatory genes in both rice shoot and root. Thus, by promoting these important functions under the control of their own promoters, maize GLK1 and GLK2 genes together dramatically improved rice photosynthetic performance and productivity. A similar approach can potentially improve the productivity of many other crops.