Abstract:Recently, intelligent analysis of lung nodules with the assistant of computer aided detection (CAD) techniques can improve the accuracy rate of lung cancer diagnosis. However, existing CAD systems and pulmonary datasets mainly focus on Computed Tomography (CT) images from one single period, while ignoring the cross spatio-temporal features associated with the progression of nodules contained in imaging data from various captured periods of lung cancer. If the evolution patterns of nodules across various periods in the patients' CT sequences can be explored, it will play a crucial role in guiding the precise screening identification of lung cancer. Therefore, a cross spatio-temporal lung nodule dataset with pathological information for nodule identification and diagnosis is constructed, which contains 328 CT sequences and 362 annotated nodules from 109 patients. This comprehensive database is intended to drive research in the field of CAD towards more practical and robust methods, and also contribute to the further exploration of precision medicine related field. To ensure patient confidentiality, we have removed sensitive information from the dataset.

What problem does this paper attempt to address?

The paper primarily aims to address the following issues: ### Research Background and Objectives - **Challenges in Lung Cancer Diagnosis**: There are numerous challenges in current lung cancer screening and diagnosis, including the time-consuming and subjective nature of manual analysis of large volumes of CT images by doctors, and the complexity and subtle differences in the morphological characteristics of malignant lung nodules that make them difficult for the human eye to recognize. - **Limitations of Computer-Aided Detection (CAD) Systems**: Existing CAD systems and lung datasets mostly focus on CT images at a single time point, neglecting the spatiotemporal features brought by the development and changes of lung nodules in images taken at different time points. ### Problems Addressed - **Construction of a Spatiotemporal Lung Nodule Dataset**: The paper aims to construct a spatiotemporal lung nodule dataset containing pathological information to support the identification and diagnosis of lung nodules. This dataset includes 328 CT sequences and 362 annotated lung nodules from 109 patients, aiming to advance CAD systems towards more practical and robust methods and promote further exploration in the field of precision medicine. ### Dataset Characteristics - **Spatiotemporal Dimension**: The dataset not only covers data from a single time point but also includes CT scans of patients at different time points, which helps observe the changes and development trends of lung nodules over time. - **Integration of Pathological Information**: All nodules are precisely annotated based on pathological reports, ensuring the authenticity and accuracy of the dataset labels. - **Data Integrity and Diversity**: The dataset covers lung nodules of different sizes and types (benign and malignant), as well as CT scans of different thicknesses (1.25 mm and 5 mm), providing a rich source of samples for research. ### Methods and Contributions - **Data Collection and Annotation**: Data were collected through the hospital's electronic medical record system, pathology information system, and picture archiving and communication system, with expert doctors guiding the annotation of the location and contours of the nodules. - **Dataset Structure**: The dataset is organized by temporal and spatial dimensions, supporting the observation of nodule changes from both time and space perspectives. - **Technical Validation**: The paper evaluates the dataset using various classification and detection networks, demonstrating the effectiveness and applicability of the dataset. In summary, this paper aims to overcome the limitations of existing CAD systems by constructing a comprehensive and high-quality spatiotemporal lung nodule dataset, providing strong support for the accurate diagnosis of lung nodules.

A Cross Spatio-Temporal Pathology-based Lung Nodule Dataset

A Lung Nodule Dataset with Histopathology-based Cancer Type Annotation

Computer-aided diagnosis of lung cancer: the effect of training datasets on classification accuracy of lung nodules.

Computer-Aided Lung Nodule Detection Based on Ct Images

Parse and Recall: Towards Accurate Lung Nodule Malignancy Prediction like Radiologists

Lung nodule pre-diagnosis and insertion path planning for chest CT images

Lung Nodule Detection Based on the Combination of Morphometric and Texture Features

A Survey of Computer-Aided Diagnosis of Lung Nodules from CT Scans Using Deep Learning

Design Computer-Aided Diagnosis System Based on Chest CT Evaluation of Pulmonary Nodules

3D Skeletonization Feature Based Computer-Aided Detection System for Pulmonary Nodules in CT Datasets

Computer-aided diagnosis of lung cancer: the effect of training data sets on classification accuracy of lung nodules

Computer Aided Lung Nodule Detection on CT Data

Computer-aided diagnosis of ground-glass opacity pulmonary nodules using radiomic features analysis.

Autonomous Detection of Solitary Pulmonary Nodules on CT Images for Computer-Aided Diagnosis

Pulmonary Nodule Classification Using a Multiview Residual Selective Kernel Network

Computerized Detection of Lung Nodules Through Radiomics

Research Progress of Computer Aided Diagnosis System for Pulmonary Nodules in CT Images.

3D-Morphomics, Morphological Features on CT scans for lung nodule malignancy diagnosis

An Anthropomorphic Diagnosis System of Pulmonary Nodules using Weak Annotation-Based Deep Learning

Evolving the pulmonary nodules diagnosis from classical approaches to deep learning-aided decision support: three decades' development course and future prospect