In current genomic research, the widely used methods for predicting antimicrobial resistance (AMR) often rely on prior knowledge of known AMR genes or reference genomes. However, these methods have limitations, potent...
详细信息
The precise reconstruction of 3D objects from a single RGB image in complex scenes presents a critical challenge in virtual reality, autonomous driving, and robotics. Existing neural implicit 3D representation methods...
详细信息
Many patients with leukoaraiosis (LA) exhibit mild and difficult-to-detect symptoms in the early stages, and due to the lack of effective detection methods, the optimal timing for treatment is often missed. This not o...
详细信息
In recent years, many fields have expanded their research methods through the integration of artificial intelligence. In the current medical field, it is widely used in image recognition to diagnose patient symptoms, ...
详细信息
Kidney stones are primarily crystals formed from ion oversaturation in urine. Currently, the diagnosis of kidney stones involves experienced professionals manually interpreting images of urinary crystals under a micro...
详细信息
Alzheimer’s disease (AD) often presents only mild symptoms in its early stages, and as there is no direct diagnostic method currently available, many patients are diagnosed only after the condition has worsened. Cons...
详细信息
We consider nonstationary multi-armed bandit problems where the model parameters of the arms change over time. We introduce the adaptive resetting bandit (ADR-bandit), a bandit algorithm class that leverages adaptive ...
详细信息
We consider nonstationary multi-armed bandit problems where the model parameters of the arms change over time. We introduce the adaptive resetting bandit (ADR-bandit), a bandit algorithm class that leverages adaptive windowing techniques from literature on data streams. We first provide new guarantees on the quality of estimators resulting from adaptive windowing techniques, which are of independent interest. Furthermore, we conduct a finite-time analysis of ADR-bandit in two typical environments: an abrupt environment where changes occur instantaneously and a gradual environment where changes occur progressively. We demonstrate that ADR-bandit has nearly optimal performance when abrupt or gradual changes occur in a coordinated manner that we call global changes. We demonstrate that forced exploration is unnecessary when we assume such global changes. Unlike the existing nonstationary bandit algorithms, ADR-bandit has optimal performance in stationary environments as well as nonstationary environments with global changes. Our experiments show that the proposed algorithms outperform the existing approaches in synthetic and real-world environments.
This study explores the feasibility of deep learning for classifying nodule neoplasms, analyzing their performance on two openly available datasets, LUNGx SPIE, and LIDC-IDRI. These datasets offer valuable diversity i...
详细信息
Narrowing of the coronary arteries is an important indicator of the severity of coronary artery disease (CAD) in patients. Previous research using deep learning to identify narrowed vessels has primarily been based on...
详细信息
In the realm of research, the global health challenge posed by lung cancer remains pronounced, contributing substantially to annual cancer-related fatalities. The critical imperative lies in the early identification o...
详细信息
ISBN:
(纸本)9798400716874
In the realm of research, the global health challenge posed by lung cancer remains pronounced, contributing substantially to annual cancer-related fatalities. The critical imperative lies in the early identification of pulmonary nodules, frequently indicative of impending lung cancer, to enhance patient outcomes and diminish mortality rates. Computed Tomography (CT) imaging stands out as a pivotal diagnostic instrument for the timely detection of these nodules. The swift proliferation of medical imaging data has underscored the pressing necessity for precise and efficient methodologies dedicated to nodule segmentation and measurement. These approaches are crucial in assisting radiologists in their diagnostic and clinical decision-making endeavors. In this study, we introduced a thorough method for analyzing lung nodules, leveraging dataset from Far Eastern Memorial Hospital (FEMH) comprising original CT images and manually annotated ground truth masks obtained with the assistance of radiologists at FEMH. This dataset is utilized for the segmentation of nodules. We employed advanced deep learning models, specifically the U-Net architecture, identified as the optimal model through our training process. We made substantial progress in nodule segmentation, attaining an Intersection over Union (IoU) score of 0.824 and a Dice Coefficient of 0.903 for the FEMH dataset. Furthermore, our performance improved when utilizing the merged dataset comprising FEMH and Luna16, yielding an IoU score of 0.862 and a Dice Coefficient of 0.926. Luna16 has been extensively utilized in numerous studies related to nodule detection and segmentation. In the next phase of the study, the best-performing model from our segmentation phase was utilized to predict nodule masks on the merged dataset. Subsequently, we measured the size of each predicted nodule by comparing it with the size ground truth mask in millimeters. In detail, this study achieved the Pearson Correlation Coefficient (PCC) at 0.
暂无评论