Introduction This study explores the use of the latest You Only Look Once (YOLO V7) object detection method to enhance kidney detection in medical imaging by training and testing a modified YOLO V7 on medical image formats. Methods Study includes 878 patients with various subtypes of renal cell carcinoma (RCC) and 206 patients with normal kidneys. A total of 5657 MRI scans for 1084 patients were retrieved. 326 patients with 1034 tumors recruited from a retrospective maintained database, and bounding boxes were drawn around their tumors. A primary model was trained on 80% of annotated cases, with 20% saved for testing (primary test set). The best primary model was then used to identify tumors in the remaining 861 patients and bounding box coordinates were generated on their scans using the model. Ten benchmark training sets were created with generated coordinates on not-segmented patients. The final model used to predict the kidney in the primary test set. We reported the positive predictive value (PPV), sensitivity, and mean average precision (mAP). Results The primary training set showed an average PPV of 0.94 +/- 0.01, sensitivity of 0.87 +/- 0.04, and mAP of 0.91 +/- 0.02. The best primary model yielded a PPV of 0.97, sensitivity of 0.92, and mAP of 0.95. The final model demonstrated an average PPV of 0.95 +/- 0.03, sensitivity of 0.98 +/- 0.004, and mAP of 0.95 +/- 0.01. Conclusion Using a semi-supervised approach with a medical image library, we developed a high-performing model for kidney detection. Further external validation is required to assess the model's generalizability.
Recent advancements in self-supervised learning have unlocked the potential to harness unlabeled data for auxiliary tasks, facilitating the learning of beneficial priors. This has been particularly advantageous in fields like medical image analysis, where labeled data are scarce. Although effective for classification tasks, this methodology has shown limitations in more complex applications, such as medical image segmentation. In this paper, we introduce Medical imaging Enhanced with Dynamic Self-Adaptive Semantic Segmentation (MedSASS), a dedicated self-supervised framework tailored for medical image segmentation. We evaluate MedSASS against existing state-of-the-art methods across four diverse medical datasets, showcasing its superiority. MedSASS outperforms existing CNN-based self-supervised methods by 3.83% and matches the performance of ViT-based methods. Furthermore, when MedSASS is trained end-to-end, covering both encoder and decoder, it demonstrates significant improvements of 14.4% for CNNs and 6% for ViT-based architectures compared to existing state-of-the-art self-supervised strategies.
Image summary, an abridged version of the original visual content, can be used to represent the scene. Thus, tasks such as scene classification, identification, indexing, etc., can be performed efficiently using the unique summary. Saliency is the most commonly used technique for generating the relevant image summary. However, the definition of saliency is subjective in nature and depends upon the application. Existing saliency detection methods using RGB-D data mainly focus on color, texture, and depth features. Consequently, the generated summary contains either foreground objects or non-stationary objects. However, applications such as scene identification require stationary characteristics of the scene, unlike state-of-the-art methods. This paper proposes a novel volumetric saliency-guided framework for indoor scene classification. The results highlight the efficacy of the proposed method.
Reward engineering has long been a challenge in Reinforcement Learning (RL) research, as it often requires extensive human effort and iterative processes of trial-and-error to design effective reward functions. In this paper, we propose RL-VLM-F, a method that automatically generates reward functions for agents to learn new tasks, using only a text description of the task goal and the agent's visual observations, by leveraging feedbacks from vision language foundation models (VLMs). The key to our approach is to query these models to give preferences over pairs of the agent's image observations based on the text description of the task goal, and then learn a reward function from the preference labels, rather than directly prompting these models to output a raw reward score, which can be noisy and inconsistent. We demonstrate that RL-VLM-F successfully produces effective rewards and policies across various domains - including classic control, as well as manipulation of rigid, articulated, and deformable objects - without the need for human supervision, outperforming prior methods that use large pretrained models for reward generation under the same assumptions.
To defend against privacy leakage of user data, differential privacy is widely used in federated learning, but it is not free. The addition of noise randomly disrupts the semantic integrity of the model and this disturbance accumulates with increased communication rounds. In this paper, we introduce a novel federated learning framework with rigorous privacy guarantees, named FedCEO, designed to strike a trade-off between model utility and user privacy by letting clients ''Collaborate with Each Other''. Specifically, we perform efficient tensor low-rank proximal optimization on stacked local model parameters at the server, demonstrating its capability to flexibly truncate high-frequency components in spectral space. This implies that our FedCEO can effectively recover the disrupted semantic information by smoothing the global semantic space for different privacy settings and continuous training processes. Moreover, we improve the SOTA utility-privacy trade-off bound by an order of $\sqrt{d}$, where $d$ is the input dimension. We illustrate our theoretical results with experiments on representative image datasets. It observes significant performance improvements and strict privacy guarantees under different privacy settings.
Short-form UGC video platforms, like Kwai and TikTok, have been an emerging and irreplaceable mainstream media form, thriving on user-friendly engagement, and kaleidoscope creation, etc. However, the advancing content-generation modes, e.g., special effects, and sophisticated processing workflows, e.g., de-artifacts, have introduced significant challenges to recent UGC video quality assessment: (i) the ambiguous contents hinder the identification of quality-determined regions. (ii) the diverse and complicated hybrid distortions are hard to distinguish. To tackle the above challenges and assist in the development of short-form videos, we establish the first large-scale Kaleidoscope short Video database for Quality assessment, termed KVQ, which comprises 600 user-uploaded short videos and 3600 processed videos through the diverse practical processing workflows, including pre-processing, transcoding, and enhancement. Among them, the absolute quality score of each video and partial ranking score among indistinguishable samples are provided by a team of professional researchers specializing in image processing. Based on this database, we propose the first short-form video quality evaluator, i.e., KSVQE, which enables the quality evaluator to identify the quality-determined semantics with the content understanding of large vision language models (i.e., CLIP) and distinguish the distortions with the distortion understanding module. Experimental results have shown the effectiveness of KSVQE on our KVQ database and popular VQA databases.
Treatment planning is currently a patient specific, time-consuming, and resource demanding task in radiotherapy. Dose-volume histogram (DVH) prediction plays a critical role in automating this process. The geometric relationship between DVHs in radiotherapy plans and organs-at-risk (OAR) and planning target volume (PTV) has been well established. This study explores the potential of deep learning models for predicting DVHs using images and subsequent human intervention facilitated by a large-language model (LLM) to enhance the planning quality. We propose a pipeline to convert unstructured images to a structured graph consisting of image-patch nodes and dose nodes. A novel Dose Graph Neural Network (DoseGNN) model is developed for predicting DVHs from the structured graph. The proposed DoseGNN is enhanced with the LLM to encode massive knowledge from prescriptions and interactive instructions from clinicians. In this study, we introduced an online human-AI collaboration (OHAC) system as a practical implementation of the concept proposed for the automation of intensity-modulated radiotherapy (IMRT) planning. In comparison to the widely-employed DL models used in radiotherapy, DoseGNN achieved mean square errors that were 80$\%$, 76$\%$ and 41.0$\%$ of those predicted by Swin U-Net Transformer, 3D U-Net CNN and vanilla MLP, respectively. Moreover, the LLM-empowered DoseGNN model facilitates seamless adjustment to treatment plans through interaction with clinicians using natural language.
Despite the recent advances in large-scale diffusion models, little progress has been made on the layout-to-image (L2I) synthesis task. Current L2I models either suffer from poor editability via text or weak alignment between the generated image and the input layout. This limits their usability in practice. To mitigate this, we propose to integrate adversarial supervision into the conventional training pipeline of L2I diffusion models (ALDM). Specifically, we employ a segmentation-based discriminator which provides explicit feedback to the diffusion generator on the pixel-level alignment between the denoised image and the input layout. To encourage consistent adherence to the input layout over the sampling steps, we further introduce the multistep unrolling strategy. Instead of looking at a single timestep, we unroll a few steps recursively to imitate the inference process, and ask the discriminator to assess the alignment of denoised images with the layout over a certain time window. Our experiments show that ALDM enables layout faithfulness of the generated images, while allowing broad editability via text prompts. Moreover, we showcase its usefulness for practical applications: by synthesizing target distribution samples via text control, we improve domain generalization of semantic segmentation models by a large margin (~12 mIoU points).
Accurate medical image segmentation demands the integration of multi-scale information, spanning from local features to global dependencies. However, it is challenging for existing methods to model long-range global information, where convolutional neural networks (CNNs) are constrained by their local receptive fields, and vision transformers (ViTs) suffer from high quadratic complexity of their attention mechanism. Recently, Mamba-based models have gained great attention for their impressive ability in long sequence modeling. Several studies have demonstrated that these models can outperform popular vision models in various tasks, offering higher accuracy, lower memory consumption, and less computational burden. However, existing Mamba-based models are mostly trained from scratch and do not explore the power of pretraining, which has been proven to be quite effective for data-efficient medical image analysis. This paper introduces a novel Mamba-based model, Swin-UMamba, designed specifically for medical image segmentation tasks, leveraging the advantages of ImageNet-based pretraining. Our experimental results reveal the vital role of ImageNet-based training in enhancing the performance of Mamba-based models. Swin-UMamba demonstrates superior performance with a large margin compared to CNNs, ViTs, and latest Mamba-based models. Notably, on AbdomenMRI, Encoscopy, and Microscopy datasets, Swin-UMamba outperforms its closest counterpart U-Mamba by an average score of 3.58%. The code and models of Swin-UMamba are publicly available at: https://github.com/JiarunLiu/Swin-UMamba
Large language models (LLMs) have recently experienced remarkable progress, where the advent of multi-modal large language models (MLLMs) has endowed LLMs with visual capabilities, leading to impressive performances in various multi-modal tasks. However, those powerful MLLMs such as GPT-4V still fail spectacularly when presented with certain image and text inputs. In this paper, we identify a typical class of inputs that baffles MLLMs, which consist of images that are highly relevant but inconsistent with answers, causing MLLMs to suffer from hallucination. To quantify the effect, we propose CorrelationQA, the first benchmark that assesses the hallucination level given spurious images. This benchmark contains 7,308 text-image pairs across 13 categories. Based on the proposed CorrelationQA, we conduct a thorough analysis on 9 mainstream MLLMs, illustrating that they universally suffer from this instinctive bias to varying degrees. We hope that our curated benchmark and evaluation results aid in better assessments of the MLLMs' robustness in the presence of misleading images. The resource is available in https://github.com/MasaiahHan/CorrelationQA.