Dataset Information

Automated AJCC (7th edition) staging of non-small cell lung cancer (NSCLC) using deep convolutional neural network (CNN) and recurrent neural network (RNN).

ABSTRACT:

Purpose

A large chunk of lung cancers are of the type non-small cell lung cancer (NSCLC). Both the treatment planning and patients' prognosis depend greatly on factors like AJCC staging which is an abstraction over TNM staging. Many significant efforts have so far been made towards automated staging of NSCLC, but the groundbreaking application of a deep neural networks (DNNs) is yet to be observed in this domain of study. DNN is capable of achieving higher level of accuracy than the traditional artificial neural networks (ANNs) as it uses deeper layers of convolutional neural network (CNN). The objective of the present study is to propose a simple yet fast CNN model combined with recurrent neural network (RNN) for automated AJCC staging of NSCLC and to compare the outcome with a few standard machine learning algorithms along with a few similar studies.

Methods

The NSCLC radiogenomics collection from the cancer imaging archive (TCIA) dataset was considered for the study. The tumor images were refined and filtered by resizing, enhancing, de-noising, etc. The initial image processing phase was followed by texture based image segmentation. The segmented images were fed into a hybrid feature detection and extraction model which was comprised of two sequential phases: maximally stable extremal regions (MSER) and the speeded up robust features (SURF). After a prolonged experiment, the desired CNN-RNN model was derived and the extracted features were fed into the model.

Results

The proposed CNN-RNN model almost outperformed the other machine learning algorithms under consideration. The accuracy remained steadily higher than the other contemporary studies.

Conclusion

The proposed CNN-RNN model performed commendably during the study. Further studies may be carried out to refine the model and develop an improved auxiliary decision support system for oncologists and radiologists.

SUBMITTER: Moitra D

PROVIDER: S-EPMC6667593 | biostudies-literature | 2019 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Automated AJCC (7th edition) staging of non-small cell lung cancer (NSCLC) using deep convolutional neural network (CNN) and recurrent neural network (RNN).

Moitra Dipanjan D Mandal Rakesh Kr RK

Health information science and systems 20190730 1

<h4>Purpose</h4>A large chunk of lung cancers are of the type non-small cell lung cancer (NSCLC). Both the treatment planning and patients' prognosis depend greatly on factors like AJCC staging which is an abstraction over TNM staging. Many significant efforts have so far been made towards automated staging of NSCLC, but the groundbreaking application of a deep neural networks (DNNs) is yet to be observed in this domain of study. DNN is capable of achieving higher level of accuracy than the trad ...[more]

PMID: 31406570

Similar Datasets

Project description:Importance:Previous studies have shown that the AJCC Cancer Staging Manual, 7th edition (AJCC 7), tumor classification for cutaneous squamous cell carcinoma (CSCC) failed to accurately stratify disease-related outcomes. The recently released 8th edition (AJCC 8) features a revised tumor classification for only head and neck CSCC (HNCSCC). Objective:To compare AJCC 7 and AJCC 8 tumor classifications for HNCSCC and to validate AJCC 8. Design, Setting, and Participants:This was a 10-year retrospective cohort study (2000-2009) at an academic tertiary care center reviewing 680 primary HNCSCC tumors in 459 patients. Main Outcomes and Measures:Primary HNCSCC tumors were reviewed for disease-related outcomes (DROs): local recurrence (LR), nodal metastasis (NM), and disease-specific death (DSD). Tumors were stratified by AJCC 7 and AJCC 8 tumor classification. Distinctiveness (outcome differences between categories), homogeneity (outcome similarity within categories), and monotonicity (outcome worsening with increasing categories) were assessed for both classifications. Results:Most of the 459 patients were white (451 [98.3%]) and male (312 [68.0%]). AJCC 8 high tumor categories (T3/T4) accounted for 121 (17.8%) of total cases but 50 of 71 DROs (70.4%) (22 of 34 of LRs [64.7%], 17 of 24 NMs [70.8%], and 11 of 13 of DSDs [84.6%]). This was a significant improvement over AJCC 7, where only 12 of 71 DROs (16.9%) (4 of 34 LRs [11.8%], 3 of 24 NMs [12.5%], and 5 of 13 DSDs [38.5%]) occurred in T3/T4 categories. However, AJCC 8 T2 and T3 were indistinct, with overlapping 95% CIs for 10-year cumulative incidences of LR, NM, and DSD. The 10-year cumulative incidence of DROs in the 119 AJCC 8 T3 cases were 19.7% (95% CI, 13.0%-29.7%) for LR, 14.1% (95% CI, 9.7%-20.7%) for NM, and 9.3% (95% CI, 6.8%-14.0% for DSD). Conclusions and Relevance:AJCC 8 demonstrates superior homogeneity and monotonicity compared with AJCC 7. It now may be possible for AJCC 8 HNCSCC T2, T3, and T4 cases to be recorded and tracked by tumor registries because they represented a 23.1% subset in this study, which includes nearly all poor outcomes (85.9%). Further work is needed to validate AJCC 8 with population-level data and to compare AJCC 8 performance against alternative tumor classifications.

Project description:ImportanceTo our knowledge, there are no validation studies to date of the prognostic value of the AJCC Cancer Staging Manual, eighth edition (AJCC 8), criteria for eyelid and periocular squamous cell carcinoma.ObjectiveTo determine the association of tumor (T) category in AJCC 8 with local recurrence, nodal metastasis, distant metastasis, and disease-specific survival (DSS) for eyelid and periocular squamous cell carcinoma.Design, setting, and participantsIn this retrospective, single-center cohort study, 109 consecutive patients with eyelid and periocular squamous cell carcinoma treated from January 1999 to April 2018 were included. Patients with secondary involvement of the periocular region were excluded.Main outcomes and measuresLocal recurrence, nodal metastasis, distance metastasis, and DSS.ResultsOf the 109 included patients, 81 (74.3%) were male, and the median (range) age was 66 (40-91) years. At presentation, 43 patients (39.4%) had recurrent tumor, 4 (3.7%) had nodal metastasis, and 1 (0.9%) had distant metastasis. The median (range) follow-up was 23 (1-161) months. During follow-up, 11 patients (10.1%) developed local recurrence, 7 (6.4%) developed nodal metastasis, 2 (1.8%) developed distant metastasis, and 9 (8.3%) died of disease. The 5-year DSS rate was 87.7% (95% CI, 79.5-96.9). Chronic immunosuppression (hazard ratio, 47.24; 95% CI, 7.33-304.30; P < .001) and presentation with recurrent squamous cell carcinoma (hazard ratio, 5.22; 95% CI, 1.12-24.31; P = .04) were associated with local recurrence during follow-up. Of the 11 patients with local recurrence during follow-up, 7 (64%) had perineural invasion. T category was associated with nodal metastasis; clinical stage of T2c or worse at presentation was associated with higher risk of nodal metastasis and death of disease but not with a higher risk of local recurrence. Distant metastasis was associated with nodal metastasis at presentation (hazard ratio, 32.50; 95% CI, 1.97-536.40; P = .02) and during follow-up. A total of 33 patients (30.3%) had different T categories depending on whether disease was staged according to the seventh or eighth edition of the AJCC Cancer Staging Manual. Compared with AJCC 7, AJCC 8 showed a better predictive value in terms of local recurrence (T3, 17% vs 14%; T4, 11% vs 16%) and DSS.Conclusions and relevanceThese findings suggest that T category in AJCC 8 is associated with nodal metastasis and DSS. Immunosuppression and presentation with recurrent disease are associated with increased risk of future local recurrence. Patients with tumors of clinical stage T2c or worse at presentation are at increased risk of nodal metastasis and worse DSS and should undergo surveillance for nodal metastasis. Future studies, ideally prospective in design, could provide greater confidence in these findings.

Project description:Radiation pneumonitis (RP) is one of the major side effects of thoracic radiotherapy. The aim of this study is to build a dose distribution based prediction model, and investigate the correlation of RP incidence and high-order features of dose distribution. A convolution 3D (C3D) neural network was used to construct the prediction model. The C3D network was pre-trained for action recognition. The dose distribution was used as input of the prediction model. With the C3D network, the convolution operation was performed in 3D space. The guided gradient-weighted class activation map (grad-CAM) was utilized to locate the regions of dose distribution which were strongly correlated with grade?2 and grade<2 RP cases, respectively. The features learned by the convolution filters were generated with gradient ascend to understand the deep network. The performance of the C3D prediction model was evaluated by comparing with three multivariate logistic regression (LR) prediction models, which used the dosimetric, normal tissue complication probability (NTCP) or dosiomics factors as input, respectively. All the prediction models were validated using 70 non-small cell lung cancer (NSCLC) patients treated with volumetric modulated arc therapy (VMAT). The area under curve (AUC) of C3D prediction model was 0.842. While the AUC of the three LR models were 0.676, 0.744 and 0.782, respectively. The guided grad-CAM indicated that the low-dose region of contralateral lung and high-dose region of ipsilateral lung were strongly correlated with the grade?2 and grade<2 RP cases, respectively. The features learned by shallow filters were simple and globally consistent, and of monotonous color. The features of deeper filters displayed more complicated pattern, which was hard or impossible to give strict mathematical definition. In conclusion, we built a C3D model for thoracic radiotherapy toxicity prediction. The results demonstrate its performance is superior over the classical LR models. In addition, CNN also offers a new perspective to further understand RP incidence.

Project description:BACKGROUND:The ?-Lactamase (BL) enzyme family is an important class of enzymes that plays a key role in bacterial resistance to antibiotics. As the newly identified number of BL enzymes is increasing daily, it is imperative to develop a computational tool to classify the newly identified BL enzymes into one of its classes. There are two types of classification of BL enzymes: Molecular Classification and Functional Classification. Existing computational methods only address Molecular Classification and the performance of these existing methods is unsatisfactory. RESULTS:We addressed the unsatisfactory performance of the existing methods by implementing a Deep Learning approach called Convolutional Neural Network (CNN). We developed CNN-BLPred, an approach for the classification of BL proteins. The CNN-BLPred uses Gradient Boosted Feature Selection (GBFS) in order to select the ideal feature set for each BL classification. Based on the rigorous benchmarking of CCN-BLPred using both leave-one-out cross-validation and independent test sets, CCN-BLPred performed better than the other existing algorithms. Compared with other architectures of CNN, Recurrent Neural Network, and Random Forest, the simple CNN architecture with only one convolutional layer performs the best. After feature extraction, we were able to remove ~95% of the 10,912 features using Gradient Boosted Trees. During 10-fold cross validation, we increased the accuracy of the classic BL predictions by 7%. We also increased the accuracy of Class A, Class B, Class C, and Class D performance by an average of 25.64%. The independent test results followed a similar trend. CONCLUSIONS:We implemented a deep learning algorithm known as Convolutional Neural Network (CNN) to develop a classifier for BL classification. Combined with feature selection on an exhaustive feature set and using balancing method such as Random Oversampling (ROS), Random Undersampling (RUS) and Synthetic Minority Oversampling Technique (SMOTE), CNN-BLPred performs significantly better than existing algorithms for BL classification.

Dataset Information

Automated AJCC (7th edition) staging of non-small cell lung cancer (NSCLC) using deep convolutional neural network (CNN) and recurrent neural network (RNN).

Purpose

Methods

Results

Conclusion

Publications

Automated AJCC (7th edition) staging of non-small cell lung cancer (NSCLC) using deep convolutional neural network (CNN) and recurrent neural network (RNN).

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets