RESUMEN
With the rapid development of image acquisition and storage, multiple images per class are commonly available for computer vision tasks (e.g., face recognition, object detection, medical imaging, etc.). Recently, the recurrent neural network (RNN) has been widely integrated with convolutional neural networks (CNN) to perform image classification on ordered (sequential) data. In this paper, by permutating multiple images as multiple dummy orders, we generalize the ordered "RNN+CNN" design (longitudinal) to a novel unordered fashion, called Multi-path x-D Recurrent Neural Network (MxDRNN) for image classification. To the best of our knowledge, few (if any) existing studies have deployed the RNN framework to unordered intra-class images to leverage classification performance. Specifically, multiple learning paths are introduced in the MxDRNN to extract discriminative features by permutating input dummy orders. Eight datasets from five different fields (MNIST, 3D-MNIST, CIFAR, VGGFace2, and lung screening computed tomography) are included to evaluate the performance of our method. The proposed MxDRNN improves the baseline performance by a large margin across the different application fields (e.g., accuracy from 46.40% to 76.54% in VGGFace2 test pose set, AUC from 0.7418 to 0.8162 in NLST lung dataset). Additionally, empirical experiments show the MxDRNN is more robust to category-irrelevant attributes (e.g., expression, pose in face images), which may introduce difficulties for image classification and algorithm generalizability. The code is publicly available.
RESUMEN
OBJECTIVE: Lung cancer is the leading cancer killer in women, resulting in more deaths than breast, cervical and ovarian cancer combined. Screening for lung cancer has been shown to significantly reduce mortality, with some evidence that women may have a greater benefit. This study demonstrates that a population of women being screened for breast cancer may greatly benefit from screening for lung cancer. METHODS: Data from 18,040 women who were screened for breast cancer in 2015 at two imaging facilities that also performed lung screening were reviewed. A natural language-processing algorithm followed by a manual chart review identified women eligible for lung cancer screening by U.S. Preventive Services Task Force (USPSTF) criteria. A chart review of these eligible women was performed to determine subsequent enrollment in a lung screening program (2016-2019), current screening eligibility, cancer diagnoses and cancer-related outcomes. RESULTS: Natural language processing identified 685 women undergoing screening mammography who were also potentially eligible for lung screening based on age and smoking history. Manual chart review confirmed 251 were eligible under USPSTF criteria. By June 2019, 63 (25%) had enrolled in lung screening, of which three were diagnosed with screening-detected lung cancer resulting in zero deaths. Of 188 not screened, seven were diagnosed with lung cancer resulting in five deaths by study end. Four women received a diagnosis of breast cancer with no deaths. CONCLUSION: Women screened for breast cancer are dying from lung cancer. We must capitalize on reducing barriers to improve screening for lung cancer among high-risk women.
Asunto(s)
Neoplasias de la Mama , Neoplasias Pulmonares , Neoplasias de la Mama/diagnóstico , Detección Precoz del Cáncer , Femenino , Humanos , Neoplasias Pulmonares/diagnóstico , Neoplasias Pulmonares/epidemiología , Mamografía , Tamizaje MasivoRESUMEN
PURPOSE: To develop a model to estimate lung cancer risk using lung cancer screening CT and clinical data elements (CDEs) without manual reading efforts. MATERIALS AND METHODS: Two screening cohorts were retrospectively studied: the National Lung Screening Trial (NLST; participants enrolled between August 2002 and April 2004) and the Vanderbilt Lung Screening Program (VLSP; participants enrolled between 2015 and 2018). Fivefold cross-validation using the NLST dataset was used for initial development and assessment of the co-learning model using whole CT scans and CDEs. The VLSP dataset was used for external testing of the developed model. Area under the receiver operating characteristic curve (AUC) and area under the precision-recall curve were used to measure the performance of the model. The developed model was compared with published risk-prediction models that used only CDEs or imaging data alone. The Brock model was also included for comparison by imputing missing values for patients without a dominant pulmonary nodule. RESULTS: A total of 23 505 patients from the NLST (mean age, 62 years ± 5 [standard deviation]; 13 838 men, 9667 women) and 147 patients from the VLSP (mean age, 65 years ± 5; 82 men, 65 women) were included. Using cross-validation on the NLST dataset, the AUC of the proposed co-learning model (AUC, 0.88) was higher than the published models predicted with CDEs only (AUC, 0.69; P < .05) and with images only (AUC, 0.86; P < .05). Additionally, using the external VLSP test dataset, the co-learning model had a higher performance than each of the published individual models (AUC, 0.91 [co-learning] vs 0.59 [CDE-only] and 0.88 [image-only]; P < .05 for both comparisons). CONCLUSION: The proposed co-learning predictive model combining chest CT images and CDEs had a higher performance for lung cancer risk prediction than models that contained only CDE or only image data; the proposed model also had a higher performance than the Brock model.Keywords: Computer-aided Diagnosis (CAD), CT, Lung, Thorax Supplemental material is available for this article. © RSNA, 2021.
RESUMEN
PURPOSE: The aim of this study was to identify predictors of appropriate follow-up for clinically significant incidental findings (IFs) detected with low-dose CT during lung cancer screening. METHODS: Charts of 1,458 prospectively enrolled lung screening patients from January 1, 2015, to October 31, 2018, were reviewed. IFs, other than coronary artery calcification and emphysema, were identified. ACR practice guidelines defined appropriate patient follow-up. Patient demographic and social characteristics were obtained from the initial shared decision-making visit and the electronic medical record. Factors of interest included age, gender, race, education level, and insurance status. Education level was reported as high school graduate or less or education past high school. A multivariate logistic regression was estimated to assess patient factors associated with appropriate follow-up. RESULTS: One hundred thirty-eight participants (9%) with 141 actionable IFs were identified. The overall appropriate follow-up rate was 82%. The most common IFs were renal lesions (16%), dilated thoracic aorta (10%), and pulmonary fibrosis (10%). Univariate analysis of appropriate patient follow-up revealed a significant difference for education level (P = .02). A greater than high school education remained strongly associated with appropriate follow-up after controlling for other demographic factors. CONCLUSIONS: Appropriate patient follow-up of clinically significant IFs from lung cancer screening is a well-recognized avenue to improve population health. Education level is a significant independent predictor of appropriate follow-up of IFs, whether as a surrogate for low socioeconomic status or as an indication of health literacy. To address these realities, lung screening shared decision making should adapt to consider health care access and health literacy.
Asunto(s)
Detección Precoz del Cáncer , Neoplasias Pulmonares , Escolaridad , Estudios de Seguimiento , Humanos , Hallazgos Incidentales , Neoplasias Pulmonares/diagnóstico por imagen , Neoplasias Pulmonares/epidemiologíaRESUMEN
Annual low dose computed tomography (CT) lung screening is currently advised for individuals at high risk of lung cancer (e.g., heavy smokers between 55 and 80 years old). The recommended screening practice significantly reduces all-cause mortality, but the vast majority of screening results are negative for cancer. If patients at very low risk could be identified based on individualized, image-based biomarkers, the health care resources could be more efficiently allocated to higher risk patients and reduce overall exposure to ionizing radiation. In this work, we propose a multi-task (diagnosis and prognosis) deep convolutional neural network to improve the diagnostic accuracy over a baseline model while simultaneously estimating a personalized cancer-free progression time (CFPT). A novel Censored Regression Loss (CRL) is proposed to perform weakly supervised regression so that even single negative screening scans can provide small incremental value. Herein, we study 2287 scans from 1433 de-identified patients from the Vanderbilt Lung Screening Program (VLSP) and Molecular Characterization Laboratories (MCL) cohorts. Using five-fold cross-validation, we train a 3D attention-based network under two scenarios: (1) single-task learning with only classification, and (2) multi-task learning with both classification and regression. The single-task learning leads to a higher AUC compared with the Kaggle challenge winner pre-trained model (0.878 v. 0.856), and multi-task learning significantly improves the single-task one (AUC 0.895, p<0.01, McNemar test). In summary, the image-based predicted CFPT can be used in follow-up year lung cancer prediction and data assessment.
RESUMEN
BACKGROUND: Detection of early-stage lung cancer improves during subsequent rounds of screening with low-dose CT and potentially leads to saving lives with curative treatment. Therefore, adherence to annual lung screening is important. We hypothesized that adherence to annual screening would increase after hiring of a dedicated program coordinator. METHODS: We performed a mixed-methods study in a retrospective cohort of patients who underwent lung screening at our academic institution between January 1, 2014, and March 31, 2018. Patients with baseline lung screening examinations performed between January 1, 2014, and September 30, 2016, with Lung CT Screening Reporting & Data System 1 or 2 scores and a 12-month follow-up recommendation were included. We tracked patient adherence to annual follow-up lung screening over time (before and after hiring of a program coordinator) and conducted a cross-sectional survey of patients nonadherent to annual follow-up to elicit quantitative and qualitative feedback. RESULTS: Of the 319 patients who completed baseline lung screening with normal results, 189 (59%) were adherent to annual follow-up recommendations and 130 (41%) were nonadherent. Patient adherence varied over time: 21.7% adherence (10 of 46) before hiring a program coordinator and 65.6% adherence (179 of 273) after the program coordinator's hire date. Patients reported the following reasons for nonadherence to annual lung screening: lack of transportation, financial cost, lack of communication by physicians, and lack of current symptoms. CONCLUSIONS: Adherence to annual lung screening after normal baseline studies increased significantly over time. Hiring a full-time program coordinator was likely associated with this increased in adherence.
Asunto(s)
Neoplasias Pulmonares , Estudios Transversales , Detección Precoz del Cáncer , Humanos , Pulmón , Neoplasias Pulmonares/diagnóstico por imagen , Tamizaje Masivo , Estudios Retrospectivos , Tomografía Computarizada por Rayos XRESUMEN
The Long Short-Term Memory (LSTM) network is widely used in modeling sequential observations in fields ranging from natural language processing to medical imaging. The LSTM has shown promise for interpreting computed tomography (CT) in lung screening protocols. Yet, traditional image-based LSTM models ignore interval differences, while recently proposed interval-modeled LSTM variants are limited in their ability to interpret temporal proximity. Meanwhile, clinical imaging acquisition may be irregularly sampled, and such sampling patterns may be commingled with clinical usages. In this paper, we propose the Distanced LSTM (DLSTM) by introducing time-distanced (i.e., time distance to the last scan) gates with a temporal emphasis model (TEM) targeting at lung cancer diagnosis (i.e., evaluating the malignancy of pulmonary nodules). Briefly, (1) the time distance of every scan to the last scan is modeled explicitly, (2) time-distanced input and forget gates in DLSTM are introduced across regular and irregular sampling sequences, and (3) the newer scan in serial data is emphasized by the TEM. The DLSTM algorithm is evaluated with both simulated data and real CT images (from 1794 National Lung Screening Trial (NLST) patients with longitudinal scans and 1420 clinical studied patients). Experimental results on simulated data indicate the DLSTM can capture families of temporal relationships that cannot be detected with traditional LSTM. Cross-validation on empirical CT datasets demonstrates that DLSTM achieves leading performance on both regularly and irregularly sampled data (e.g., improving LSTM from 0.6785 to 0.7085 on F1 score in NLST). In external-validation on irregularly acquired data, the benchmarks achieved 0.8350 (CNN feature) and 0.8380 (with LSTM) on AUC score, while the proposed DLSTM achieves 0.8905. In conclusion, the DLSTM approach is shown to be compatible with families of linear, quadratic, exponential, and log-exponential temporal models. The DLSTM can be readily extended with other temporal dependence interactions while hardly increasing overall model complexity.
Asunto(s)
Memoria a Corto Plazo , Redes Neurales de la Computación , Algoritmos , Humanos , Procesamiento de Lenguaje Natural , Tomografía Computarizada por Rayos XRESUMEN
The field of lung nodule detection and cancer prediction has been rapidly developing with the support of large public data archives. Previous studies have largely focused cross-sectional (single) CT data. Herein, we consider longitudinal data. The Long Short-Term Memory (LSTM) model addresses learning with regularly spaced time points (i.e., equal temporal intervals). However, clinical imaging follows patient needs with often heterogeneous, irregular acquisitions. To model both regular and irregular longitudinal samples, we generalize the LSTM model with the Distanced LSTM (DLSTM) for temporally varied acquisitions. The DLSTM includes a Temporal Emphasis Model (TEM) that enables learning across regularly and irregularly sampled intervals. Briefly, (1) the temporal intervals between longitudinal scans are modeled explicitly, (2) temporally adjustable forget and input gates are introduced for irregular temporal sampling; and (3) the latest longitudinal scan has an additional emphasis term. We evaluate the DLSTM framework in three datasets including simulated data, 1794 National Lung Screening Trial (NLST) scans, and 1420 clinically acquired data with heterogeneous and irregular temporal accession. The experiments on the first two datasets demonstrate that our method achieves competitive performance on both simulated and regularly sampled datasets (e.g. improve LSTM from 0.6785 to 0.7085 on F1 score in NLST). In external validation of clinically and irregularly acquired data, the benchmarks achieved 0.8350 (CNN feature) and 0.8380 (LSTM) on area under the ROC curve (AUC) score, while the proposed DLSTM achieves 0.8905.