Tumor Segmentation in Intraoperative Fluorescence Images Based on Transfer Learning and Convolutional Neural Networks.

OBJECTIVE: To propose a transfer learning based method of tumor segmentation in intraoperative fluorescence images, which will assist surgeons to efficiently and accurately identify the boundary of tumors of interest. METHODS: We employed transfer learning and deep convolutional neural networks (DCNNs) for tumor segmentation. Specifically, we first pre-trained four networks on the ImageNet dataset to extract low-level features. Subsequently, we fine-tuned these networks on two fluorescence image datasets (ABFM and DTHP) separately to enhance the segmentation performance of fluorescence images. Finally, we tested the trained models on the DTHL dataset. The performance of this approach was compared and evaluated against DCNNs trained end-to-end and the traditional level-set method. RESULTS: The transfer learning-based UNet++ model achieved high segmentation accuracies of 82.17% on the ABFM dataset, 95.61% on the DTHP dataset, and 85.49% on the DTHL test set. For the DTHP dataset, the pre-trained Deeplab v3 + network performed exceptionally well, with a segmentation accuracy of 96.48%. Furthermore, all models achieved segmentation accuracies of over 90% when dealing with the DTHP dataset. CONCLUSION: To the best of our knowledge, this study explores tumor segmentation on intraoperative fluorescent images for the first time. The results show that compared to traditional methods, deep learning has significant advantages in improving segmentation performance. Transfer learning enables deep learning models to perform better on small-sample fluorescence image data compared to end-to-end training. This discovery provides strong support for surgeons to obtain more reliable and accurate image segmentation results during surgery.

Deep learning radiomics based prediction of axillary lymph node metastasis in breast cancer.

This study aimed to develop and validate a deep learning radiomics nomogram (DLRN) for the preoperative evaluation of axillary lymph node (ALN) metastasis status in patients with a newly diagnosed unifocal breast cancer. A total of 883 eligible patients with breast cancer who underwent preoperative breast and axillary ultrasound were retrospectively enrolled between April 1, 2016, and June 30, 2022. The training cohort comprised 621 patients from Hospital I; the external validation cohorts comprised 112, 87, and 63 patients from Hospitals II, III, and IV, respectively. A DLR signature was created based on the deep learning and handcrafted features, and the DLRN was then developed based on the signature and four independent clinical parameters. The DLRN exhibited good performance, yielding areas under the receiver operating characteristic curve (AUC) of 0.914, 0.929, and 0.952 in the three external validation cohorts, respectively. Decision curve and calibration curve analyses demonstrated the favorable clinical value and calibration of the nomogram. In addition, the DLRN outperformed five experienced radiologists in all cohorts. This has the potential to guide appropriate management of the axilla in patients with breast cancer, including avoiding overtreatment.

Automated peripancreatic vessel segmentation and labeling based on iterative trunk growth and weakly supervised mechanism.

Peripancreatic vessel segmentation and anatomical labeling are pivotal aspects in aiding surgical planning and prognosis for patients with pancreatic tumors. Nevertheless, prevailing techniques often fall short in achieving satisfactory segmentation performance for the peripancreatic vein (PPV), leading to predictions characterized by poor integrity and connectivity. Besides, unsupervised labeling algorithms usually cannot deal with complex anatomical variation while fully supervised methods require a large number of voxel-wise annotations for training, which is very labor-intensive and time-consuming. To address these two problems, we propose an Automated Peripancreatic vEssel Segmentation and lAbeling (APESA) framework, to not only highly improve the segmentation performance for PPV, but also efficiently identify the peripancreatic artery (PPA) branches. There are two core modules in our proposed APESA framework: iterative trunk growth module (ITGM) for vein segmentation and weakly supervised labeling mechanism (WSLM) for artery labeling. The ITGM is composed of a series of iterative submodules, each of which chooses the largest connected component of the previous PPV segmentation as the trunk of a tree structure, seeks for the potential missing branches around the trunk by our designed branch proposal network, and facilitates trunk growth under the connectivity constraint. The WSLM incorporates the rule-based pseudo label generation with less expert participation, an anatomical labeling network to learn the branch distribution voxel by voxel, and adaptive radius-based postprocessing to refine the branch structures of the labeling predictions. Our achieved Dice of 94.01% for PPV segmentation on our collected dataset represents an approximately 10% accuracy improvement compared to state-of-the-art methods. Additionally, we attained a Dice of 97.01% for PPA segmentation and competitive labeling performance for PPA labeling compared to prior works. Our source codes will be publicly available at https://github.com/ZouLiwen-1999/APESA.

A Weakly Supervised Segmentation Network Embedding Cross-Scale Attention Guidance and Noise-Sensitive Constraint for Detecting Tertiary Lymphoid Structures of Pancreatic Tumors.

The presence of tertiary lymphoid structures (TLSs) on pancreatic pathological images is an important prognostic indicator of pancreatic tumors. Therefore, TLSs detection on pancreatic pathological images plays a crucial role in diagnosis and treatment for patients with pancreatic tumors. However, fully supervised detection algorithms based on deep learning usually require a large number of manual annotations, which is time-consuming and labor-intensive. In this paper, we aim to detect the TLSs in a manner of few-shot learning by proposing a weakly supervised segmentation network. We firstly obtain the lymphocyte density maps by combining a pretrained model for nuclei segmentation and a domain adversarial network for lymphocyte nuclei recognition. Then, we establish a cross-scale attention guidance mechanism by jointly learning the coarse-scale features from the original histopathology images and fine-scale features from our designed lymphocyte density attention. A noise-sensitive constraint is introduced by an embedding signed distance function loss in the training procedure to reduce tiny prediction errors. Experimental results on two collected datasets demonstrate that our proposed method significantly outperforms the state-of-the-art segmentation-based algorithms in terms of TLSs detection accuracy. Additionally, we apply our method to study the congruent relationship between the density of TLSs and peripancreatic vascular invasion and obtain some clinically statistical results.

CTG-Net: an efficient cascaded framework driven by terminal guidance mechanism for dilated pancreatic duct segmentation.

Pancreatic duct dilation indicates a high risk of various pancreatic diseases. Segmentation for dilated pancreatic duct (DPD) on computed tomography (CT) image shows the potential to assist the early diagnosis, surgical planning and prognosis. Because of the DPD's tiny size, slender tubular structure and the surrounding distractions, most current researches on DPD segmentation achieve low accuracy and always have segmentation errors on the terminal DPD regions. To address these problems, we propose a cascaded terminal guidance network to efficiently improve the DPD segmentation performance. Firstly, a basic cascaded segmentation architecture is established to get the pancreas and coarse DPD segmentation, a DPD graph structure is build on the coarse DPD segmentation to locate the terminal DPD regions. Then, a terminal anatomy attention module is introduced for jointly learning the local intensity from the CT images, feature cues from the coarse DPD segmentation and global anatomy information from the designed pancreas anatomy-aware maps. Finally, a terminal distraction attention module which explicitly learns the distribution of the terminal distraction regions is proposed to reduce the false positive and false negative predictions. We also propose a new metric called tDice to measure the terminal segmentation accuracy for targets with tubular structures and two other metrics for segmentation error evaluation. We collect our dilated pancreatic duct segmentation dataset with 150 CT scans from patients with five types of pancreatic tumors. Experimental results on our dataset show that our proposed approach boosts DPD segmentation accuracy by nearly 20% compared with the existing results, and achieves more than 9% improvement for the terminal segmentation accuracy compared with the state-of-the-art methods.

A two-stage data-model driven pancreas segmentation strategy embedding directional information of the boundary intensity gradient and deep adaptive pointwise parameters.

In this paper, we propose a two-stage data-model driven pancreas segmentation method that combines a 3D convolution neural network with adaptive pointwise parametric hybrid variational model embedding the directional and magnitude information of the boundary intensity gradient. Firstly, nnU-net is used to segment the entire abdominal CT image with the aim of obtaining the region of the interest of pancreas. Secondly, an adaptive pointwise parametric variational model with a new edge term containing the directional and magnitude information of the boundary intensity gradient is used to refine the predicted results from CNN. Although CNN is good at extracting texture information, it does not capture weak boundary information very well. In order to well acquire more weak boundary information of the pancreas, we utilize not only the magnitude of the gradient, but also the directional information of the boundary intensity gradient to obtain more accurate results in the new edge term. In addition, the probability value for each pixel obtained by calculating the softmax function is exploited twice. Actually, it is applied firstly to generate the binary map as the initial contour of the variational model and then to design the adaptive pointwise weight parameters of internal and external area terms of the variational model rather than constants. It not only eliminates the trouble of manual parameter adjustment, but also, most importantly, provides a more accurate pointwise evolutionary trend of the level set contour, i.e. determine the tendency of the level set contour to pointwisely contract inward or expand outward. Our method is evaluated on three public datasets and outperformed the state-of-the-art pancreas segmentation methods. Accurate pancreatic segmentation allows for more reliable quantitative analysis of local morphological changes in the pancreas, which can assist in early diagnosis and treatment planning.

Measuring the rogue wave pattern triggered from Gaussian perturbations by deep learning.

Weak Gaussian perturbations on a plane wave background could trigger lots of rogue waves (RWs), due to modulational instability. Numerical simulations showed that these RWs seemed to have similar unit structure. However, to the best of our knowledge, there are no relative results to prove that these RWs have the similar patterns for different perturbations, partly due to that it is hard to measure the RW pattern automatically. In this work, we address these problems from the perspective of computer vision via using deep neural networks. We propose a rogue wave detection network (RWD-Net) model to automatically and accurately detect RWs in the images, which directly indicates they have the similar computer vision patterns. For this purpose, we herein meanwhile have designed and release the corresponding dataset, termed as rogue wave dataset-10K (RWD-10K), which has 10191 RW images with bounding box annotations for each RW unit. In our detection experiments, we get 99.29% average precision on the test splits of the proposed dataset. Finally, we derive our metric, termed as the density of RW units, to characterize the evolution of Gaussian perturbations and obtain the statistical results on them.

Combined Use of Trichoderma atroviride CCTCCSBW0199 and Brassinolide to Control Botrytis cinerea Infection in Tomato.

Tomato gray mold caused by Botrytis cinerea is one of the main diseases of tomato and significantly impacts the yield and quality of tomato fruit. The overuse of chemical fungicides has resulted in the development of fungicide-resistant strains. Biological control is becoming an alternative method for the control of plant diseases to replace or decrease the application of traditional synthetic chemical fungicides and genus Trichoderma is widely used as a biological agent for controlling tomato gray mold. Brassinolide (BR) is a plant-growth-promoting steroid. To enhance the efficiency and stability of Trichoderma activity against B. cinerea, an optimal combination of Trichoderma atroviride CCTCCSBW0199 and BR that controls B. cinerea infection in tomato was identified. Strain CCTCCSBW0199 was found to have antagonistic activity against B. cinerea both in vitro and in vivo. In addition, a fermented culture of chlamydospores and metabolites, or metabolites only of strain CCTCCSBW0199 also reduced growth of B. cinerea. BR reduced growth of B. cinerea and had no effect on the sporulation and growth of Trichoderma spp. An application of metabolites of a Trichoderma sp. + BR reduced gray mold on tomato leaves by approximately 70.0%. Furthermore, the activities of induced defense response-related enzyme, such as peroxidase, superoxide dismutase, catalase, and phenylalanine ammonia-lyase were increased in tomato plants treated with a Trichoderma sp. + BR. Our data suggested that applying a mix of metabolites of T. atroviride CCTCCSBW0199 + BR was effective at reducing gray mold of tomato and may lay a theoretical foundation for the development of novel biofungicides.

Reprogrammed endophytic microbial community in maize stalk induced by Trichoderma asperellum biocontrol agent against Fusarium diseases and mycotoxin accumulation.

Maize stalk rot and ear rot, caused by Fusarium graminearum and Fusarium verticillioides, respectively, are major diseases that threaten the sustainable production of maize. In this study, an artificial inoculation assay demonstrated that the control efficacy of maize stalk rot and ear rot by Trichoderma asperellum granules were 49.83 % and 39.63 %, respectively. By high-throughput sequencing of maize plants, a total of 76 196 ITS1 sequences and 887 226 V3V4 16S rRNA sequences were analyzed and were grouped into 2934 fungal and 24 248 bacterial operational taxonomic units (OTUs), respectively. It revealed a significantly higher endophytic microbial abundance in the stem tissue of plants grown in T. asperellum-treated soil than in those grown in the control, with the largest increase observed in the basal stem section. In addition, the endophytic microbial diversity and corresponding control effects all gradually decreased from the basal to apical parts of the stem in plants grown in Trichoderma-treated soil, indicating that Trichoderma stimulated a more significant effect on the defense system in the basal section of the stalk than in the apical parts of plants. Furthermore, the accumulation of deoxynivalenol (DON) and fumonisin B1 (FB1) decreased in the stem and ear of maize grown in T. asperellum-treated soil.

Fitting nonstationary general-time-reversible models to obtain edge-lengths and frequencies for the barry-hartigan model.

Among models of nucleotide evolution, the Barry and Hartigan (BH) model (also known as the General Markov Model) is very flexible as it allows separate arbitrary substitution matrices along edges. For a given tree, the estimates of the BH model are a set of joint probability matrices, each giving the pairwise frequencies of nucleotides at the ends of the edge. We have previously shown that, due to an identifiability problem, these cannot be expected to consistently estimate the actual pairwise frequencies. A further consequence is that internal node frequency estimates are likely to be incorrect. Here we define a nonstationary GTR model for each edge that we refer to as the NSGTR model. We fit the NSGTR model by minimizing the sums of squares between the estimates of transition probabilities under the NSGTR model and the estimates provided by a fitted BH model. This NSGTR model provides estimates that avoid the identifiability difficulties of the BH model while closely fitting it. With the best-fitting NSGTR estimates, we are able to get interpretable frequency vectors at internal nodes as well as edge length estimates that are otherwise not yielded by the BH model. These edge lengths are interpretable as the expected number of substitutions along an edge for the model. We also show that for a nonstationary continuous-time model these are not the same as the edge length parameters for conventional substitution matrices that are output by nonstationary model phylogenetic estimation programs such as nhPhyML.