Towards an affordable magnetomyography instrumentation and low model complexity approach for labour imminency prediction using a novel multiresolution analysis

Abstract

The ability to predict the onset of labour is seen to be an important tool in a clinical setting. Magnetomyography has shown promise in the area of labour imminency prediction, but its clinical application remains limited due to high resource consumption associated with its broad number of channels. In this study, five electrode channels, which account for 3.3% of the total, are used alongside a novel signal decomposition algorithm and low complexity classifiers (logistic regression and linear‐SVM) to classify between labour imminency due within 0 to 48 hours and >48 hours. The results suggest that the parsimonious representation comprising of five electrode channels and novel signal decomposition method alongside the candidate classifiers could allow for greater affordability and hence clinical viability of the magnetomyography‐based prediction model, which carries a good degree of model interpretability. The results showed around a 20% increase on average for the novel decomposition method, alongside a reduced group of features across the various classification metrics considered for both the logistic regression and support vector machine.

Multiscale depth of anaesthesia prediction for surgery using frontal cortex electroencephalography

“…The decomposition method is efficient and works in the time domain with a set of heuristically tuned linear thresholds as a basis function, in addition to a peak identification sequencewhich together ultimately yields a series of decomposed signals based on amplitude bands-followed by a learning process which is performed via a performance index that is used to select an optimal region from the candidate decompositions [20,[24][25][26]. The decomposition method has seen subsequent application in various aspects of clinical medicine involving the analysis of nonlinear and stochastic physiological signals such as rehabilitation, pregnancy medicine, and more recently in the prediction of adolescent schizophrenia from EEG signals [20,27,28].…”

On an Affordable Approach towards the Diagnosis and Care for Prostate Cancer Patients Using Urine, FTIR and Prediction Machines

“…The LSDL, as mentioned, is an AI-driven method towards the signal deconvolution process built upon metaheuristic reasoning, with the original application of the approach based around source separation of a high-frequency acoustic emission signal in order to infer the particle size distribution of a heterogenous powder mixture [ 28 , 29 , 30 , 31 , 32 , 33 ]. Its performance has been observed to surpass that of the wavelet transform in terms of accuracy and computational efficiency [ 28 , 29 , 30 , 31 , 32 , 33 ].…”

“…Following its original application involving source separation and estimation of particle size distribution, the LSDL has seen further applications in case studies that involve brain–machine interface control of bionic upper limb prostheses, prediction of preterm pregnancies from uterine contraction signals, prediction of adolescent schizophrenia from EEG brainwave signals and, more recently, prediction of depth of anesthesia during surgery using frontal cortex neural oscillations [ 27 , 31 , 32 , 33 , 37 , 38 ].…”

“…For the calculation of the optimal threshold region with the LSDL, the Cancer vs. No Cancer datasets were used, obtained using the implementation parameters from Nsugbe et al [ 31 ] and Nsugbe and Sanusi [ 32 ]. The results can be seen in Table 1 and Table 2 .…”

“…The linear series decomposition learner (LSDL) is an artificial intelligence (AI) approach formed in the area of metaheuristics, that is capable of performing a sequence of decompositions on a candidate signal with the use of a linear basis function with the goal of performing signal separation in order to obtain the optimal region of the signal which minimizes uncertainty and redundancy within the signal and maximizes the information quality within the candidate signal [ 23 , 24 , 25 , 26 , 27 ]. The LSDL has seen applications beyond its inception case study, which was based around source separation of an acoustic emission signal, and has seen use in the processing of various other time series in clinical areas spanning rehabilitation, pregnancy and psychiatric medicine, where the inclusion of the LSDL as a signal decomposition tool prior to modeling and prediction exercises was noted to boost the overall prediction accuracy within each case study [ 28 , 29 , 30 , 31 , 32 , 33 ]. An expanded description of the architecture of the LSDL can be seen in Section 2.2 .…”