Search

Early detection of eye diseases like glaucoma, macular degeneration, and diabetic retinopathy is crucial for preventing vision loss. While artificial intelligence (AI) foundation models hold significant promise for addressing these challenges, existing ophthalmic foundation models primarily focus on a single modality, whereas diagnosing eye diseases requires multiple modalities. A critical yet often overlooked aspect is harnessing the multi-view information across various modalities for the same patient. Additionally, due to the long-tail nature of ophthalmic diseases, standard fully supervised or unsupervised learning approaches often struggle. Therefore, it is essential to integrate clinical text to capture a broader spectrum of diseases. We propose EyeCLIP, a visual-language foundation model developed using over 2.77 million multi-modal ophthalmology images with partial text data. To fully leverage the large multi-modal unlabeled and labeled data, we introduced a pretraining strategy that combines self-supervised reconstructions, multi-modal image contrastive learning, and image-text contrastive learning to learn a shared representation of multiple modalities. Through evaluation using 14 benchmark datasets, EyeCLIP can be transferred to a wide range of downstream tasks involving ocular and systemic diseases, achieving state-of-the-art performance in disease classification, visual question answering, and cross-modal retrieval. EyeCLIP represents a significant advancement over previous methods, especially showcasing few-shot, even zero-shot capabilities in real-world long-tail scenarios.

We proposed a low-complexity SVM-based signal recovery algorithm and evaluated it in 100G-PON with 25G-class devices. For the first time, it experimentally achieved 24 dB power budget @ FEC threshold 1E-3 over 40 km SMF, improving receiver sensitivity over 2 dB compared to FFE&DFE., Comment: I have discovered issues with the paper and need to rewrite parts of it, hence I am applying to withdraw the submission

We proposed and experimentally demonstrated a look-up table boosted fast CDR and equalization scheme for the burst-mode 50/100 Gbps bandwidth-limited flexible PON, requiring no preamble for convergence and achieved the same bit error rate performance as in the case of long preambles.

We consider scheduling two types of jobs (A-job and B-job) to $p$ machines and minimizing their makespan. A group of same type of jobs processed consecutively by a machine is called a batch. For machine $v$, processing $x$ A-jobs in a batch takes $k^A_vx^2$ time units for a given speed $k^A_v$, and processing $x$ B-jobs in a batch takes $k^B_vx^2$ time units for a given speed $k^B_v$. We give an $O(n^2p\log(n))$ algorithm based on dynamic programming and binary search for solving this problem, where $n$ denotes the maximal number of A-jobs and B-jobs to be distributed to the machines. Our algorithm also fits the easier linear case where each batch of length $x$ of $A$-jobs takes $k^A_v x$ time units and each batch of length $x$ of $B$-jobs takes $k^B_vx$ time units. The running time is the same as the above case.

In project management, a project is typically described as an activity-on-edge network (AOE network), where each activity / job is represented as an edge of some network $N$ (which is a DAG). Some jobs must be finished before others can be started, as described by the topology structure of $N$. It is known that job $j_i$ in normal speed would require $b_i$ days to be finished after it is started. Given the network $N$ with the associated edge lengths $b_1,\ldots,b_m$, the duration of the project is determined, which equals the length of the critical path (namely, the longest path) of $N$. To speed up the project (i.e. reduce the duration), the manager can crash a few jobs (namely, reduce the length of the corresponding edges) by investing extra resources into that job. However, the time for completing $j_i$ has a lower bound due to technological limits -- it requires at least $a_i$ days to be completed. Moreover, it is expensive to buy resources. Given $N$ and an integer $k\geq 1$, the $k$-crashing problem asks the minimum amount of resources required to speed up the project by $k$ days. We show a simple and efficient algorithm with an approximation ratio $\frac{1}{1}+\ldots+\frac{1}{k}$ for this problem. We also study a related problem called $k$-LIS, in which we are given a sequence $\omega$ of numbers and we aim to find $k$ disjoint increasing subsequence of $\omega$ with the largest total length. We show a $(1-\frac{1}{e})$-approximation algorithm which is simple and efficient.

Cross-modal medical image-report retrieval task plays a significant role in clinical diagnosis and various medical generative tasks. Eliminating heterogeneity between different modalities to enhance semantic consistency is the key challenge of this task. The current Vision-Language Pretraining (VLP) models, with cross-modal contrastive learning and masked reconstruction as joint training tasks, can effectively enhance the performance of cross-modal retrieval. This framework typically employs dual-stream inputs, using unmasked data for cross-modal contrastive learning and masked data for reconstruction. However, due to task competition and information interference caused by significant differences between the inputs of the two proxy tasks, the effectiveness of representation learning for intra-modal and cross-modal features is limited. In this paper, we propose an efficient VLP framework named Masked Contrastive and Reconstruction (MCR), which takes masked data as the sole input for both tasks. This enhances task connections, reducing information interference and competition between them, while also substantially decreasing the required GPU memory and training time. Moreover, we introduce a new modality alignment strategy named Mapping before Aggregation (MbA). Unlike previous methods, MbA maps different modalities to a common feature space before conducting local feature aggregation, thereby reducing the loss of fine-grained semantic information necessary for improved modality alignment. Qualitative and quantitative experiments conducted on the MIMIC-CXR dataset validate the effectiveness of our approach, demonstrating state-of-the-art performance in medical cross-modal retrieval tasks.

Automatic segmentation of fluid in Optical Coherence Tomography (OCT) images is beneficial for ophthalmologists to make an accurate diagnosis. Although semi-supervised OCT fluid segmentation networks enhance their performance by introducing additional unlabeled data, the performance enhancement is limited. To address this, we propose Superpixel and Confident Learning Guide Point Annotations Network (SCLGPA-Net) based on the teacher-student architecture, which can learn OCT fluid segmentation from limited fully-annotated data and abundant point-annotated data. Specifically, we use points to annotate fluid regions in unlabeled OCT images and the Superpixel-Guided Pseudo-Label Generation (SGPLG) module generates pseudo-labels and pixel-level label trust maps from the point annotations. The label trust maps provide an indication of the reliability of the pseudo-labels. Furthermore, we propose the Confident Learning Guided Label Refinement (CLGLR) module identifies error information in the pseudo-labels and leads to further refinement. Experiments on the RETOUCH dataset show that we are able to reduce the need for fully-annotated data by 94.22\%, closing the gap with the best fully supervised baselines to a mean IoU of only 2\%. Furthermore, We constructed a private 2D OCT fluid segmentation dataset for evaluation. Compared with other methods, comprehensive experimental results demonstrate that the proposed method can achieve excellent performance in OCT fluid segmentation., Comment: Submission to MIA

The application of sodium laser guide star is the key difference between modern adaptive optics system and traditional adaptive optics system. Especially in system like multi-conjugate adaptive optics, sodium laser guide star asterism which is formed by several laser guide stars in certain pattern is required to probe more atmospheric turbulence in different directions. To achieve this, a sodium laser guide star asterism launching platform is required. In this paper, we will introduce the sodium laser guide star asterism launching platform built and tested on the 1.8m telescope of the Gaomeigu Observatory. The platform has two functions: one is to compare the performance of sodium laser guide stars generated by different lasers at the same place; the other is to generate sodium laser guide star asterism with adjustable shape. The field test results at the beginning of 2021 verify the important role of the platform, which is also the first time to realize sodium laser guide star asterism in China.

Although the growing electric vehicle (EV) population is leading us into a more sustainable world, it is also bringing challenges for the manufacturers's production planning, the charging facility providers's expansion plan, and the energy system's adaption to greater electricity demand. To tackle these challenges, a model to predict EV growth in geographical scope would be helpful. In this study, an agent-based parameterized bass network diffusion model was developed for EV population data in Washington. The model included income levels and number of neighbors adopted as two key factors in determining EV diffusion probabilities. With the parameters estimated from simulation, the resulting model achieve a high estimation accuracy for EV adoption in Washington in both temporal and geographical scopes. This model could be used to predict EV growth in Washington, and to be adopted to other geographical areas.

Architected materials can achieve enhanced properties compared to their plain counterparts. Specific architecting serves as a powerful design lever to achieve targeted behavior without changing the base material. Thus, the connection between architected structure and resultant properties remains an open field of great interest to many fields, from aerospace to civil to automotive applications. Here, we focus on properties related to mechanical compression, and design hierarchical honeycomb structures to meet specific values of stiffness and compressive stress. To do so, we employ a combination of techniques in a singular workflow, starting with molecular dynamics simulation of the forward design problem, augmenting with data-driven artificial intelligence models to address the inverse design problem, and verifying the behavior of de novo structures with experimentation of additively manufactured samples. We thereby demonstrate an approach for architected design that is generalizable to multiple material properties and agnostic to the identity of the base material.

This paper describes the adaptation of an open-source ecological momentary assessment smart-watch platform with three sets of micro-survey wellness-related questions focused on i) infectious disease (COVID-19) risk perception, ii) privacy and distraction in an office context, and iii) triggers of various movement-related behaviors in buildings. This platform was previously used to collect data for thermal comfort, and this work extends its use to other domains. Several research participants took part in a proof-of-concept experiment by wearing a smartwatch to collect their micro-survey question preferences and perception responses for two of the question sets. Participants were also asked to install an indoor localization app on their phone to detect where precisely in the building they completed the survey. The experiment identified occupant information such as the tendencies for the research participants to prefer privacy in certain spaces and the difference between infectious disease risk perception in naturally versus mechanically ventilated spaces.

Sequence partition problems arise in many fields, such as sequential data analysis, information transmission, and parallel computing. In this paper, we study the following partition problem variant: given a sequence of $n$ items $1,\ldots,n$, where each item $i$ is associated with weight $w_i$ and another parameter $s_i$, partition the sequence into several consecutive subsequences, so that the total weight of each subsequence is no more than a threshold $w_0$, and the sum of the largest $s_i$ in each subsequence is minimized. This problem admits a straightforward solution based on dynamic programming, which costs $O(n^2)$ time and can be improved to $O(n\log n)$ time easily. Our contribution is an $O(n)$ time algorithm, which is nontrivial yet easy to implement. We also study the corresponding tree partition problem. We prove that the problem on the tree is NP-complete and we present an $O(w_0 n^2)$ time ($O(w_0^2n^2)$ time, respectively) algorithm for the unit weight (integer weight, respectively) case.

We study the following rearrangement problem: Given $n$ words, rearrange and concatenate them so that the obtained string is lexicographically smallest (or largest, respectively). We show that this problem reduces to sorting the given words so that their repeating strings are non-decreasing (or non-increasing, respectively), where the repeating string of a word $A$ refers to the infinite string $AAA\ldots$. Moreover, for fixed size alphabet $\Sigma$, we design an $O(L)$ time sorting algorithm of the words (in the mentioned orders), where $L$ denotes the total length of the input words. Hence we obtain an $O(L)$ time algorithm for the rearrangement problem. Finally, we point out that comparing primitive words via comparing their repeating strings leads to a total order, which can further be extended to a total order on the finite words (or all words).

Capacity of dense coding via correlated noisy channel is greater than that in uncorrelated noisy channel. It is shown that weak measurement and reversal measurement can make further effort to improve quantum dense coding capacity in correlated amplitude damping channel, but this effort is very small in correlated phase damping channel and correlated depolarizing channel., Comment: 13 pages, 4 figures

The dynamics of coherence in a two-qubit Heisenberg XXX model with magnetic field in the presence of the intrinsic decoherence is investigated. We discuss the influence of dipole parameter, spin distance, magnetic field and initial state parameter on coherence of the long time limit situation. Then, we discuss how decoherence changes with time with the initial state parameter and the decoherence parameter., Comment: 11 pages,7 figures,Accepted by Physica A

Accurate segmentation is a crucial step in medical image analysis and applying supervised machine learning to segment the organs or lesions has been substantiated effective. However, it is costly to perform data annotation that provides ground truth labels for training the supervised algorithms, and the high variance of data that comes from different domains tends to severely degrade system performance over cross-site or cross-modality datasets. To mitigate this problem, a novel unsupervised domain adaptation (UDA) method named dispensed Transformer network (DTNet) is introduced in this paper. Our novel DTNet contains three modules. First, a dispensed residual transformer block is designed, which realizes global attention by dispensed interleaving operation and deals with the excessive computational cost and GPU memory usage of the Transformer. Second, a multi-scale consistency regularization is proposed to alleviate the loss of details in the low-resolution output for better feature alignment. Finally, a feature ranking discriminator is introduced to automatically assign different weights to domain-gap features to lessen the feature distribution distance, reducing the performance shift of two domains. The proposed method is evaluated on large fluorescein angiography (FA) retinal nonperfusion (RNP) cross-site dataset with 676 images and a wide used cross-modality dataset from the MM-WHS challenge. Extensive results demonstrate that our proposed network achieves the best performance in comparison with several state-of-the-art techniques., Comment: 10 pages