Camera traps often produce massive images, and empty images that do not contain animals are usually overwhelming. Deep learning is a machine‐learning algorithm and widely used to identify empty camera trap images automatically. Existing methods with high accuracy are based on millions of training samples (images) and require a lot of time and personnel costs to label the training samples manually. Reducing the number of training samples can save the cost of manually labeling images. However, the deep learning models based on a small dataset produce a large omission error of animal images that many animal images tend to be identified as empty images, which may lead to loss of the opportunities of discovering and observing species. Therefore, it is still a challenge to build the DCNN model with small errors on a small dataset. Using deep convolutional neural networks and a small‐size dataset, we proposed an ensemble learning approach based on conservative strategies to identify and remove empty images automatically. Furthermore, we proposed three automatic identifying schemes of empty images for users who accept different omission errors of animal images. Our experimental results showed that these three schemes automatically identified and removed 50.78%, 58.48%, and 77.51% of the empty images in the dataset when the omission errors were 0.70%, 1.13%, and 2.54%, respectively. The analysis showed that using our scheme to automatically identify empty images did not omit species information. It only slightly changed the frequency of species occurrence. When only a small dataset was available, our approach provided an alternative to users to automatically identify and remove empty images, which can significantly reduce the time and personnel costs required to manually remove empty images. The cost savings were comparable to the percentage of empty images removed by models. 相似文献
Optical coherence tomography angiography (OCTA) offers a noninvasive label-free solution for imaging retinal vasculatures at the capillary level resolution. In principle, improved resolution implies a better chance to reveal subtle microvascular distortions associated with eye diseases that are asymptomatic in early stages. However, massive screening requires experienced clinicians to manually examine retinal images, which may result in human error and hinder objective screening. Recently, quantitative OCTA features have been developed to standardize and document retinal vascular changes. The feasibility of using quantitative OCTA features for machine learning classification of different retinopathies has been demonstrated. Deep learning-based applications have also been explored for automatic OCTA image analysis and disease classification. In this article, we summarize recent developments of quantitative OCTA features, machine learning image analysis, and classification. 相似文献
Photoacoustic computed tomography (PACT) is a non‐invasive imaging technique offering high contrast, high resolution, and deep penetration in biological tissues. We report a PACT system equipped with a high frequency linear transducer array for mapping the microvascular network of a whole mouse brain with the skull intact and studying its hemodynamic activities. The linear array was scanned in the coronal plane to collect data from different angles, and full‐view images were synthesized from the limited‐view images in which vessels were only partially revealed. We investigated spontaneous neural activities in the deep brain by monitoring the concentration of hemoglobin in the blood vessels and observed strong interhemispherical correlations between several chosen functional regions, both in the cortical layer and in the deep regions. We also studied neural activities during an epileptic seizure and observed the epileptic wave spreading around the injection site and the wave propagating in the opposite hemisphere.
本文提出了一种基于卷积神经网络和循环神经网络的深度学习模型,通过分析基因组序列数据,识别人基因组中环形RNA剪接位点.首先,根据预处理后的核苷酸序列,设计了2种网络深度、8种卷积核大小和3种长短期记忆(long short term memory,LSTM)参数,共8组16个模型;其次,进一步针对池化层进行均值池化和最大池化的测试,并加入GC含量提高模型的预测能力;最后,对已经实验验证过的人类精浆中环形RNA进行了预测.结果表明,卷积核尺寸为32×4、深度为1、LSTM参数为32的模型识别率最高,在训练集上为0.9824,在测试数据集上准确率为0.95,并且在实验验证数据上的正确识别率为83%.该模型在人的环形RNA剪接位点识别方面具有较好的性能. 相似文献
Photoacoustic computed tomography (PACT) has centimeter-level imaging ability and can be used to detect the human body. However, strong photoacoustic signals from skin cover deep tissue information, hindering the frontal display and analysis of photoacoustic images of deep regions of interest. Therefore, we propose a 2.5 D deep learning model based on feature pyramid structure and single-type skin annotation to extract the skin region, and design a mask generation algorithm to remove skin automatically. PACT imaging experiments on the human periphery blood vessel verified the correctness our proposed skin-removal method. Compared with previous studies, our method exhibits high robustness to the uneven illumination, irregular skin boundary, and reconstruction artifacts in the images, and the reconstruction errors of PACT images decreased by 20% ~ 90% with a 1.65 dB improvement in the signal-to-noise ratio at the same time. This study may provide a promising way for high-definition PACT imaging of deep tissues. 相似文献
This review covers original articles using deep learning in the biophotonic field published in the last years. In these years deep learning, which is a subset of machine learning mostly based on artificial neural network geometries, was applied to a number of biophotonic tasks and has achieved state‐of‐the‐art performances. Therefore, deep learning in the biophotonic field is rapidly growing and it will be utilized in the next years to obtain real‐time biophotonic decision‐making systems and to analyze biophotonic data in general. In this contribution, we discuss the possibilities of deep learning in the biophotonic field including image classification, segmentation, registration, pseudostaining and resolution enhancement. Additionally, we discuss the potential use of deep learning for spectroscopic data including spectral data preprocessing and spectral classification. We conclude this review by addressing the potential applications and challenges of using deep learning for biophotonic data. 相似文献
Non‐invasive photoacoustic tomography (PAT) of mouse brains with intact skulls has been a challenge due to the skull's strong acoustic attenuation, aberration, and reverberation, especially in the high‐frequency range (>15 MHz). In this paper, we systematically investigated the impacts of the murine skull on the photoacoustic wave propagation and on the PAT image reconstruction. We studied the photoacoustic acoustic wave aberration due to the acoustic impedance mismatch at the skull boundaries and the mode conversion between the longitudinal wave and shear wave. The wave's reverberation within the skull was investigated for both longitudinal and shear modes. In the inverse process, we reconstructed the transcranial photoacoustic computed tomography (PACT) and photoacoustic microscopy (PAM) images of a point target enclosed by the mouse skull, showing the skull's different impacts on both modalities. Finally, we experimentally validated the simulations by imaging an in vitro mouse skull phantom using representative transcranial PAM and PACT systems. The experimental results agreed well with the simulations and confirmed the accuracy of our forward and inverse models. We expect that our results will provide better understanding of the impacts of the murine skull on transcranial photoacoustic brain imaging and pave the ways for future technical improvements. 相似文献
A time‐consuming challenge faced by camera trap practitioners is the extraction of meaningful data from images to inform ecological management. An increasingly popular solution is automated image classification software. However, most solutions are not sufficiently robust to be deployed on a large scale due to lack of location invariance when transferring models between sites. This prevents optimal use of ecological data resulting in significant expenditure of time and resources to annotate and retrain deep learning models.
We present a method ecologists can use to develop optimized location invariant camera trap object detectors by (a) evaluating publicly available image datasets characterized by high intradataset variability in training deep learning models for camera trap object detection and (b) using small subsets of camera trap images to optimize models for high accuracy domain‐specific applications.
We collected and annotated three datasets of images of striped hyena, rhinoceros, and pigs, from the image‐sharing websites FlickR and iNaturalist (FiN), to train three object detection models. We compared the performance of these models to that of three models trained on the Wildlife Conservation Society and Camera CATalogue datasets, when tested on out‐of‐sample Snapshot Serengeti datasets. We then increased FiN model robustness by infusing small subsets of camera trap images into training.
In all experiments, the mean Average Precision (mAP) of the FiN trained models was significantly higher (82.33%–88.59%) than that achieved by the models trained only on camera trap datasets (38.5%–66.74%). Infusion further improved mAP by 1.78%–32.08%.
Ecologists can use FiN images for training deep learning object detection solutions for camera trap image processing to develop location invariant, robust, out‐of‐the‐box software. Models can be further optimized by infusion of 5%–10% camera trap images into training data. This would allow AI technologies to be deployed on a large scale in ecological applications. Datasets and code related to this study are open source and available on this repository: https://doi.org/10.5061/dryad.1c59zw3tx.
Counting animal populations is fundamental to understand ecological processes. Counts make it possible to estimate the size of an animal population at specific points in time, which is essential information for understanding demographic change. However, in the case of large populations, counts are time-consuming, particularly if carried out manually. Here, we took advantage of convolutional neural networks (CNN) to count the total number of nest-entrances in 222 photographs covering the largest known Psittaciformes (Aves) colony in the world. We conducted our study at the largest Burrowing Parrot Cyanoliseus patagonus colony, located on a cliff facing the Atlantic Ocean in the vicinity of El Cóndor village, in north-eastern Patagonia, Argentina. We also aimed to investigate the distribution of nest-entrances along the cliff with the colony. For this, we used three CNN architectures, U-Net, ResUnet, and DeepLabv3. The U-Net architecture showed the best performance, counting a mean of 59,842 Burrowing Parrot nest-entrances across the colony, with a mean absolute error of 2.7 nest-entrances over the testing patches, measured as the difference between actual and predicted counts per patch. Compared to a previous study conducted at El Cóndor colony more than 20 years ago, the CNN architectures also detected noteworthy differences in the distribution of the nest-entrances along the cliff. We show that the strong changes observed in the distribution of nest-entrances are a measurable effect of a long record of human-induced disturbance to the Burrowing Parrot colony at El Cóndor. Given the paramount importance of the Burrowing Parrot colony at El Cóndor, which concentrates 71% of the world's population of this species, we advocate that it is imperative to reduce such a degree of disturbance before the parrots reach the limit of their capacity of adaptation. 相似文献
The benchmark method for the evaluation of breast cancers involves microscopic testing of a hematoxylin and eosin (H&E)‐stained tissue biopsy. Resurgery is required in 20% to 30% of cases because of incomplete excision of malignant tissues. Therefore, a more accurate method is required to detect the cancer margin to avoid the risk of recurrence. In the recent years, convolutional neural networks (CNNs) has achieved excellent performance in the field of medical images diagnosis. It automatically extracts the features from the images and classifies them. In the proposed study, we apply a pretrained Inception‐v3 CNN with reverse active learning for the classification of healthy and malignancy breast tissue using optical coherence tomography (OCT) images. This proposed method attained the sensitivity, specificity and accuracy is 90.2%, 91.7% and 90%, respectively, with testing datasets collected from 48 patients (22 normal fibro‐adipose tissue and 26 Invasive ductal carcinomas cancerous tissues). The trained network utilizes for the breast cancer margin assessment to predict the tumor with negative margins. Additionally, the network output is correlated with the corresponding histology image. Our results lay the foundation for the future that the proposed method can be used to perform automatic intraoperative identification of breast cancer margins in real‐time and to guide core needle biopsies. 相似文献
We describe a novel scanning approach for miniaturized photoacoustic tomography (PAT), based on fan‐shaped scanning of a single transducer at one or two discrete positions. This approach is tested and evaluated using several phantom and animal experiments. The results obtained show that this new scanning approach provides high image quality in the configuration of miniaturized handheld or endoscopic PAT with improved effective field of view and penetration depth. 相似文献
Intrinsically disordered regions (IDR) play an important role in key biological processes and are closely related to human diseases. IDRs have great potential to serve as targets for drug discovery, most notably in disordered binding regions. Accurate prediction of IDRs is challenging because their genome wide occurrence and a low ratio of disordered residues make them difficult targets for traditional classification techniques. Existing computational methods mostly rely on sequence profiles to improve accuracy which is time consuming and computationally expensive. This article describes an ab initio sequence-only prediction method—which tries to overcome the challenge of accurate prediction posed by IDRs—based on reduced amino acid alphabets and convolutional neural networks (CNNs). We experiment with six different 3-letter reduced alphabets. We argue that the dimensional reduction in the input alphabet facilitates the detection of complex patterns within the sequence by the convolutional step. Experimental results show that our proposed IDR predictor performs at the same level or outperforms other state-of-the-art methods in the same class, achieving accuracy levels of 0.76 and AUC of 0.85 on the publicly available Critical Assessment of protein Structure Prediction dataset (CASP10). Therefore, our method is suitable for proteome-wide disorder prediction yielding similar or better accuracy than existing approaches at a faster speed. 相似文献
Wildlife conservation and the management of human–wildlife conflicts require cost‐effective methods of monitoring wild animal behavior. Still and video camera surveillance can generate enormous quantities of data, which is laborious and expensive to screen for the species of interest. In the present study, we describe a state‐of‐the‐art, deep learning approach for automatically identifying and isolating species‐specific activity from still images and video data.
We used a dataset consisting of 8,368 images of wild and domestic animals in farm buildings, and we developed an approach firstly to distinguish badgers from other species (binary classification) and secondly to distinguish each of six animal species (multiclassification). We focused on binary classification of badgers first because such a tool would be relevant to efforts to manage Mycobacterium bovis (the cause of bovine tuberculosis) transmission between badgers and cattle.
We used two deep learning frameworks for automatic image recognition. They achieved high accuracies, in the order of 98.05% for binary classification and 90.32% for multiclassification. Based on the deep learning framework, a detection process was also developed for identifying animals of interest in video footage, which to our knowledge is the first application for this purpose.
The algorithms developed here have wide applications in wildlife monitoring where large quantities of visual data require screening for certain species.
We report a framework based on a generative adversarial network that performs high‐fidelity color image reconstruction using a single hologram of a sample that is illuminated simultaneously by light at three different wavelengths. The trained network learns to eliminate missing‐phase‐related artifacts, and generates an accurate color transformation for the reconstructed image. Our framework is experimentally demonstrated using lung and prostate tissue sections that are labeled with different histological stains. This framework is envisaged to be applicable to point‐of‐care histopathology and presents a significant improvement in the throughput of coherent microscopy systems given that only a single hologram of the specimen is required for accurate color imaging. 相似文献
The standard medical practice for cancer diagnosis requires histopathology, which is an invasive and time-consuming procedure. Optical coherence tomography (OCT) is an alternative that is relatively fast, noninvasive, and able to capture three-dimensional structures of epithelial tissue. Unlike most previous OCT systems, which cannot capture crucial cellular-level information for squamous cell carcinoma (SCC) diagnosis, the full-field OCT (FF-OCT) technology used in this paper is able to produce images at sub-micron resolution and thereby facilitates the development of a deep learning algorithm for SCC detection. Experimental results show that the SCC detection algorithm can achieve a classification accuracy of 80% for mouse skin. Using the sub-micron FF-OCT imaging system, the proposed SCC detection algorithm has the potential for in-vivo applications. 相似文献
Deep learning offers new approaches to investigate the mechanisms underlying complex biological phenomena, such as subgenome dominance. Subgenome dominance refers to the dominant expression and/or biased fractionation of genes in one subgenome of allopolyploids, which has shaped the evolution of a large group of plants. However, the underlying cause of subgenome dominance remains elusive. Here, we adopt deep learning to construct two convolutional neural network (CNN) models, binary expression model (BEM) and homoeolog contrast model (HCM), to investigate the mechanism underlying subgenome dominance using DNA sequence and methylation sites. We apply these CNN models to analyze three representative polyploidization systems, Brassica, Gossypium, and Cucurbitaceae, each with available ancient and neo/synthetic polyploidized genomes. The BEM shows that DNA sequence of the promoter region can accurately predict whether a gene is expressed or not. More importantly, the HCM shows that the DNA sequence of the promoter region predicts dominant expression status between homoeologous gene pairs retained from ancient polyploidizations, thus predicting subgenome dominance associated with these events. However, HCM fails to predict gene expression dominance between new homoeologous gene pairs arising from the neo/synthetic polyploidizations. These results are consistent across the three plant polyploidization systems, indicating broad applicability of our models. Furthermore, the two models based on methylation sites produce similar results. These results show that subgenome dominance is associated with long-term sequence differentiation between the promoters of homoeologs, suggesting that subgenome expression dominance precedes and is the driving force or even the determining factor for sequence divergence between subgenomes following polyploidization. 相似文献
