Use of Deep Learning to Detect the Maternal Heart Rate and False Signals on Fetal Heart Rate Recordings.

We have developed deep learning models for automatic identification of the maternal heart rate (MHR) and, more generally, false signals (FSs) on fetal heart rate (FHR) recordings. The models can be used to preprocess FHR data prior to automated analysis or as a clinical alert system to assist the practitioner. Three models were developed and used to detect (i) FSs on the MHR channel (the FSMHR model), (ii) the MHR and FSs on the Doppler FHR sensor (the FSDop model), and (iii) FSs on the scalp ECG channel (the FSScalp model). The FSDop model was the most useful because FSs are far more frequent on the Doppler FHR channel. All three models were based on a multilayer, symmetric, GRU, and were trained on data recorded during the first and second stages of delivery. The FSMHR and FSDop models were also trained on antepartum recordings. The training dataset contained 1030 expert-annotated periods (mean duration: 36 min) from 635 recordings. In an initial evaluation of routine clinical practice, 30 fully annotated recordings for each sensor type (mean duration: 5 h for MHR and Doppler sensors, and 3 h for the scalp ECG sensor) were analyzed. The sensitivity, positive predictive value (PPV) and accuracy were respectively 62.20%, 87.1% and 99.90% for the FSMHR model, 93.1%, 95.6% and 99.68% for the FSDop model, and 44.6%, 87.2% and 99.93% for the FSScalp model. We built a second test dataset with a more solid ground truth by selecting 45 periods (lasting 20 min, on average) on which the Doppler FHR and scalp ECG signals were recorded simultaneously. Using scalp ECG data, the experts estimated the true FHR value more reliably and thus annotated the Doppler FHR channel more precisely. The models achieved a sensitivity of 53.3%, a PPV of 62.4%, and an accuracy of 97.29%. In comparison, two experts (blinded to the scalp ECG data) respectively achieved a sensitivity of 15.7%, a PPV of 74.3%, and an accuracy of 96.91% and a sensitivity of 60.7%, a PPV of 83.5% and an accuracy of 98.24%. Hence, the models performed at expert level (better than one expert and worse than the other), although a well-trained expert with good knowledge of FSs could probably do better in some cases. The models and datasets have been included in the Fetal Heart Rate Morphological Analysis open-source MATLAB toolbox and can be used freely for research purposes.

Oligoclonal Band Straightening Based on Optimized Hierarchical Warping for Multiple Sclerosis Diagnosis.

The detection of immunoglobulin G (IgG) oligoclonal bands (OCB) in cerebrospinal fluid (CSF) by isoelectric focusing (IEF) is a valuable tool for the diagnosis of multiple sclerosis. Over the last decade, the results of our clinical research have suggested that tears are a non-invasive alternative to CSF. However, since tear samples have a lower IgG concentration than CSF, a sensitive OCB detection is therefore required. We are developing the first automatic tool for IEF analysis, with a view to speeding up the current visual inspection method, removing user variability, reducing misinterpretation, and facilitating OCB quantification and follow-up studies. The removal of band distortion is a key image enhancement step in increasing the reliability of automatic OCB detection. Here, we describe a novel, fully automatic band-straightening algorithm. The algorithm is based on a correlation directional warping function, estimated using an energy minimization procedure. The approach was optimized via an innovative coupling of a hierarchy of image resolutions to a hierarchy of transformation, in which band misalignment is corrected at successively finer scales. The algorithm's performance was assessed in terms of the bands' standard deviation before and after straightening, using a synthetic dataset and a set of 200 lanes of CSF, tear, serum and control samples on which experts had manually delineated the bands. The number of distorted bands was divided by almost 16 for the synthetic lanes and by 7 for the test dataset of real lanes. This method can be applied effectively to different sample types. It can realign minimal contrast bands and is robust for non-uniform deformations.

Toward an automatic tool for oligoclonal band detection in cerebrospinal fluid and tears for multiple sclerosis diagnosis: lane segmentation based on a ribbon univariate open active contour.

The latest revision of multiple sclerosis diagnosis guidelines emphasizes the role of oligoclonal band detection in isoelectric focusing images of cerebrospinal fluid. Recent studies suggest tears as a promising noninvasive alternative to cerebrospinal fluid. We are developing the first automatic method for isoelectric focusing image analysis and oligoclonal band detection in cerebrospinal fluid and tear samples. The automatic analysis would provide an accurate, fast analysis and would reduce the expert-dependent variability and errors of the current visual analysis. In this paper, we describe a new effective model for the fully automated segmentation of highly distorted lanes in isoelectric focusing images. This approach is a new formulation of the classic parametric active contour problem, in which an open active contour is constrained to move from the top to the bottom of the image, and the x-axis coordinate is expressed as a function of the y-axis coordinate. The left and right edges of the lane evolved together in a ribbon-like shape so that the full width of the lane was captured reliably. The segmentation algorithm was implemented using a multiresolution approach in which the scale factor and the active contour control points were progressively increased. The lane segmentation algorithm was tested on a database of 51 isoelectric focusing images containing 419 analyzable lanes. The new model gave robust results for highly curved lanes, weak edges, and low-contrast lanes. A total of 98.8% of the lanes were perfectly segmented, and the remaining 1.2% had only minor errors. The computation time (1 s per membrane) is negligible. This method precisely defines the region of interest in each lane and thus is a major step toward the first fully automatic tool for oligoclonal band detection in isoelectric focusing images. Graphical abstract.

Fetal heart rate baseline computation with a weighted median filter.

BACKGROUND: Automated fetal heart rate (FHR) analysis removes inter- and intra-expert variability, and is a promising solution for reducing the occurrence of fetal acidosis and the implementation of unnecessary medical procedures. The first steps in automated FHR analysis are determination of the baseline, and detection of accelerations and decelerations (A/D). We describe a new method in which a weighted median filter baseline (WMFB) is computed and A/Ds are then detected. METHOD: The filter weightings are based on the prior probability that the sampled FHR is in the baseline state or in an A/D state. This probability is computed by estimating the signal's stability at low frequencies and by progressively trimming the signal. Using a competition dataset of 90 previously annotated FHR recordings, we evaluated the WMFB method and 11 recently published literature methods against the ground truth of an expert consensus. The level of agreement between the WMFB method and the expert consensus was estimated by calculating several indices (primarily the morphological analysis discordance index, MADI). The agreement indices were then compared with the values for eleven other methods. We also compared the level of method-expert agreement with the level of interrater agreement. RESULTS: For the WMFB method, the MADI indicated a disagreement of 4.02% vs. the consensus; this value is significantly lower (p<10-13) than that calculated for the best of the 11 literature methods (7.27%, for Lu and Wei's empirical mode decomposition method). The level of inter-expert agreement (according to the MADI) and the level of WMFB-expert agreement did not differ significantly (p=0.22). CONCLUSION: The WMFB method reproduced the expert consensus analysis better than 11 other methods. No differences in performance between the WMFB method and individual experts were observed. The method Matlab source code is available under General Public Licence at http://utsb.univ-catholille.fr/fhr-wmfb.

Cervicocephalic relocation test to evaluate cervical proprioception in adolescent idiopathic scoliosis.

PURPOSE: Adolescent idiopathic scoliosis (AIS) is a three-dimensional deformity of the spine associated with disturbed postural control. Cervical proprioception participates in controlling orthostatic posture via its influence on head stabilization. We hypothesized that patients with AIS exhibit altered cervical proprioception. METHODS: We conducted a case-control study to evaluate cervical proprioception using the cervicocephalic relocation test (CRT) in 30 adolescents with AIS (15.5 ± 1.5 years; Cobb 24.8° ± 9.5°) versus 14 non-scoliotic controls (14.6 ± 2.0 years). CRT evaluates cervical proprioception by measuring the capacity to relocate the head on the trunk after active rotation of the head in the transversal plane without visual control. Each subject performed ten right and then ten left head rotations. RESULTS: The CRT results were pathological in 12 AIS patients (40 %). The CRT mean was significantly different between AIS patients with a pathological CRT (5° ± 1.4° for right rotation; 4.2° ± 0.9° for left rotation) compared with AIS patients with a normal CRT (2.7° ± 0.6° for right rotation; 2.9° ± 0.8° for left rotation) or with the control group (3.5° ± 2.1° for right rotation; 3.1° ± 1.2° for left rotation). CONCLUSION: Cervical proprioception is impaired in certain AIS patients. This anomaly may worsen the prognosis of AIS (headache; balance disorders; worsened spinal deformity; complication after spinal fusion). We recommend systematic screening for altered cervical proprioception in AIS patients.

Baseline fetal heart rate analysis: eleven automatic methods versus expert consensus.

Visual analysis of fetal heart rate (FHR) during labor is subject to inter- and intra-observer variability that is particularly troublesome for anomalous recordings. Automatic FHR analysis has been proposed as a promising way to reduce this variability. The major difficulty with automatic analysis is to determine the baseline from which accelerations and decelerations will be detected. Eleven methods for automatic FHR analysis were reprogrammed using description from the literature and applied to 66 FHR recordings collected during the first stage of delivery. The FHR baselines produced by the automatic methods were compared with the baseline defined by agreement among a panel of three experts. The better performance of the automatic methods described by Mongelli, Lu, Wrobel and Pardey was noted despite their different approaches on signal processing. Nevertheless, for several recordings, none of the automatic studied methods produced a baseline similar to that defined by the experts.

Evaluation of an Expanded Disability Status Scale (EDSS) modeling strategy in multiple sclerosis.

The Expanded Disability Status Scale (EDSS) is the most widely used scale to evaluate the degree of neurological impairment in multiple sclerosis (MS). In this paper, we report on the evaluation of an EDSS modeling strategy based on recurrence quantification analysis (RQA) of posturographic data (i.e., center of pressure, COP). A total of 133 volunteers with EDSS ranging from 0 to 4.5 participated in this study, with eyes closed. After selection of time delay (τ), embedding dimension (m) as well as threshold (radius, r) to identify recurrent points, several RQA measures were calculated for each COP's position and velocity data in the mono- and multi-dimensional RQAs. Estimation results lead to the selection of the recurrence rate (RR) of the COP's position as the most pertinent RQA measure. The performance of the models versus raw and noisy data was higher in the mono-dimensional analysis than in the multi-dimensional. This study suggests that the posturographic signal's mono-dimensional RQA is a more pertinent method to quantify disability in MS than the multi-dimensional RQA.

Prostate contours delineation using interactive directional active contours model and parametric shape prior model.

Prostate contours delineation on Magnetic Resonance (MR) images is a challenging and important task in medical imaging with applications of guiding biopsy, surgery and therapy. While a fully automated method is highly desired for this application, it can be a very difficult task due to the structure and surrounding tissues of the prostate gland. Traditional active contours-based delineation algorithms are typically quite successful for piecewise constant images. Nevertheless, when MR images have diffuse edges or multiple similar objects (e.g. bladder close to prostate) within close proximity, such approaches have proven to be unsuccessful. In order to mitigate these problems, we proposed a new framework for bi-stage contours delineation algorithm based on directional active contours (DAC) incorporating prior knowledge of the prostate shape. We first explicitly addressed the prostate contour delineation problem based on fast globally DAC that incorporates both statistical and parametric shape prior model. In doing so, we were able to exploit the global aspects of contour delineation problem by incorporating a user feedback in contours delineation process where it is shown that only a small amount of user input can sometimes resolve ambiguous scenarios raised by DAC. In addition, once the prostate contours have been delineated, a cost functional is designed to incorporate both user feedback interaction and the parametric shape prior model. Using data from publicly available prostate MR datasets, which includes several challenging clinical datasets, we highlighted the effectiveness and the capability of the proposed algorithm. Besides, the algorithm has been compared with several state-of-the-art methods.

Development of CBCT-based prostate setup correction strategies and impact of rectal distension.

BACKGROUND: Cone-beam computed tomography (CBCT) image-guided radiotherapy (IGRT) systems are widely used tools to verify and correct the target position before each fraction, allowing to maximize treatment accuracy and precision. In this study, we evaluate automatic three-dimensional intensity-based rigid registration (RR) methods for prostate setup correction using CBCT scans and study the impact of rectal distension on registration quality. METHODS: We retrospectively analyzed 115 CBCT scans of 10 prostate patients. CT-to-CBCT registration was performed using (a) global RR, (b) bony RR, or (c) bony RR refined by a local prostate RR using the CT clinical target volume (CTV) expanded with 1-to-20-mm varying margins. After propagation of the manual CT contours, automatic CBCT contours were generated. For evaluation, a radiation oncologist manually delineated the CTV on the CBCT scans. The propagated and manual CBCT contours were compared using the Dice similarity and a measure based on the bidirectional local distance (BLD). We also conducted a blind visual assessment of the quality of the propagated segmentations. Moreover, we automatically quantified rectal distension between the CT and CBCT scans without using the manual CBCT contours and we investigated its correlation with the registration failures. To improve the registration quality, the air in the rectum was replaced with soft tissue using a filter. The results with and without filtering were compared. RESULTS: The statistical analysis of the Dice coefficients and the BLD values resulted in highly significant differences (p<10(-6)) for the 5-mm and 8-mm local RRs vs the global, bony and 1-mm local RRs. The 8-mm local RR provided the best compromise between accuracy and robustness (Dice median of 0.814 and 97% of success with filtering the air in the rectum). We observed that all failures were due to high rectal distension. Moreover, the visual assessment confirmed the superiority of the 8-mm local RR over the bony RR. CONCLUSION: The most successful CT-to-CBCT RR method proved to be the 8-mm local RR. We have shown the correlation between its registration failures and rectal distension. Furthermore, we have provided a simple (easily applicable in routine) and automatic method to quantify rectal distension and to predict registration failure using only the manual CT contours.

Dual adaptive filtering by optimal projection applied to filter muscle artifacts on EEG and comparative study.

Muscle artifacts constitute one of the major problems in electroencephalogram (EEG) examinations, particularly for the diagnosis of epilepsy, where pathological rhythms occur within the same frequency bands as those of artifacts. This paper proposes to use the method dual adaptive filtering by optimal projection (DAFOP) to automatically remove artifacts while preserving true cerebral signals. DAFOP is a two-step method. The first step consists in applying the common spatial pattern (CSP) method to two frequency windows to identify the slowest components which will be considered as cerebral sources. The two frequency windows are defined by optimizing convolutional filters. The second step consists in using a regression method to reconstruct the signal independently within various frequency windows. This method was evaluated by two neurologists on a selection of 114 pages with muscle artifacts, from 20 clinical recordings of awake and sleeping adults, subject to pathological signals and epileptic seizures. A blind comparison was then conducted with the canonical correlation analysis (CCA) method and conventional low-pass filtering at 30 Hz. The filtering rate was 84.3% for muscle artifacts with a 6.4% reduction of cerebral signals even for the fastest waves. DAFOP was found to be significantly more efficient than CCA and 30 Hz filters. The DAFOP method is fast and automatic and can be easily used in clinical EEG recordings.

Evaluation of the AFOP/DAFOP method for automatic filtering of EEGs of patients with epilepsy.

OBJECTIVE: Further developments in EEG monitoring necessitate new methods of filtering to eliminate artifacts, without transforming relevant signals. This article presents an automatic filtering of EEG recordings, based on a spatio-temporal method called Adaptive Filtering by Optimal Projection or Dual Adaptive Filtering by Optimal Projection. Evaluation of filtering methods is difficult, and comparisons between methods remain a challenge; here, we present a method to score the visual assessment of the EEG. The aim of this study was to evaluate an automatic filtering method, called Adaptive Filtering by Optimal Projection, improved by Dual Adaptive Filtering by Optimal Projection, of EEG recordings of patients with epilepsy. METHODS: Two hundred forty-eight nonfiltered EEG segments of 20 seconds each were selected from 35 EEG recordings of 27 different patients by 3 clinical neurophysiologists based on their content. The reading quality as well as the proportions of artifacts and of cerebral activity removed after filtering were evaluated on a scale of 0 to 4. The mean square difference of amplitude before and after filtering was computed in specific spectral band. RESULTS: The artifacts were largely removed (82% for muscular, 72% for ocular, and 71% for electrode artifacts). The readability was improved on an average by two points for pages containing epileptic seizures, and by one point for those containing alpha rhythms, slow waves, and spikes. After filtering, consistency tests showed a consensus (Spearman correlation [0.69-0.79]) on the removal of the artifact versus loss of information. The spectral analysis showed equivalent results (0.16% mean square difference in the alpha band). CONCLUSIONS: Our filtering method is effective in removing artifacts without altering relevant signals. The significance is that we evaluated a new automated method of filtering EEG that is easy to use for both for the analysis of routine EEG and in the field of epilepsy at large.

BioSigPlot: an opensource tool for the visualization of multi-channel biomedical signals with Matlab.

This paper presents a Matlab-based software (MathWorks inc.) called BioSigPlot for the visualization of multi-channel biomedical signals, particularly for the EEG. This tool is designed for researchers on both engineering and medicine who have to collaborate to visualize and analyze signals. It aims to provide a highly customizable interface for signal processing experimentation in order to plot several kinds of signals while integrating the common tools for physician. The main advantages compared to other existing programs are the multi-dataset displaying, the synchronization with video and the online processing. On top of that, this program uses object oriented programming, so that the interface can be controlled by both graphic controls and command lines. It can be used as EEGlab plug-in but, since it is not limited to EEG, it would be distributed separately. BioSigPlot is distributed free of charge (http://biosigplot.sourceforge.net), under the terms of GNU Public License for non-commercial use and open source development.

Automatic prostate segmentation in cone-beam computed tomography images using rigid registration.

We propose to evaluate automatic three-dimensional gray-value rigid registration (RR) methods for prostate localization on cone-beam computed tomography (CBCT) scans. In total, 103 CBCT scans of 9 prostate patients have been analyzed. Each one was registered to the planning CT scan using different methods: (a) global RR, (b) pelvis bone structure RR, (c) bone RR refined by local soft-tissue RR using the CT clinical target volume (CTV) expanded with a 1, 3, 5, 8, 10, 12, 15 or 20-mm margin. To evaluate results, a radiation oncologist was asked to manually delineate the CTV on the CBCT scans. The Dice coefficients between each automatic CBCT segmentation - derived from the transformation of the manual CT segmentation - and the manual CBCT segmentation were calculated. Global or bone CT/CBCT RR has been shown to yield insufficient results in average. Local RR with an 8-mm margin around the CTV after bone RR was found to be the best candidate for systematically significantly improving prostate localization.

Expanded Disability Status Scale (EDSS) estimation in multiple sclerosis from posturographic data.

Expanded Disability Status Scale (EDSS) is the most widely used clinical scale to evaluate levels of multiple sclerosis (MS). As MS can lead to disruptions in the regulation of balance and the disability can be evaluated by force platform posturography, we have developed in this study a new strategy to estimate EDSS from posturographic data. 118 volunteers with EDSS ranging from 0 to 4.5 participated in this study, with eyes closed. By using second-order polynomial regression models, EDSS was estimated from two postural sway parameters, respectively, the length and the surface and four recurrence quantification analysis (RQA) parameters: percentage of recurrence (%Rec), Shannon entropy (Ent), mean diagonal line length (LL) and trapping time (TT). In addition, all four RQA parameters were calculated for position, instantaneous velocity and acceleration of the center of pressure. In order to select the most accurate method for estimating EDSS, four statistical indices (percentage of agreement, underestimation and overestimation, as well as Mean error) were calculated comparing clinical and estimated EDSS scores. The results demonstrate that estimations of EDSS from surface, %Rec and LL of position, best agreed with clinical scores. This study emphasizes the possibility of distinguishing EDSS scores using postural sway and RQA parameters.

A global approach for automatic artifact removal for standard EEG record.

The EEG signal is a record of the brain activity using multiple electrodes placed on the scalp. Unfortunately, it can be hardly contaminated by a lot of noises called artifacts. These latter can be generated by various actions such as eye blinks, eye movements or the skeletal muscle activities (jaw, forehead, ...). This study will focus on a global artifact removal method using independent component analysis (ICA) on signals cut in frequency bands. The interest of this method resides in automatizing the artifactual source identification and enables a global filtering of records using constant bases. A brief overview of the project will be made in order to introduce the method used. Next, the results will be presented and their validation will be discussed in the conclusion.