Lung cancer has brought tough challenges to human health due to its high incidence and mortality rate in the current practice. Nowadays, computed tomography (CT) imaging is still the most preferred diagnostic tool for early screening of lung cancer. However, a great challenge brought from accumulative CT imaging data can not meet the demand of the current clinical practice. As a novel kind of artificial intelligence technique aimed to deal with medical images, a computer-aided diagnosis has been found to provide useful auxiliary information, attenuate the workload of doctors, and significantly improve the efficiency and accuracy for clinical diagnosis of lung cancer. Therefore, an effective combination of computer-aided techniques and CT imaging has increasingly become an active area of investigation in early diagnosis of lung cancer. This review aims to summarize the latest progress on the diagnostic value of computer-aided technology with regard to early stage lung cancer from the perspectives of machine learning and deep learning.
Objective To develop an innovative recognition algorithm that aids physicians in the identification of pulmonary nodules. MethodsPatients with pulmonary nodules who underwent thoracoscopic surgery at the Department of Thoracic Surgery, Affiliated Drum Tower Hospital of Nanjing University Medical School in December 2023, were enrolled in the study. Chest surface exploration data were collected at a rate of 60 frames per second and a resolution of 1 920×1 080. Frame images were saved at regular intervals for subsequent block processing. An algorithm database for lung nodule recognition was developed using the collected data. ResultsA total of 16 patients were enrolled, including 9 males and 7 females, with an average age of (54.9±14.9) years. In the optimized multi-topology convolutional network model, the test results demonstrated an accuracy rate of 94.39% for recognition tasks. Furthermore, the integration of micro-variation amplification technology into the convolutional network model enhanced the accuracy of lung nodule identification to 96.90%. A comprehensive evaluation of the performance of these two models yielded an overall recognition accuracy of 95.59%. Based on these findings, we conclude that the proposed network model is well-suited for the task of lung nodule recognition, with the convolutional network incorporating micro-variation amplification technology exhibiting superior accuracy. Conclusion Compared to traditional methods, our proposed technique significantly enhances the accuracy of lung nodule identification and localization, aiding surgeons in locating lung nodules during thoracoscopic surgery.
Surgical resection is the only radical method for the treatment of early-stage non-small cell lung cancer. Intraoperative frozen section (FS) has the advantages of high accuracy, wide applicability, few complications and real-time diagnosis of pulmonary nodules. It is one of the main means to guide surgical strategies for pulmonary nodules. Therefore, we searched PubMed, Web of Science, CNKI, Wanfang and other databases for nearly 30 years of relevant literature and research data, held 3 conferences, and formulated this consensus by using the Delphi method. A total of 6 consensus contents were proposed: (1) Rapid intraoperative FS diagnosis of benign and malignant diseases; (2) Diagnosis of lung cancer types including adenocarcinoma, squamous cell carcinoma, others, etc; (3) Diagnosis of lung adenocarcinoma infiltration degree; (4) Histological subtype diagnosis of invasive adenocarcinoma; (5) The treatment strategy of lung adenocarcinoma with inconsistent diagnosis on degree of invasion between intraoperative FS and postoperative paraffin diagnosis; (6) Intraoperative FS diagnosis of tumor spread through air space, visceral pleural invasion and lymphovascular invasion. Finally, we gave 11 recommendations in the above 6 consensus contents to provide a reference for diagnosis of pulmonary nodules and guiding surgical decision-making for peripheral non-small cell lung cancer using FS, and to further improve the level of individualized and precise diagnosis and treatment of early-stage lung cancer.
ObjectiveTo reveal and demonstrate the hotspots and further research directions in screening technology for early lung cancer, and provide references for the future studies. MethodsResearches related to lung cancer screening from 2011 to 2021 in the Web of Science database were included. Biblioshiny, a bibliometrics program based on R language, was used to perform content analysis and visualization of the included literature information. ResultsResearches related to lung cancer screening were increasing year by year. Six major cooperation groups were formed between countries. The current research hotspots in the field of early lung cancer screening technology mainly focused on the multi-directional fusion of radiographic imaging, liquid biopsy and artificial intelligence. ConclusionLow-dose spiral CT screening is still the most important and mainstream method for the screening of early lung cancer at present. The combination and integration of artificial intelligence with various screening methods and the innovation of novel testing and diagnostic equipment are the current research hotspots and the future research trend in this field.
Objective To investigate the diagnostic value of tumor marker combining the probability of malignancy model in pulmonary nodules. Methods A total of 117 patients with pulmonary nodules diagnosed between January 2013 and January 2016 were retrospectively analyzed. Seventy-six cases of the patients diagnosed with cancer were selected as a lung cancer group. Forty-one cases of the patients diagnosed with benign lesions were selected as a benign group. Tumor markers were detected and the probability of malignancy were calculated. Results The positive rate of carcinoembryonic antigen (CEA), cancer antigen 125 (CA125), neuron-specific enolase (NSE), cytokeratin marker (CYFRA21-1), and the probability of malignancy in the lung caner group were significantly higher than those of the benign group. The sensitivity, specificity, and accuracy of CEA, CA125, NSE, CYFRA21-1 combined detection were 72.37%, 73.17%, and 72.65%, respectively. Using the probability of malignancy model to calculate each pulmonary nodules, the area under ROC curve was 0.743 which was higher than 0.7; and 28.5% was selected as cut-off value based on clinical practice and ROC curve. The sensitivity, specificity, and accuracy of the probability of malignancy model were 63.16%, 78.05%, and 68.68%, respectively. The sensitivity, specificity, and accuracy of tumor marker combining the probability of malignancy model were 93.42%, 68.29%, and 92.31%, respectively. The sensitivity and accuracy of tumor marker combining the probability of malignancy model were significantly improved compared with tumor markers or the probability of malignancy model single detection (P<0.01). Conclusion The tumor marker combining the probability of malignancy model can improve the sensitivity and accuracy in diagnosis of pulmonary nodules.
Objective The management of pulmonary nodules is a common clinical problem, and this study constructed a nomogram model based on FUT7 methylation combined with CT imaging features to predict the risk of adenocarcinoma in patients with pulmonary nodules. Methods The clinical data of 219 patients with pulmonary nodules diagnosed by histopathology at the First Affiliated Hospital of Zhengzhou University from 2021 to 2022 were retrospectively analyzed. The FUT7 methylation level in peripheral blood were detected, and the patients were randomly divided into training set (n=154) and validation set (n=65) according to proportion of 7:3. They were divided into a lung adenocarcinoma group and a benign nodule group according to pathological results. Single-factor analysis and multi-factor logistic regression analysis were used to construct a prediction model in the training set and verified in the validation set. The receiver operating characteristic (ROC) curve was used to evaluate the discrimination of the model, the calibration curve was used to evaluate the consistency of the model, and the clinical decision curve analysis (DCA) was used to evaluate the clinical application value of the model. The applicability of the model was further evaluated in the subgroup of high-risk CT signs (located in the upper lobe, vascular sign, and pleural sign). Results Multivariate logistic regression analysis showed that female, age, FUT7_CpG_4, FUT7_CpG_6, sub-solid nodules, lobular sign and burr sign were independent risk factors for lung adenocarcinoma (P<0.05). A column-line graph prediction model was constructed based on the results of the multifactorial analysis, and the area under the ROC curve was 0.925 (95%CI 0.877 - 0.972 ), and the maximum approximate entry index corresponded to a critical value of 0.562, at which time the sensitivity was 89.25%, the specificity was 86.89%, the positive predictive value was 91.21%, and the negative predictive value was 84.13%. The calibration plot predicted the risk of adenocarcinoma of pulmonary nodules was highly consistent with the risk of actual occurrence. The DCA curve showed a good clinical net benefit value when the threshold probability of the model was 0.02 - 0.80, which showed a good clinical net benefit value. In the upper lobe, vascular sign and pleural sign groups, the area under the ROC curve was 0.903 (95%CI 0.847 - 0.959), 0.897 (95%CI 0.848 - 0.945), and 0.894 (95%CI 0.831 - 0.956). Conclusions This study developed a nomogram model to predict the risk of lung adenocarcinoma in patients with pulmonary nodules. The nomogram has high predictive performance and clinical application value, and can provide a theoretical basis for the diagnosis and subsequent clinical management of pulmonary nodules.
Lung cancer is a disease with high incidence rate and high mortality rate worldwide. Its diagnosis and treatment mode is developing constantly. Among them, multi-disciplinary team (MDT) can provide more accurate diagnosis and more individualized treatment, which can not only benefit more early patients, but also prolong the survival time of late patients. However, MDT diagnosis and treatment mode is only carried out more in provincial and municipal tertiary hospitals and has not been popularized. This article intends to introduce MDT mode and its advantages, hoping that MDT mode can be popularized and applied.
The widespread application of low-dose computed tomography (LDCT) has significantly increased the detection of pulmonary small nodules, while accurate prediction of their growth patterns is crucial to avoid overdiagnosis or underdiagnosis. This article reviews recent research advances in predicting pulmonary nodule growth based on CT imaging, with a focus on summarizing key factors influencing nodule growth, such as baseline morphological parameters, dynamic indicators, and clinical characteristics, traditional prediction models (exponential and Gompertzian models), and the applications and limitations of radiomics-based and deep learning models. Although existing studies have achieved certain progress in predicting nodule growth, challenges such as small sample sizes and lack of external validation persist. Future research should prioritize the development of personalized and visualized prediction models integrated with larger-scale datasets to enhance predictive accuracy and clinical applicability.
Lung cancer, as one of the malignant tumors with the fastest increasing morbidity and mortality in the world, has a serious impact on people's health. With the continuous advancement of medical technology, more and more medical methods are applied to lung cancer screening, which has gradually increased the detection rate of early lung cancer. At present, the standard operation for the treatment of early non-small cell lung cancer (NSCLC) is still lobectomy and mediastinal lymph node dissection. There is a growing trend to use segmentectomy for the treatment of early stage lung cancer. Anatomical segmentectomy not only removes the lesions to the maximum extent, but also preserves the lung function to the greatest extent, and its advantages are also obvious. This article reviews the progress of anatomical segmentectomy in the treatment of early NSCLC.
Objective To explore the efficacy of a novel detection technique of circulating tumor cells (CTCs) to identify benign and malignant lung nodules. Methods Nanomagnetic CTC detection based on polypeptide with epithelial cell adhesion molecule (EpCAM)-specific recognition was performed on enrolled patients with pulmonary nodules. There were 73 patients including 48 patients with malignant lesions as a malignant group and 25 patients with benign lesion as a benign group. There were 13 males and 35 females at age of 57.0±11.9 years in the malignant group and 11 males and 14 females at age of 53.1±13.2 years in the benign group. e calculated the differential diagnostic efficacy of CTC count, and conducted subgroup analysis according to the consolidation-tumor ratio, while compared with PET/CT on the efficacy. Results CTC count of the malignant group was significantly higher than that of the benign group (0.50/ml vs. 0.00/ml, P<0.05). Subgroup analysis according to consolidation tumor ratio (CTR) revealed that the difference was statistically significant in pure ground glass (pGGO) nodules 1.00/mlvs. 0.00/ml, P<0.05), but not in part-solid or pure solid nodules. For pGGO nodules, the area under the receiver operating characteristic (ROC) curve of CTC count was 0.833, which was significantly higher than that of maximum of standardized uptake value (SUVmax) (P<0.001). Its sensitivity and specificity was 80.0% and 83.3%, respectively. Conclusion The peptide-based nanomagnetic CTC detection system can differentiate malignant tumor and benign lesions in pulmonary nodules presented as pGGO. It is of great clinical potential as a noninvasive, nonradiating method to identify malignancies in pulmonary nodules.