Improved FAST algorithm for non-uniform rotational distortion correction in OCT endoscopic imaging.

Optical Coherence Tomography (OCT) is widely used for endoscopic imaging in endoluminal organs because of its high imaging accuracy and resolution. However, OCT endoscopic imaging suffers from Non-Uniform Rotational Distortion (NURD), which can be caused by many factors, such as irregular motor rotation and changes in friction between the probe and the sheath. Correcting this distortion is essential to obtaining high-quality Optical Coherence Tomography Angiography (OCTA) images. There are two main approaches for correcting NURD: hardware-based methods and algorithm-based methods. Hardware-based methods can be costly, challenging to implement, and may not eliminate NURD. Algorithm-based methods, such as image registration, can be effective for correcting NURD but can also be prone to the problem of NURD propagation. To address this issue, we process frames by coarse and fine registration, respectively. The new reference frame is generated by filtering out the A-scan that may have the NURD problem by coarse registration. And the fine registration uses this frame to achieve the final NURD correction. In addition, we have improved the Features from Accelerated Segment Test (FAST) algorithm and put it into coarse and fine registration process. Four evaluation functions were used for the experimental results, including signal-to-noise ratio (SNR), peak signal-to-noise ratio (PSNR), mean squared error (MSE), and structural similarity index measure (SSIM). By comparing with Scale-invariant feature transform (SIFT), Speeded up robust features (SURF), Oriented FAST and Rotated BRIEF (ORB), intensity-based (Cross-correlation), and Optical Flow algorithms, our algorithm has a higher similarity between the corrected frames. Moreover, the noise in the OCTA data is better suppressed, and the vascular information is well preserved. Our image registration-based algorithm reduces the problem of NURD propagation between B-scan frames and improves the imaging quality of OCT endoscopic images.

DaylilyNet: A Multi-Task Learning Method for Daylily Leaf Disease Detection.

Timely detection and management of daylily diseases are crucial to prevent yield reduction. However, detection models often struggle with handling the interference of complex backgrounds, leading to low accuracy, especially in detecting small targets. To address this problem, we propose DaylilyNet, an object detection algorithm that uses multi-task learning to optimize the detection process. By incorporating a semantic segmentation loss function, the model focuses its attention on diseased leaf regions, while a spatial global feature extractor enhances interactions between leaf and background areas. Additionally, a feature alignment module improves localization accuracy by mitigating feature misalignment. To investigate the impact of information loss on model detection performance, we created two datasets. One dataset, referred to as the 'sliding window dataset', was obtained by splitting the original-resolution images using a sliding window. The other dataset, known as the 'non-sliding window dataset', was obtained by downsampling the images. Experimental results in the 'sliding window dataset' and the 'non-sliding window dataset' demonstrate that DaylilyNet outperforms YOLOv5-L in mAP@0.5 by 5.2% and 4.0%, while reducing parameters and time cost. Compared to other models, our model maintains an advantage even in scenarios where there is missing information in the training dataset.

Multi-label classification of retinal diseases based on fundus images using Resnet and Transformer.

Retinal disorders are a major cause of irreversible vision loss, which can be mitigated through accurate and early diagnosis. Conventionally, fundus images are used as the gold diagnosis standard in detecting retinal diseases. In recent years, more and more researchers have employed deep learning methods for diagnosing ophthalmic diseases using fundus photography datasets. Among the studies, most of them focus on diagnosing a single disease in fundus images, making it still challenging for the diagnosis of multiple diseases. In this paper, we propose a framework that combines ResNet and Transformer for multi-label classification of retinal disease. This model employs ResNet to extract image features, utilizes Transformer to capture global information, and enhances the relationships between categories through learnable label embedding. On the publicly available Ocular Disease Intelligent Recognition (ODIR-5 k) dataset, the proposed method achieves a mean average precision of 92.86%, an area under the curve (AUC) of 97.27%, and a recall of 90.62%, which outperforms other state-of-the-art approaches for the multi-label classification. The proposed method represents a significant advancement in the field of retinal disease diagnosis, offering a more accurate, efficient, and comprehensive model for the detection of multiple retinal conditions.

Self-attention CNN for retinal layer segmentation in OCT.

The structure of the retinal layers provides valuable diagnostic information for many ophthalmic diseases. Optical coherence tomography (OCT) obtains cross-sectional images of the retina, which reveals information about the retinal layers. The U-net based approaches are prominent in retinal layering methods, which are usually beneficial to local characteristics but not good at obtaining long-distance dependence for contextual information. Furthermore, the morphology of retinal layers with the disease is more complex, which brings more significant challenges to the task of retinal layer segmentation. We propose a U-shaped network combining an encoder-decoder architecture and self-attention mechanisms. In response to the characteristics of retinal OCT cross-sectional images, a self-attentive module in the vertical direction is added to the bottom of the U-shaped network, and an attention mechanism is also added in skip connection and up-sampling to enhance essential features. In this method, the transformer's self-attentive mechanism obtains the global field of perception, thus providing the missing context information for convolutions, and the convolutional neural network also efficiently extracts local features, compensating the local details the transformer ignores. The experiment results showed that our method is accurate and better than other methods for segmentation of the retinal layers, with the average Dice scores of 0.871 and 0.820, respectively, on two public retinal OCT image datasets. To perform the layer segmentation of retinal OCT image better, the proposed method incorporates the transformer's self-attention mechanism in a U-shaped network, which is helpful for ophthalmic disease diagnosis.

[Accelerated algorithm of dose calculation for treatment planning system of multi-source stereotactic radiotherapy system with gamma beam].

This paper reviewed the dose calculation method for treatment planning system of multi-source stereotactic radiotherapy system with Gamma beam, and presented a fast method based on cone coordinate system and multi-core parallelization. The experiments show that the new method not only ensures the accuracy of dose calculation, but also greatly improves the speed, which can reach as 20 times as that of traditional method.

A single-step regression method based on transformer for retinal layer segmentation.

The shape and structure of retinal layers are basic characteristics for the diagnosis of many ophthalmological diseases. Based on B-Scans of optical coherence tomography, most of retinal layer segmentation methods are composed of two-steps: classifying pixels and extracting retinal layers, in which the optimization of two independent steps decreases the accuracy. Although the methods based on deep learning are highly accurate, they require a large amount of labeled data. This paper proposes a single-step method based on transformer for retinal layer segmentation, which is trained by axial data (A-Scans), to obtain the boundary of each layer. The proposed method was evaluated on two public data sets. The first one contains eight retinal layer boundaries for diabetic macular edema, and the second one contains nine retinal layer boundaries for healthy controls and subjects with multiple sclerosis. Its absolute average distance errors are 0.99 pixels and 3.67 pixels, respectively, for the two sets, and its root mean square error is 1.29 pixels for the latter set. In addition, its accuracy is acceptable even if the training data is reduced to 0.3. The proposed method achieves state-of-the-art performance while maintaining the correct topology and requires less labeled data.

[Non-rigid medical image registration based on mutual information and thin-plate spline].

To get precise and complete details, the contrast in different images is needed in medical diagnosis and computer assisted treatment. The image registration is the basis of contrast, but the regular rigid registration does not satisfy the clinic requirements. A non-rigid medical image registration method based on mutual information and thin-plate spline was present. Firstly, registering two images globally based on mutual information; secondly, dividing reference image and global-registered image into blocks and registering them; then getting the thin-plate spline transformation according to the shift of blocks' center; finally, applying the transformation to the global-registered image. The results show that the method is more precise than the global rigid registration based on mutual information and it reduces the complexity of getting control points and satisfy the clinic requirements better by getting control points of the thin-plate transformation automatically.