Algorithm-assisted diagnosis of Hirschsprung's disease - evaluation of robustness and comparative image analysis on data from various labs and slide scanners.

BACKGROUND: Differences in the preparation, staining and scanning of digital pathology slides create significant pre-analytic variability. Algorithm-assisted tools must be able to contend with this variability in order to be applicable in clinical practice. In a previous study, a decision support algorithm was developed to assist in the diagnosis of Hirschsprung's disease. In the current study, we tested the robustness of this algorithm while assessing for pre-analytic factors which may affect its performance. METHODS: The decision support algorithm was used on digital pathology slides obtained from four different medical centers (A-D) and scanned by three different scanner models (by Philips, Hamamatsu and 3DHISTECH). A total of 192 cases and 1782 slides were used in this study. RGB histograms were constructed to compare images from the various medical centers and scanner models and highlight the differences in color and contrast. RESULTS: The algorithm was able to correctly identify ganglion cells in 99.2% of cases, from all medical centers (All scanned by the Philips slide scanner) as well as 95.5% and 100% of the slides scanned by the 3DHISTECH and Hamamatsu brand slide scanners, respectively. The total error rate for center D was lower than the other medical centers (3.9% vs 7.1%, 10.8% and 6% for centers A-C, respectively), the vast majority of errors being false positives (3.45% vs 0.45% false negatives). The other medical centers showed a higher rate of false negatives in relation to false positives (6.81% vs 0.29%, 9.8% vs 1.2% and 5.37% vs 0.63% for centers A-C, respectively). The total error rates for the Philips, Hamamatsu and 3DHISTECH brand scanners were 3.9%, 3.2% and 9.8%, respectively. RGB histograms demonstrated significant differences in pixel value distribution between the four medical centers, as well as between the 3DHISTECH brand scanner when compared to the Philips and Hamamatsu brand scanners. CONCLUSIONS: The results reported in this paper suggest that the algorithm-based decision support system has sufficient robustness to be applicable for clinical practice. In addition, the novel method used in its development - Hierarchial-Contexual Analysis (HCA) may be applicable to the development of algorithm-assisted tools in other diseases, for which available datasets are limited. Validation of any given algorithm-assisted support system should nonetheless include data from as many medical centers and scanner models as possible.

Perineural invasion detection in pancreatic ductal adenocarcinoma using artificial intelligence.

Perineural invasion (PNI) refers to the presence of cancer cells around or within nerves, raising the risk of residual tumor. Linked to worse prognosis in pancreatic ductal adenocarcinoma (PDAC), PNI is also being explored as a therapeutic target. The purpose of this work was to build a PNI detection algorithm to enhance accuracy and efficiency in identifying PNI in PDAC specimens. Training used 260 manually segmented nerve and tumor HD images from 6 scanned PDAC cases; Analytical performance analysis used 168 additional images; clinical analysis used 59 PDAC cases. The algorithm pinpointed key areas of tumor-nerve proximity for pathologist confirmation. Analytical performance reached sensitivity of 88% and 54%, and specificity of 78% and 85% for the detection of nerve and tumor, respectively. Incorporating tumor-nerve distance in clinical evaluation raised PNI detection from 52 to 81% of all cases. Interestingly, pathologist analysis required an average of only 24 s per case. This time-efficient tool accurately identifies PNI in PDAC, even with a small training cohort, by imitating pathologist thought processes.

Short Training Significantly Improves Ganglion Cell Detection Using an Algorithm-Assisted Approach.

CONTEXT.­: Medical education in pathology relies on the accumulation of experience gained through inspection of numerous samples from each entity. Acquiring sufficient teaching material for rare diseases, such as Hirschsprung disease (HSCR), may be difficult, especially in smaller institutes. The current study makes use of a previously developed decision support system using a decision support algorithm meant to aid pathologists in the diagnosis of HSCR. OBJECTIVE.­: To assess the effect of a short training session on algorithm-assisted HSCR diagnosis. DESIGN.­: Five pathologists reviewed a data set of 568 image sets (1704 images in total) selected from 50 cases by the decision support algorithm and were tasked with scoring the images for the presence or absence of ganglion cells. The task was repeated a total of 3 times. Each pathologist had to complete a short educational presentation between the second and third iterations. RESULTS.­: The training resulted in a significantly increased rate of correct diagnoses (true positive/negative) and a decreased need for referrals for expert consultation. No statistically significant changes in the rate of false positives/negatives were detected. CONCLUSIONS.­: A very short (<10 minutes) training session can greatly improve the pathologist's performance in the algorithm-assisted diagnosis of HSCR. The same approach may be feasible in training for the diagnosis of other rare diseases.

Automatic ganglion cell detection for improving the efficiency and accuracy of hirschprung disease diagnosis.

Histopathologic diagnosis of Hirschsprung's disease (HSCR) is time consuming and requires expertise. The use of artificial intelligence (AI) in digital pathology is actively researched and may improve the diagnosis of HSCR. The purpose of this research was to develop an algorithm capable of identifying ganglion cells in digital pathology slides and implement it as an assisting tool for the pathologist in the diagnosis of HSCR. Ninety five digital pathology slides were used for the construction and training of the algorithm. Fifty cases suspected for HSCR (727 slides) were used as a validation cohort. Image sets suspected to contain ganglion cells were chosen by the algorithm and then reviewed and scored by five pathologists, one HSCR expert and 4 non-experts. The algorithm was able to identify ganglion cells with 96% sensitivity and 99% specificity (in normal colon) as well as to correctly identify a case previously misdiagnosed as non-HSCR. The expert was able to achieve perfectly accurate diagnoses based solely on the images suggested by the algorithm, with over 95% time saved. Non-experts would require expert consultation in 20-58% of the cases to achieve similar results. The use of AI in the diagnosis of HSCR can greatly reduce the time and effort required for diagnosis and improve accuracy.

Dynamic Spatial Predicted Background.

We present a novel method for online background modeling for static video cameras - Dynamic Spatial Predicted Background (DSPB). Our unique method employs a small subset of image pixels to predict the whole scene by exploiting pixel correlations (distant and close). DSPB acts as a hybrid model combining successful elements taken from two major approaches: local-adaptive that propose to fit a distribution pixelwise, and global-linear that reconstruct the background by finding a lowrank version of the scene. To our knowledge, this is the first attempt to combine these approaches in a unified system. DSPB models the scene as a superposition of illumination effects and predicts each pixel's value by a linear estimator comprised of only 5 pixels of the scene and can initialize the background starting from the 5th frame. By doing so, we keep the computational load low, allowing our method to be used in many real-time applications using simple hardware. The suggested prediction model of scene appearance is novel, and the scheme is very accurate and efficient computationally. We show the method merits on an application for video FG-BG separation, and how some of the main existing approaches may be challenged and how their drawbacks are less dominant in our model. Experimental results validate our findings, by computation speed and mean F-measure values on several public datasets. We also examine how results may improve by analyzing each video individually according to its content. DSPB can be successfully incorporated in other image processing tasks like change detection, video compression and video inpainting.

Geometry and Radiometry Invariant Matched Manifold Detection.

Consider a set of deformable objects undergoing geometric and radiometric transformations. As a result of the action of these transformations, the set of different realizations of each object is generally a manifold in the space of observations. Assuming the geometric deformations an object undergoes, belong to some finite dimensional family, it has been shown that the universal manifold embedding (UME) provides a set of nonlinear operators that universally maps each of the different manifolds, where each manifold is generated by the set all of possible appearances of a single object, into a distinct linear subspace of an Euclidean space. In this paper, we generalize this framework to the case where the observed object undergoes both an affine geometric transformation, and a monotonic radiometric transformation, and present a novel framework for the detection and recognition of the deformable objects. Applying to each of the observations an operator that makes it invariant to monotonic amplitude transformations, but is geometry-covariant with the affine transformation, the set of all possible observations on that object is mapped by the UME into a single linear subspace-invariant with respect to both the geometric and radiometric transformations. The embedding of the space of observations is independent of the specific observed object; hence it is universal. The invariant representation of the object is the basis of a matched manifold detection and tracking framework of objects that undergo complex geometric and radiometric deformations: the observed surface is tessellated into a set of tiles such that the deformation of each one is well approximated by an affine geometric transformation and a monotonic transformation of the measured intensities. Since each tile is mapped by the radiometry invariant UME to a distinct linear subspace, the detection and tracking problems are solved by evaluating distances between linear subspaces. Classification in this context becomes a problem of determining which labeled subspace in a Grassmannian is closest to a subspace in the same Grassmannian, where the latter has been generated by radiometry invariant UME from an unlabeled observation.