Computer Aided Diagnosis Using Multilevel Image Features on Large-Scale Evaluation

Abstract

Abstract. Computer aided diagnosis (CAD) of cancerous anatomical structures via 3D medical images has emerged as an intensively studied research area. In this paper, we present a principled three-tiered image feature learning approach to capture task specific and data-driven class discriminative statistics from an annotated image database. It integrates voxel-, instance-, and database-level feature learning, aggregation and parsing. The initial segmentation is proceeded as robust voxel labeling and thresholding. After instance-level spatial aggregation, extracted features can also be flexibly tuned for classifying lesions, or discriminating different subcategories of lesions. We demonstrate the effectiveness in the lung nodule detection task which handles all types of solid, partial-solid, and ground-glass nodules using the same set of learned features. Our hierarchical feature learning framework, which was extensively trained and validated on large-scale multiple site datasets of 879 CT volumes (510 training and 369 validation), achieves superior performance than other state-of-the-art CAD systems. The proposed method is also shown to be applicable for colonic polyp detection, including all polyp morphological subcategories, via 770 tagged-prep CT scans from multiple medical sites (358 training and 412 validation).

Improving Computer-Aided Detection UsingConvolutional Neural Networks and Random View Aggregation

“…highly tuned CADe systems for colonic polyp detection in CTC, such as [37], [40], [38]. Note that our system achieves significantly higher sensitivities of 95%, 98% at 1 or 3 FP/vol.…”

“…There exist two types of cascaded CADe classification architectures for false positive reduction are two types: 1) extraction of new image features followed by retraining of a classifier on all candidates [39], [38], [6], [20], [40] (from Sec. II-D) or 2) design of application dependent post-filtering components [41], [42], [43].…”

“…For all imaging data sets used in this study, the image patches were centered at each CADe coordinate (of candidate VOI centroid from pre-existing CADe systems [4], [26], [25], [27]) with 32 × 32 pixels in resolution. All patches were sampled at 4 scales of s = [30,35,40,45] mm ROI edge length in physical image space, after isotropic resampling of the input CT images (see Fig. 4).…”

“…Each three-channel image patch (as a 2.5D view) is centered at a CADe coordinate with 32 × 32 pixels. Again, all patches are sampled at 4 scales: s = [30,35,40,45] mm for the VOI edge length in the physical image space, after isotropic resampling of the CT images (see Fig. 4).…”

Novel image markers for non-small cell lung cancer classification and survival prediction

“…However, none of these methods could robustly handle touching cell segmentation challenges exhibited in lung cancer images. Lu et al [23] has proposed a supervised learning-based segmentation algorithm to support new image features extraction and polyp detection on CT images, and a flexible, hierarchical feature learning framework integrating different levels of discriminative and descriptive information is presented in [24]. Supervised learning is a potential approach to tackle these challenges, but it requires a lot of labeled training data provided by experienced pathologists.…”