MyMedR

Displaying all 4 publications

Abstract:

Sort:

Fulltext Street Sign Recognition Using Histogram of Oriented Gradients and Artificial Neural Networks

Islam KT, Wijewickrema S, Raj RG, O'Leary S

J Imaging, 2019 Apr 03;5(4).
PMID: 34460482 DOI: 10.3390/jimaging5040044

Street sign identification is an important problem in applications such as autonomous vehicle navigation and aids for individuals with vision impairments. It can be especially useful in instances where navigation techniques such as global positioning system (GPS) are not available. In this paper, we present a method of detection and interpretation of Malaysian street signs using image processing and machine learning techniques. First, we eliminate the background from an image to segment the region of interest (i.e., the street sign). Then, we extract the text from the segmented image and classify it. Finally, we present the identified text to the user as a voice notification. We also show through experimental results that the system performs well in real-time with a high level of accuracy. To this end, we use a database of Malaysian street sign images captured through an on-board camera.
Fulltext A Vision-Based Machine Learning Method for Barrier Access Control Using Vehicle License Plate Authentication

Islam KT, Raj RG, Shamsul Islam SM, Wijewickrema S, Hossain MS, Razmovski T, et al.

Sensors (Basel), 2020 Jun 24;20(12).
PMID: 32599883 DOI: 10.3390/s20123578

Automatic vehicle license plate recognition is an essential part of intelligent vehicle access control and monitoring systems. With the increasing number of vehicles, it is important that an effective real-time system for automated license plate recognition is developed. Computer vision techniques are typically used for this task. However, it remains a challenging problem, as both high accuracy and low processing time are required in such a system. Here, we propose a method for license plate recognition that seeks to find a balance between these two requirements. The proposed method consists of two stages: detection and recognition. In the detection stage, the image is processed so that a region of interest is identified. In the recognition stage, features are extracted from the region of interest using the histogram of oriented gradients method. These features are then used to train an artificial neural network to identify characters in the license plate. Experimental results show that the proposed method achieves a high level of accuracy as well as low processing time when compared to existing methods, indicating that it is suitable for real-time applications.
Electrocochleography triggered intervention successfully preserves residual hearing during cochlear implantation: Results of a randomised clinical trial

Bester C, Collins A, Razmovski T, Weder S, Briggs RJ, Wei B, et al.

Hear Res, 2022 Dec;426:108353.
PMID: 34600798 DOI: 10.1016/j.heares.2021.108353

BACKGROUND: Preservation of natural hearing during cochlear implantation is associated with improved speech outcomes, however more than half of implant recipients lose this hearing. Real-time electrophysiological monitoring of cochlear output during implantation, made possible by recording electrocochleography using the electrodes on the cochlear implant, has shown promise in predicting hearing preservation. Sudden drops in the amplitude of the cochlear microphonic (CM) have been shown to predict more severe hearing losses. Here, we report on a randomized clinical trial investigating whether immediate surgical intervention triggered by these drops can save residual hearing.
METHODS: A single-blinded placebo-controlled trial of surgical intervention triggered when CM amplitude dropped by at least 30% of a prior maximum amplitude during cochlear implantation. Intraoperative electrocochleography was recorded in 60 adults implanted with Cochlear Ltd's Thin Straight Electrode, half randomly assigned to a control group and half to an interventional group. The surgical intervention was to withdraw the electrode in ½-mm steps to recover CM amplitude. The primary outcome was hearing preservation 3 months following implantation, with secondary outcomes of speech-in-noise reception thresholds by group or CM outcome, and depth of implantation.
RESULTS: Sixty patients were recruited; neither pre-operative audiometry nor speech reception thresholds were significantly different between groups. Post-operatively, hearing preservation was significantly better in the interventional group. This was the case in absolute difference (median of 30 dB for control, 20 dB for interventional, χ² = 6.2, p = .013), as well as for relative difference (medians of 66% for the control, 31% for the interventional, χ² = 5.9, p = .015). Speech-in-noise reception thresholds were significantly better in patients with no CM drop at any point during insertion compared with patients with a CM drop; however, those with successfully recovered CMs after an initial drop were not significantly different (median gain required for speech reception score of 50% above noise of 6.9 dB for no drop, 8.6 for recovered CM, and 9.8 for CM drop, χ² = 6.8, p = .032). Angular insertion depth was not significantly different between control and interventional groups.
CONCLUSIONS: This is the first demonstration that surgical intervention in response to intraoperative hearing monitoring can save residual hearing during cochlear implantation.
Balloon catheters versus vaginal prostaglandins for labour induction (CPI Collaborative): an individual participant data meta-analysis of randomised controlled trials

Jones MN, Palmer KR, Pathirana MM, Cecatti JG, Filho OBM, Marions L, et al.

Lancet, 2022 Nov 12;400(10364):1681-1692.
PMID: 36366885 DOI: 10.1016/S0140-6736(22)01845-1

BACKGROUND: Induction of labour is one of the most common obstetric interventions globally. Balloon catheters and vaginal prostaglandins are widely used to ripen the cervix in labour induction. We aimed to compare the effectiveness and safety profiles of these two induction methods.
METHODS: We did an individual participant data meta-analysis comparing balloon catheters and vaginal prostaglandins for cervical ripening before labour induction. We systematically identified published and unpublished randomised controlled trials that completed data collection between March 19, 2019, and May 1, 2021, by searching the Cochrane Library, ClinicalTrials.gov, WHO International Clinical Trials Registry Platform, and PubMed. Further trials done before March 19, 2019, were identified through a recent Cochrane review. Data relating to the combined use of the two methods were not included, only data from women with a viable, singleton pregnancy were analysed, and no exclusion was made based on parity or membrane status. We contacted authors of individuals trials and participant-level data were harmonised and recoded according to predefined definitions of variables. Risk of bias was assessed with the ROB2 tool. The primary outcomes were caesarean delivery, indication for caesarean delivery, a composite adverse perinatal outcome, and a composite adverse maternal outcome. We followed the intention-to-treat principle for the main analysis. The primary meta-analysis used two-stage random-effects models and the sensitivity analysis used one-stage mixed models. All models were adjusted for maternal age and parity. This meta-analysis is registered with PROSPERO (CRD42020179924).
FINDINGS: Individual participant data were available from 12 studies with a total of 5460 participants. Balloon catheters, compared with vaginal prostaglandins, did not lead to a significantly different rate of caesarean delivery (12 trials, 5414 women; crude incidence 27·0%; adjusted OR [aOR] 1·09, 95% CI 0·95-1·24; I2=0%), caesarean delivery for failure to progress (11 trials, 4601 women; aOR 1·20, 95% CI 0·91-1·58; I2=39%), or caesarean delivery for fetal distress (10 trials, 4441 women; aOR 0·86, 95% CI 0·71-1·04; I2=0%). The composite adverse perinatal outcome was lower in women who were allocated to balloon catheters than in those allocated to vaginal prostaglandins (ten trials, 4452 neonates, crude incidence 13·6%; aOR 0·80, 95% CI 0·70-0·92; I2=0%). There was no significant difference in the composite adverse maternal outcome (ten trials, 4326 women, crude incidence 22·7%; aOR 1·02, 95% CI 0·89-1·18; I2=0%).
INTERPRETATION: In induction of labour, balloon catheters and vaginal prostaglandins have comparable caesarean delivery rates and maternal safety profiles, but balloon catheters lead to fewer adverse perinatal events.
FUNDING: Australian National Health and Medical Research Council and Monash Health Emerging Researcher Fellowship.