Machine learning approaches used to analyze auditory evoked responses from the human auditory brainstem: A systematic review.

BACKGROUND: The application of machine learning algorithms for assessing the auditory brainstem response has gained interest over recent years with a considerable number of publications in the literature. In this systematic review, we explore how machine learning has been used to develop algorithms to assess auditory brainstem responses. A clear and comprehensive overview is provided to allow clinicians and researchers to explore the domain and the potential translation to clinical care. METHODS: The systematic review was performed based on PRISMA guidelines. A search was conducted of PubMed, IEEE-Xplore, and Scopus databases focusing on human studies that have used machine learning to assess auditory brainstem responses. The duration of the search was from January 1, 1990, to April 3, 2021. The Covidence systematic review platform (www.covidence.org) was used throughout the process. RESULTS: A total of 5812 studies were found through the database search and 451 duplicates were removed. The title and abstract screening process further reduced the article count to 89 and in the proceeding full-text screening, 34 articles met our full inclusion criteria. CONCLUSION: Three categories of applications were found, namely neurologic diagnosis, hearing threshold estimation, and other (does not relate to neurologic or hearing threshold estimation). Neural networks and support vector machines were the most commonly used machine learning algorithms in all three categories. Only one study had conducted a clinical trial to evaluate the algorithm after development. Challenges remain in the amount of data required to train machine learning models. Suggestions for future research avenues are mentioned with recommended reporting methods for researchers.

Comparison of machine learning models to classify Auditory Brainstem Responses recorded from children with Auditory Processing Disorder.

INTRODUCTION: Auditory brainstem responses (ABRs) offer a unique opportunity to assess the neural integrity of the peripheral auditory nervous system in individuals presenting with listening difficulties. ABRs are typically recorded and analyzed by an audiologist who manually measures the timing and quality of the waveforms. The interpretation of ABRs requires considerable experience and training, and inappropriate interpretation can lead to incorrect judgments about the integrity of the system. Machine learning (ML) techniques may be a suitable approach to automate ABR interpretation and reduce human error. OBJECTIVES: The main objective of this paper was to identify a suitable ML technique to automate the analysis of ABR responses recorded as a part of the electrophysiological testing in the Auditory Processing Disorder clinical test battery. METHODS: ABR responses recorded during routine clinical assessment from 136 children being evaluated for auditory processing difficulties were analyzed using several common ML algorithms: Support Vector Machines (SVM), Random Forests (RF), Decision Trees (DT), Gradient Boosting (GB), Extreme Gradient Boosting (Xgboost), and Neural Networks (NN). A variety of signal feature extraction techniques were used to extract features from the ABR waveforms as inputs to the ML algorithms. Statistical significance testing and confusion matrices were used to identify the most robust model capable of accurately identifying neurological abnormalities present in ABRs. RESULTS: Clinically significant features in the time-frequency representation of the signal were identified. The ML model trained using the Xgboost algorithm was identified as the most robust model with an accuracy of 92% compared to other models. CONCLUSION: The findings of the present study demonstrate that it is possible to develop accurate ML models to automate the process of analyzing ABR waveforms recorded at suprathreshold levels. There is currently no ML-based application to screen children with listening difficulties. Therefore, it is expected that this work will be translated into an evaluation tool that can be used by audiologists in the clinic. Furthermore, this work may aid future researchers in exploring ML paradigms to improve clinical test batteries used by audiologists in achieving accurate diagnoses.

Robust 2-D-3-D Registration Optimization for Motion Compensation During 3-D TRUS-Guided Biopsy Using Learned Prostate Motion Data.

In magnetic resonance (MR)-targeted, 3-D transrectal ultrasound (TRUS)-guided biopsy, prostate motion during the procedure increases the needle targeting error and limits the ability to accurately sample MR-suspicious tumor volumes. The robustness of the 2-D-3-D registration methods for prostate motion compensation is impacted by local optima in the search space. In this paper, we analyzed the prostate motion characteristics and investigated methods to incorporate such knowledge into the registration optimization framework to improve robustness against local optima. Rigid motion of the prostate was analyzed adopting a mixture-of-Gaussian (MoG) model using 3-D TRUS images acquired at bilateral sextant probe positions with a mechanically assisted biopsy system. The learned motion characteristics were incorporated into Powell's direction set method by devising multiple initial search positions and initial search directions. Experiments were performed on data sets acquired during clinical biopsy procedures, and registration error was evaluated using target registration error (TRE) and converged image similarity metric values after optimization. After incorporating the learned initialization positions and directions in Powell's method, 2-D-3-D registration to compensate for motion during prostate biopsy was performed with rms ± std TRE of 2.33 ± 1.09 mm with ~3 s mean execution time per registration. This was an improvement over 3.12 ± 1.70 mm observed in Powell's standard approach. For the data acquired under clinical protocols, the converged image similarity metric value improved in ≥8% of the registrations whereas it degraded only ≤1% of the registrations. The reported improvements in optimization indicate useful advancements in robustness to ensure smooth clinical integration of a registration solution for motion compensation that facilitates accurate sampling of the smallest clinically significant tumors.

Automating dicentric chromosome detection from cytogenetic biodosimetry data.

We present a prototype software system with sufficient capacity and speed to estimate radiation exposures in a mass casualty event by counting dicentric chromosomes (DCs) in metaphase cells from many individuals. Top-ranked metaphase cell images are segmented by classifying and defining chromosomes with an active contour gradient vector field (GVF) and by determining centromere locations along the centreline. The centreline is extracted by discrete curve evolution (DCE) skeleton branch pruning and curve interpolation. Centromere detection minimises the global width and DAPI-staining intensity profiles along the centreline. A second centromere is identified by reapplying this procedure after masking the first. Dicentrics can be identified from features that capture width and intensity profile characteristics as well as local shape features of the object contour at candidate pixel locations. The correct location of the centromere is also refined in chromosomes with sister chromatid separation. The overall algorithm has both high sensitivity (85 %) and specificity (94 %). Results are independent of the shape and structure of chromosomes in different cells, or the laboratory preparation protocol followed. The prototype software was recoded in C++/OpenCV; image processing was accelerated by data and task parallelisation with Message Passaging Interface and Intel Threading Building Blocks and an asynchronous non-blocking I/O strategy. Relative to a serial process, metaphase ranking, GVF and DCE are, respectively, 100 and 300-fold faster on an 8-core desktop and 64-core cluster computers. The software was then ported to a 1024-core supercomputer, which processed 200 metaphase images each from 1025 specimens in 1.4 h.

Intensity integrated Laplacian-based thickness measurement for detecting human metaphase chromosome centromere location.

Accurate detection of the human metaphase chromosome centromere is an important step in many chromosome analysis and medical diagnosis algorithms. The centromere location can be utilized to derive information such as the chromosome type, polarity assignment, etc. Methods available in the literature yield unreliable results mainly due to high variability of morphology in metaphase chromosomes and boundary noise present in the image. In this paper, we have proposed a multistaged algorithm which includes the use of discrete curve evolution, gradient vector flow active contours, functional approximation of curve segments, and support vector machine classification. The standard Laplacian thickness measurement algorithm was enhanced to incorporate both contour information as well as intensity information to obtain a more accurate centromere location. In addition to segmentation and width profile measurement, the proposed algorithm can also correct for sister chromatid separation in cell images. The proposed method was observed to be more accurate and statistically significant as compared to a centerline-based method when tested with 226 human metaphase chromosomes.

2D-3D rigid registration to compensate for prostate motion during 3D TRUS-guided biopsy.

PURPOSE: Three-dimensional (3D) transrectal ultrasound (TRUS)-guided systems have been developed to improve targeting accuracy during prostate biopsy. However, prostate motion during the procedure is a potential source of error that can cause target misalignments. The authors present an image-based registration technique to compensate for prostate motion by registering the live two-dimensional (2D) TRUS images acquired during the biopsy procedure to a preacquired 3D TRUS image. The registration must be performed both accurately and quickly in order to be useful during the clinical procedure. METHODS: The authors implemented an intensity-based 2D-3D rigid registration algorithm optimizing the normalized cross-correlation (NCC) metric using Powell's method. The 2D TRUS images acquired during the procedure prior to biopsy gun firing were registered to the baseline 3D TRUS image acquired at the beginning of the procedure. The accuracy was measured by calculating the target registration error (TRE) using manually identified fiducials within the prostate; these fiducials were used for validation only and were not provided as inputs to the registration algorithm. They also evaluated the accuracy when the registrations were performed continuously throughout the biopsy by acquiring and registering live 2D TRUS images every second. This measured the improvement in accuracy resulting from performing the registration, continuously compensating for motion during the procedure. To further validate the method using a more challenging data set, registrations were performed using 3D TRUS images acquired by intentionally exerting different levels of ultrasound probe pressures in order to measure the performance of our algorithm when the prostate tissue was intentionally deformed. In this data set, biopsy scenarios were simulated by extracting 2D frames from the 3D TRUS images and registering them to the baseline 3D image. A graphics processing unit (GPU)-based implementation was used to improve the registration speed. They also studied the correlation between NCC and TREs. RESULTS: The root-mean-square (RMS) TRE of registrations performed prior to biopsy gun firing was found to be 1.87 ± 0.81 mm. This was an improvement over 4.75 ± 2.62 mm before registration. When the registrations were performed every second during the biopsy, the RMS TRE was reduced to 1.63 ± 0.51 mm. For 3D data sets acquired under different probe pressures, the RMS TRE was found to be 3.18 ± 1.6 mm. This was an improvement from 6.89 ± 4.1 mm before registration. With the GPU based implementation, the registrations were performed with a mean time of 1.1 s. The TRE showed a weak correlation with the similarity metric. However, the authors measured a generally convex shape of the metric around the ground truth, which may explain the rapid convergence of their algorithm to accurate results. CONCLUSIONS: Registration to compensate for prostate motion during 3D TRUS-guided biopsy can be performed with a measured accuracy of less than 2 mm and a speed of 1.1 s, which is an important step toward improving the targeting accuracy of a 3D TRUS-guided biopsy system.

Quantification of prostate deformation due to needle insertion during TRUS-guided biopsy: comparison of hand-held and mechanically stabilized systems.

PURPOSE: Prostate biopsy is the clinical standard for the definitive diagnosis of prostate cancer. To overcome the limitations of 2D TRUS-guided biopsy systems when targeting preplanned locations, systems have been developed with 3D guidance to improve the accuracy of cancer detection. Prostate deformation due to needle insertion and biopsy gun firing is a potential source of error that can cause target misalignments during biopsies. METHODS: The authors used nonrigid registration of 2D TRUS images to quantify the deformation that occurs during the needle insertion and the biopsy gun firing procedure and compare this effect in biopsies performed using a hand-held TRUS probe to those performed using a mechanically assisted 3D TRUS-guided biopsy system. The authors calculated a spatially varying 95% confidence interval on the prostate tissue motion and analyzed this motion both as a function of distance to the biopsy needle and as a function of distance to the lower piercing point of the prostate. The former is relevant because biopsy targets lie along the needle axis, and the latter is of particular importance due to the reported high concentration of prostate cancer in the peripheral zone, a substantial portion of which lies on the posterior side of the prostate where biopsy needles enter the prostate after penetrating the rectal wall during transrectal biopsy. RESULTS: The results show that for both systems, the tissue deformation is such that throughout the length of the needle axis, including regions proximal to the lower piercing point, spherical tumors with a radius of 2.1 mm or more can be sampled with 95% confidence under the assumption of zero error elsewhere in the biopsy system. More deformation was observed in the direction orthogonal to the needle axis compared to the direction parallel to the needle axis; this is of particular importance given the long, narrow shape of the biopsy core. The authors measured lateral tissue motion proximal to the needle axis of not more than 1.5 mm, with 95% confidence. The authors observed a statistically significant and clinically insignificant maximum difference of 0.38 mm in the deformation, resulting from the hand-held and mechanically assisted systems along the needle axis, and the mechanical system resulted in a lower relative increase in deformation proximal to the needle axis during needle insertion, as well as lower variability of deformation during biopsy gun firing. CONCLUSIONS: Given the clinical need to biopsy tumors of volume greater than or equal to 0.5 cm3, corresponding to spherical tumors with a radius of 5 mm or more, the tissue motion induced by needle insertion and gun firing is an important consideration when setting the design specifications for TRUS-guided prostate biopsy systems.

Quantification of prostate deformation due to needle insertion during TRUS-guided biopsy.

Prostate biopsy is the clinical standard for the diagnosis of prostate cancer, and technologies for 3D guidance to targets and recording of biopsy locations are promising approaches to reducing the need for repeated biopsies. In this study, we use image-based non-rigid registration to quantify prostate deformation during needle insertion and biopsy gun firing, in order to provide information useful to the overall assessment of a TRUS-guided biopsy system's expected targeting error. We recorded mean tissue displacements of up to 0.4 mm, accounting for 16% of the clinically-motivated maximum desired RMS error of a guidance system.

Order-preserving moves for graph-cut-based optimization.

In the last decade, graph-cut optimization has been popular for a variety of labeling problems. Typically, graph-cut methods are used to incorporate smoothness constraints on a labeling, encouraging most nearby pixels to have equal or similar labels. In addition to smoothness, ordering constraints on labels are also useful. For example, in object segmentation, a pixel with a "car wheel" label may be prohibited above a pixel with a "car roof" label. We observe that the commonly used graph-cut \alpha-expansion move algorithm is more likely to get stuck in a local minimum when ordering constraints are used. For a certain model with ordering constraints, we develop new graph-cut moves which we call order-preserving. The advantage of order-preserving moves is that they act on all labels simultaneously, unlike \alpha-expansion. More importantly, for most labels \alpha, the set of \alpha-expansion moves is strictly smaller than the set of order-preserving moves. This helps to explain why in practice optimization with order-preserving moves performs significantly better than \alpha-expansion in the presence of ordering constraints. We evaluate order-preserving moves for the geometric class scene labeling (introduced by Hoiem et al.) where the goal is to assign each pixel a label such as "sky," "ground," etc., so ordering constraints arise naturally. In addition, we use order-preserving moves for certain simple shape priors in graph-cut segmentation, which is a novel contribution in itself.

Nonrigid registration of three-dimensional ultrasound and magnetic resonance images of the carotid arteries.

Atherosclerosis at the carotid bifurcation can result in cerebral emboli, which in turn can block the blood supply to the brain causing ischemic strokes. Noninvasive imaging tools that better characterize arterial wall, and atherosclerotic plaque structure and composition may help to determine the factors which lead to the development of unstable lesions, and identify patients at risk of plaque disruption and stroke. Carotid magnetic resonance (MR) imaging allows for the characterization of carotid vessel wall and plaque composition, the characterization of normal and pathological arterial wall, the quantification of plaque size, and the detection of plaque integrity. On the other hand, various ultrasound (US) measurements have also been used to quantify atherosclerosis, carotid stenosis, intima-media thickness, total plaque volume, total plaque area, and vessel wall volume. Combining the complementary information provided by 3D MR and US carotid images may lead to a better understanding of the underlying compositional and textural factors that define plaque and wall vulnerability, which may lead to better and more effective stroke prevention strategies and patient management. Combining these images requires nonrigid registration to correct the nonlinear misalignments caused by relative twisting and bending in the neck due to different head positions during the two image acquisition sessions. The high degree of freedom and large number of parameters associated with existing nonrigid image registration methods causes several problems including unnatural plaque morphology alteration, high computational complexity, and low reliability. Thus, a "twisting and bending" model was used with only six parameters to model the normal movement of the neck for nonrigid registration. The registration technique was evaluated using 3D US and MR carotid images at two field strengths, 1.5 and 3.0 T, of the same subject acquired on the same day. The mean registration error between the segmented carotid artery wall boundaries in the target US image and the registered MR images was calculated using a distance-based error metric after applying a "twisting and bending" model based nonrigid registration algorithm. An average registration error of 1.4 +/- 0.3 mm was obtained for 1.5 T MR and 1.5 +/- 0.4 mm for 3.0 T MR, when registered with 3D US images using the nonrigid registration technique presented in this paper. Visual inspection of segmented vessel surfaces also showed a substantial improvement of alignment with this nonrigid registration technique compared to rigid registration.

Robust and efficient feature tracking for indoor navigation.

Robust feature tracking is a requirement for many computer vision tasks such as indoor robot navigation. However, indoor scenes are characterized by poorly localizable features. As a result, indoor feature tracking without artificial markers is challenging and remains an attractive problem. We propose to solve this problem by constraining the locations of a large number of nondistinctive features by several planar homographies which are strategically computed using distinctive features. We experimentally show the need for multiple homographies and propose an illumination-invariant local-optimization scheme for motion refinement. The use of a large number of nondistinctive features within the constraints imposed by planar homographies allows us to gain robustness. Also, the lesser computation cost in estimating these nondistinctive features helps to maintain the efficiency of the proposed method. Our local-optimization scheme produces subpixel accurate feature motion. As a result, we are able to achieve robust and accurate feature tracking.

A "twisting and bending" model-based nonrigid image registration technique for 3-D ultrasound carotid images.

Atherosclerosis at the carotid bifurcation resulting in cerebral emboli is a major cause of ischemic stroke. Most strokes associated with carotid atherosclerosis can be prevented by lifestyle/dietary changes and pharmacological treatments if identified early by monitoring carotid plaque changes. Registration of 3-D ultrasound (US) images of carotid plaque obtained at different time points is essential for sensitive monitoring of plaque changes in volume and surface morphology. This registration technique should be nonrigid, since different head positions during image acquisition sessions cause relative bending and torsion in the neck, producing nonlinear deformations between the images. We modeled the movement of the neck using a "twisting and bending" model with only six parameters for nonrigid registration. We evaluated the algorithm using 3-D US carotid images acquired at two different head positions to simulate images acquired at different times. We calculated the mean registration error (MRE) between the segmented vessel surfaces in the target image and the registered image using a distance-based error metric after applying our "twisting and bending" model-based nonrigid registration algorithm. We achieved an average registration error of 0.80 +/-0.26 mm using our nonrigid registration technique, which was a significant improvement in registration accuracy over rigid registration, even with reduced degrees-of-freedom compared to the other nonrigid registration algorithms.

Prostate segmentation by feature enhancement using domain knowledge and adaptive region based operations.

Estimation of prostate location and volume is essential in determining a dose plan for ultrasound-guided brachytherapy, a common prostate cancer treatment. However, manual segmentation is difficult, time consuming and prone to variability. In this paper, we present a semi-automatic discrete dynamic contour (DDC) model based image segmentation algorithm, which effectively combines a multi-resolution model refinement procedure together with the domain knowledge of the image class. The segmentation begins on a low-resolution image by defining a closed DDC model by the user. This contour model is then deformed progressively towards higher resolution images. We use a combination of a domain knowledge based fuzzy inference system (FIS) and a set of adaptive region based operators to enhance the edges of interest and to govern the model refinement using a DDC model. The automatic vertex relocation process, embedded into the algorithm, relocates deviated contour points back onto the actual prostate boundary, eliminating the need of user interaction after initialization. The accuracy of the prostate boundary produced by the proposed algorithm was evaluated by comparing it with a manually outlined contour by an expert observer. We used this algorithm to segment the prostate boundary in 114 2D transrectal ultrasound (TRUS) images of six patients scheduled for brachytherapy. The mean distance between the contours produced by the proposed algorithm and the manual outlines was 2.70 +/- 0.51 pixels (0.54 +/- 0.10 mm). We also showed that the algorithm is insensitive to variations of the initial model and parameter values, thus increasing the accuracy and reproducibility of the resulting boundaries in the presence of noise and artefacts.