The epidermal growth factor receptor (EGFR) mutation status play a key role in clinical decision support and prognosis for non-small-cell lung cancer (NSCLC). In this study, we present an automatic end-to-end pipeline to classify the EGFR mutation status according to the features extracted from medical images via deep learning. We tried to solve this problem with three steps: (I) locating tumor candidates via a 3D convolutional neural network (CNN), (II) extracting features via pre-trained lower convolutional layers (layers before the fully connected layers) of VGG16 network, (III) classifying EGFR mutation status according to the extracted features with a logistic regression model. In the experiments, the dataset contains 83 Chest CT series collecting from patients with non-small-cell lung cancer, half of whom are positive for a mutation in EGFR. The whole dataset was divided into two splits for training and testing with 66 CT series and 17 CT series respectively. Our pipeline achieves AUC of 0.725 (±0.009) when running a five-fold cross validation on training dataset and AUC of 0.75 on testing dataset, which validates the efficacy and generalizability of our approach and shows potential usage of non-invasive medical image analysis in detecting EGFR mutation status.
Radiomics analysis has been shown to have considerable potential power for treatment assessment, cancer genetics analysis and clinical decision support. A broad set of quantitative features extracted from medical images is expected to build a descriptive and predictive model, which relating the image features to phenotypes or gene-protein signatures. As a common wrapper strategy, Backward Feature Elimination (BFE) algorithm is widely used to reduce the dimensionality of feature space. In this paper, we propose an effective BFE algorithm utilizing Random Forest (RF) to automatically select the optimal feature subset and try to predict the EGFR mutations using CT images. Firstly, the whole dataset was shuffled and the features were ranked by RF importance measures. Then, LASSO regression was iteratively used to perform both regularization and accuracy calculation in the BFE, ending when any further removals do not result in an improvement, to gain a series of feature subsets. Lastly, we gathered all the feature subsets in a feature counter and final feature subset was determined by hard voting with equal weight. The dataset consists of 130 CT image series with EGFR-mutated lung adenocarcinoma harboring Ex19 (n=56) and Ex21 (n=74) and more than 2000 radiomic features were extracted in each series. Seven features were selected as the set to predict EGFR mutation and all of the features were from Wavelet and Gabor filtered image. It reached best classification result (AUC 0.74, 95% CI, 0.67-0.84) on the K-nearest neighbors (KNN) model.
At present, computer aided systems for liver surgery design and risk evaluation are widely used in clinical all over the world. However, most systems are local applications that run on high-performance workstations, and the images have to processed offline. Compared with local applications, a web-based system is accessible anywhere and for a range of regardless of relative processing power or operating system. RayPlus (http://rayplus.life.hust.edu.cn), a B/S platform for medical image processing, was developed to give a jump start on web-based medical image processing. In this paper, we implement a computer aided system for liver surgery planning on the architecture of RayPlus. The system consists of a series of processing to CT images including filtering, segmentation, visualization and analyzing. Each processing is packaged into an executable program and runs on the server side. CT images in DICOM format are processed step by to interactive modeling on browser with zero-installation and server-side computing. The system supports users to semi-automatically segment the liver, intrahepatic vessel and tumor from the pre-processed images. Then, surface and volume models are built to analyze the vessel structure and the relative position between adjacent organs. The results show that the initial implementation meets satisfactorily its first-order objectives and provide an accurate 3D delineation of the liver anatomy. Vessel labeling and resection simulation are planned to add in the future. The system is available on Internet at the link mentioned above and an open username for testing is offered.
Liver segmentation in CT images has been acknowledged as a basic and indispensable part in systems of computer aided liver surgery for operation design and risk evaluation. In this paper, we will introduce and implement a web-based procedure for liver segmentation to help radiologists and surgeons get an accurate result efficiently and expediently. Several clinical datasets are used to evaluate the accessibility and the accuracy. This procedure seems a promising approach for extraction of liver volumetry of various shapes. Moreover, it is possible for user to access the segmentation wherever the Internet is available without any specific machine.
In order to build high quality geometric models for liver containing vascular system, multi-phase CT series used in a computer–aided diagnosis and surgical planning system aims at liver diseases have to be accurately registered. In this paper we model the segmented liver containing vascular system as a complex shape and propose a two-step registration method. Without any tree modeling for vessel this method can carry out a simultaneous registration for both liver tissue and vascular system inside. Firstly a rigid aligning using vessel as feature is applied on the complex shape model while genetic algorithm is used as the optimization method. Secondly we achieve the elastic shape registration by combine the incremental free form deformation (IFFD) with a modified iterative closest point (ICP) algorithm. Inspired by the concept of demons method, we propose to calculate a fastest diffusion vector (FDV) for each control point on the IFFD lattice to replace the points correspondence needed in ICP iterations. Under the iterative framework of the modified ICP, the optimal solution of control points’ displacement in every IFFD level can be obtained efficiently. The method has been quantitatively evaluated on clinical multi-phase CT series.
Liver segmentation is a basic and indispensable function in systems of computer aided liver surgery for volume calculation, operation designing and risk evaluation. Traditional manual segmentation is very time consuming because of the complicated contours of liver and the big amount of images. For increasing the efficiency of the clinical work, in this paper, a fully-automatic method was proposed to segment the liver from multi-phase contrast-enhanced computed tomography (CT) images. As an advanced region growing method, we applied various pre- and post-processing to get better segmentation from the different phases. Fifteen sets of clinical abdomens CT images of five patients were segmented by our algorithm, and the results were acceptable and evaluated by an experienced surgeon. The running-time is about 30 seconds for a single-phase data which includes more than 200 slices.
It is of vital importance that providing detailed and accurate information about hepatic vein (HV) for liver surgery
planning, such as pre-operative planning of living donor liver transplantation (LDLT). Due to the different blood flow
rate of intra-hepatic vascular systems and the restrictions of CT scan, it is common that HV and hepatic portal vein (HPV) are both filled with contrast medium during the scan and in high intensity in the hepatic venous phase images. As a result, the HV segmentation result obtained from the hepatic venous phase images is always contaminated by HPV which makes accurate HV modeling difficult. In this paper, we proposed a method for quick and accurate HV extraction. Based on the topological structure of intra-hepatic vessels, we analyzed the anatomical features of HV and HPV. According to the analysis, three conditions were presented to identify the nodes that connect HV with HPV in the topological structure, and thus to distinguish HV from HPV. The method costs less than one minute to extract HV and provides a correct and detailed HV model even with variations in vessels. Evaluated by two experienced radiologists, the accuracy of the HV model obtained from our method is over 97%. In the following work, we will extend our work to a comprehensive clinical evaluation and apply this method to actual LDLT surgical planning.
Liver tumor, one of the most wide-spread diseases, has a very high mortality in China. To improve success rates of liver
surgeries and life qualities of such patients, we implement an interactive liver surgery planning system based on contrastenhanced
liver CT images. The system consists of five modules: pre-processing, segmentation, modeling, quantitative
analysis and surgery simulation. The Graph Cuts method is utilized to automatically segment the liver based on an
anatomical prior knowledge that liver is the biggest organ and has almost homogeneous gray value. The system supports
users to build patient-specific liver segment and sub-segment models using interactive portal vein branch labeling, and to
perform anatomical resection simulation. It also provides several tools to simulate atypical resection, including resection
plane, sphere and curved surface. To match actual surgery resections well and simulate the process flexibly, we extend
our work to develop a virtual scalpel model and simulate the scalpel movement in the hepatic tissue using multi-plane
continuous resection. In addition, the quantitative analysis module makes it possible to assess the risk of a liver surgery.
The preliminary results show that the system has the potential to offer an accurate 3D delineation of the liver anatomy, as
well as the tumors' location in relation to vessels, and to facilitate liver resection surgeries. Furthermore, we are testing
the system in a full-scale clinical trial.