5 September 2022 in Knowledge-Based Systems
15 January 2024 in Nature Machine Intelligence
Generative models for molecules based on sequential line notation (for example, the simplified molecular-input line-entry system) or graph representation have attracted an increasing interest in the field of structure-based drug design, but they struggle to capture important three-dimensional (3D) spatial interactions and often produce undesirable molecular structures. To address these challenges, we introduce Lingo3DMol, a pocket-based 3D molecule generation method that combines language models and geometric deep learning technology. A new molecular representation, the fragment-based simplified molecular-input line-entry system with local and global coordinates, was developed to assist the model in learning molecular topologies and atomic spatial positions. Additionally, we trained a separate non-covalent interaction predictor to provide essential binding pattern information for the generative model. Lingo3DMol can efficiently traverse drug-like chemical spaces, preventing the formation of unusual structures. The Directory of Useful Decoys-Enhanced dataset was used for evaluation. Lingo3DMol outperformed state-of-the-art methods in terms of drug likeness, synthetic accessibility, pocket binding mode and molecule generation speed.
1 September 2020 in International Journal of Intelligent Systems
Radiotherapy is an indispensable part of adjuvant therapy for cancer that improves local control, overall survival, and the opportunity for good quality of life. Organ delineation and dose plan design are the key steps in the treatment. Organ delineation controls the area of radiotherapy and dose planning controls its intensity. However, both tasks are time‐consuming, exhausting, and subjective, and automated methods are desirable. Although automated methods have been studied, the previous studies either focus on organ segmentation or dose prediction, without considering them from a holistic perspective. In this paper, we treat organ segmentation and dose prediction as similar tasks, and propose an error correction framework to improve their performance based on the same mechanism. The proposed error correction framework consists of a prediction network and a calibration network. The biggest difference between our framework and previous studies is that the state‐of‐the‐art networks can be used as a prediction network or calibration network, and then the performance can be improved by the error correction mechanism. To evaluate the framework, we conducted a series of experiments on dose prediction and organ segmentation. These experimental results show that the framework is superior to other state‐of‐the‐art methods in both tasks.
4 February 2022 in IEEE Transactions on Medical Imaging
Lesion volume segmentation in medical imaging is an effective tool for assessing lesion/tumor sizes and monitoring changes in growth. Since manually segmentation of lesion volume is not only time-consuming but also requires radiological experience, current practices rely on an imprecise surrogate called response evaluation criteria in solid tumors (RECIST). Although RECIST measurement is coarse compared with voxel-level annotation, it can reflect the lesion’s location, length, and width, resulting in a possibility of segmenting lesion volume directly via RECIST measurement. In this study, a novel weakly-supervised method called RECISTSup is proposed to automatically segment lesion volume via RECIST measurement. Based on RECIST measurement, a new RECIST measurement propagation algorithm is proposed to generate pseudo masks, which are then used to train the segmentation networks. Due to the spatial prior knowledge provided by RECIST measurement, two new losses are also designed to make full use of it. In addition, the automatically segmented lesion results are used to supervise the model training iteratively for further improving segmentation performance. A series of experiments are carried out on three datasets to evaluate the proposed method, including ablation experiments, comparison of various methods, annotation cost analyses, visualization of results. Experimental results show that the proposed RECISTSup achieves the state-of-the-art result compared with other weakly-supervised methods. The results also demonstrate that RECIST measurement can produce similar performance to voxel-level annotation while significantly saving the annotation cost.
12 May 2021 in International Journal of Intelligent Systems
Computed tomography (CT) scanning is a fast and painless procedure that can capture clear imaging information beneath the abdomen and is widely used to help diagnose and monitor disease progress. The pelvic lymph node is a key indicator of colorectal cancer metastasis. In the traditional process, an experienced radiologist must read all the CT scanning images slice by slice to track the lymph nodes for future diagnosis. However, this process is time‐consuming, exhausting, and subjective due to the complex pelvic structure, numerous blood vessels, and small lymph nodes. Therefore, automated methods are desirable to make this process easier. Currently, the available open‐source CTLNDataset only contains large lymph nodes. Consequently, a new data set called PLNDataset, which is dedicated to lymph nodes within the pelvis, is constructed to solve this issue. A two‐level annotation calibration method is proposed to guarantee the quality and correctness of pelvic lymph node annotation. Moreover, a novel system composed of a keyframe localization network and a lymph node detection network is proposed to detect pelvic lymph nodes in CT scanning images. The proposed method makes full use of two kinds of prior knowledge: spatial prior knowledge for keyframe localization and anchor prior knowledge for lymph node detection. A series of experiments are carried out to evaluate the proposed method, including ablation experiments, comparing other state‐of‐the‐art methods, and visualization of results. The experimental results demonstrate that our proposed method outperforms other methods on PLNDataset and CTLNDataset. This system is expected to be applied in future clinical practice.
13 June 2023 in IEEE Transactions on Artificial Intelligence
Survival models based on whole slide images (WSIs) are widely used in precision medicine to treat cancer patients better. Most previous studies attempted to address the challenge of WSIs' gigapixel resolutions to survival models, but they failed in terms of the computational efficiency and interpretability of models. This study proposes a coarse-to-fine survival model called WSISur based on graph neural networks, which not only solves the above two problems but also achieves the best survival prediction performance. To solve the issue of computational efficiency, coarse WSI graphs are first constructed on low-resolution images in WSIs, and then fine WSI graphs are built with high-resolution images on the basis of coarse WSI graphs. Subsequent survival analysis is performed on the constructed WSI graphs. To solve the issue of interpretability of the model, WSIs' regions most relevant to patients' lifetimes are identified by gradient-weighted class activation mapping. Nevertheless, due to the imbalance of data labels, there is a problem of probability failure in end-to-end mini-batch training. To this end, survival event sampling strategy is proposed to balance data labels. In addition, a series of experiments are carried out on multi-center datasets to evaluate the proposed model. Experimental results show that WSISur achieves the state-of-the-art result compared with other existing related methods, and has stronger domain invariance. Results also demonstrate the interpretability of WSISur.