ABSTRACT
The behavioural nature of pure-tone audiometry (PTA) limits those who can participate in the test, and therefore those who can access accurate hearing threshold measurements. Event Related Potentials (ERPs) from brain signals has shown limited utility on adult subjects, and a neural response that can consistently be identified as a result of pure-tone auditory stimulus has yet to be identified. The in doing so challenge is worsened by the nature of PTA, where stimulus amplitude decrease to a patient's lower threshold of hearing. We investigate whether EEGNet, a compact Convolutional Neural Network, could help in this domain. We trained EEGNet on a dataset collected whilst patients underwent a test designed to mimic a pure-tone audiogram, then assessed EEGNet performance in the detection task. For comparison, we also trained Support Vector Machines (SVMs) and Common Spatial Patterns + Linear Discriminant Analysis (CSPLDA) on the same task, with the same training paradigms. The results show that EEGNet is capable of detecting hearing events with 81.5% accuracy on unseen participants, outperforming SVMs by just over 5%. Whilst EEGNet outperformed SVMs and CSPLDA, it did not, however, always show a statistically significant improvement. Further analysis of EEGNet predictions revealed that, with sufficient test repetition, EEGNet has the potential to accurately ascertain hearing thresholds. The implication of these results is for a brain-signal based hearing test for those with physical or mental disabilities that limit their participation in a PTA. While this research is promising, future research will be needed to address the complexity of test setup, the duration of testing, and to further improve accuracy.
Subject(s)
Hearing , Neural Networks, Computer , Adult , Humans , Auditory Threshold/physiology , Audiometry, Pure-Tone/methodsABSTRACT
Analysis of heart rate variability (HRV) can reveal a range of useful information regarding the dynamics of the autonomic nervous system (ANS). It is considered a robust and reliable tool to understand even some subtle changes in ANS activity. Here, we study the "hidden" characteristic changes in HRV during visually induced motion sickness; using nonlinear analytical methods, supplemented by conventional time- and frequency-domain analyses. We computed HRV from electrocardiograms (ECG) of 14 healthy participants measured at baseline and during nausea. Primarily hypothesizing evident differences in measures of physiologic complexity (SampEn; sample entropy, FuzzyEn; fuzzy entropy), chaos (LLE; largest Lyapunov exponent) and Poincaré/Lorenz (CSI; cardiac sympathetic activity, CVI; cardiac vagal index) between the two states. We found that during nausea, participants showed a markedly higher degree of regularity (SampEn, p = 0.0275; FuzzyEn, p = 0.0006), with a less chaotic ANS response (LLE, p = 0.0004). CSI significantly increased during nausea compared to baseline (p = 0.0005), whereas CVI did not appear to be statistically different between the two states (p = 0.182). Our findings suggest that motion sickness-induced ANS perturbations may be quantifiable via nonlinear HRV indices. These findings have implications for understanding the malaise of motion sickness and in turn, aid development of therapeutic interventions to relieve motion sickness symptoms.Clinical relevance- The study suggests potential indices of physiologic complexity and chaos that may be useful in monitoring motion sickness during clinical studies.
Subject(s)
Electrocardiography , Motion Sickness , Humans , Heart Rate/physiology , Autonomic Nervous System/physiology , Motion Sickness/etiology , NauseaABSTRACT
Transcutaneous auricular vagus nerve stimulation (taVNS) is a novel neuromodulation application for vagal afferent stimulation. Owing to its non-invasive nature, taVNS is a potent therapeutic tool for a diverse array of diseases and disorders that ail us. Herein, we investigated taVNS-induced effects on neural activity of participants during visually induced motion sickness. 64-channel electroencephalography (EEG) recordings were obtained from 15 healthy participants in a randomized, within-subjects, cross-over design during sham and taVNS conditions. To assess motion sickness severity, we used the motion sickness assessment questionnaire (MSAQ). We observed that taVNS attenuated theta (4-8 Hz) brain activity in the right frontal, right parietal and occipital cortices when compared to sham condition. The total MSAQ scores, and central, peripheral and sopite MSAQ categorical scores were significantly lower after taVNS compared to sham. These findings reveal for the first time the potential therapeutic role of taVNS toward counter-motion sickness, and suggest that taVNS may be reliable in alleviating symptoms of motion sickness in real-time, non-pharmacologically.Clinical relevance- This suggests taVNS potential to offset motion sickness-induced nausea; which may be of translational value to counter e.g., chemotherapy-induced nausea.
Subject(s)
Motion Sickness , Transcutaneous Electric Nerve Stimulation , Vagus Nerve Stimulation , Humans , Motion Sickness/etiology , Motion Sickness/therapy , Nausea , Pilot Projects , Cross-Over StudiesABSTRACT
Perturbations in the autonomic nervous system occur in individuals experiencing increasing levels of motion sickness. Here, we investigated the effects of transauricular electrical stimulation (tES) on autonomic function during visually induced motion sickness, through the analysis of spectral and time-frequency heart rate variability. To determine the efficacy of tES, we compared sham and tES conditions in a randomized, within-subjects, cross-over design in 14 healthy participants. We found that tES reduced motion sickness symptoms by significantly increasing normalized high-frequency (HF) power and decreasing both normalized low-frequency (LF) power and the power ratio of LF and HF components (LF/HF ratio). Furthermore, behavioral data recorded using the motion sickness assessment questionnaire (MSAQ) showed significant differences in decreased symptoms during tES compared to sham condition for the total MSAQ scores and, central and sopite categories of the MSAQ. Our preliminary findings suggest that by administering tES, parasympathetic modulation is increased, and autonomic imbalance induced by motion sickness is restored. This study provides first evidence that tES may have potential as a non-pharmacological neuromodulation tool to keep motion sickness at bay. Thus, these findings may have implications towards protecting people from becoming motion sick and possible accelerated recovery from the malady.
Subject(s)
Autonomic Nervous System Diseases , Motion Sickness , Humans , Autonomic Nervous System/physiology , Electric Stimulation , Heart Rate/physiology , Cross-Over Studies , Healthy VolunteersABSTRACT
BACKGROUND: Process improvement in healthcare is informed by knowledge from the private sector. Skilled individuals may aid the adoption of this knowledge by frontline care delivery workers through knowledge brokering. However, the effectiveness of those who broker knowledge is limited when the context they work within proves unreceptive to their efforts. We therefore need greater insight into the contextual conditions that support individuals to broker process improvement knowledge to the frontline of care delivery, and how policy makers and organizations might generate such conditions. METHODS: Our research took place in a healthcare system within an Australian State. We undertook a qualitative, embedded single case study over the four year period of a process improvement intervention encompassing 57 semi-structured interviews (with knowledge brokers, policy makers, and executive sponsors), 12 focus groups, and 137 h of observation, which included the frontline implementation of actual process improvement initiatives, where knowledge brokering took place. RESULTS: We identified four phases of the process improvement intervention that moved towards a more mature collaboration within which knowledge brokering by improvement advisors began to emerge as effective. In the first phase knowledge brokering was not established. In the second phase, whilst knowledge brokering had been initiated, the knowledge being brokered lacked legitimacy amongst frontline practitioners, resulting in resistance. Only in the fourth and final phase of the intervention did the collective experience of policy makers result in reflections on how they might engender a more receptive context for knowledge brokering. CONCLUSION: We highlight a number of suggested actions that policy makers might consider, if they wish to engender contextual conditions that support knowledge brokering. Policy makers might consider: ensuring they respect local context and experience, by pulling good ideas upward, rather than imposing foreign knowledge from on high; facilitating the lateral diffusion of knowledge by building cultural linkages between people and organizations; strengthening collaboration, not competition, so that trans-organisational flow of ideas might be encouraged; being friend, not foe, to healthcare organizations on their knowledge integration journey. In sum, we suggest that top-down approaches to facilitating the diffusion and adoption of new ideas ought to be reconsidered.
Subject(s)
Administrative Personnel , Knowledge , Humans , Australia , Qualitative Research , Delivery of Health CareABSTRACT
This paper evaluates a range of deep learning frameworks for detecting respiratory anomalies from input audio. Audio recordings of respiratory cycles collected from patients are transformed into time-frequency spectrograms to serve as front-end two-dimensional features. Cropped spectrogram segments are then used to train a range of back-end deep learning networks to classify respiratory cycles into predefined medically-relevant categories. A set of those trained high-performance deep learning frameworks are then fused to obtain the best score. Our experiments on the ICBHI benchmark dataset achieve the highest ICBHI score to date of 57.3%. This is derived from a late fusion of inception based and transfer learning based deep learning frameworks, easily outperforming other state-of-the-art systems. Clinical relevance--- Respiratory disease, wheeze, crackle, inception, convolutional neural network, transfer learning.
Subject(s)
Deep Learning , Humans , Neural Networks, Computer , Respiratory RateABSTRACT
This study examines the neural activities of participants undergoing vibro-motor reprocessing therapy (VRT) while experiencing motion sickness. We evaluated the efficacy of vibro-motor reprocessing therapy, a novel therapeutic technique based on eye movement desensitization and reprocessing (EMDR), in reducing motion sickness. Based on visually induced motion sickness in two sets of performed sessions, eight participants were exposed to VRT stimulation in a VRT/non-VRT setting. Simultaneously, brain activity changes were recorded using electroencephalography (EEG) at baseline and during stimulus exposure, and comparisons made across the VRT/non-VRT conditions. A significant reduction in the alpha (8-12 Hz) spectral power was observed in the frontal and occipital locations, consistent across all participants. Furthermore, significant reductions were also found in the frontal and occipital delta (0.5-4 Hz) and theta (4-8 Hz) spectral power frequency bands between non-VRT and VRT conditions (p < 0.05). Our results offer novel insights for a potential nonpharmacological treatment and attenuation of motion sickness. Furthermore, symptoms can be observed, and alleviated, in real-time using the reported techniques. Clinical relevance - Instead of using drugs to treat motion sickness, patients could safely use this VRT technique.
Subject(s)
Motion Sickness , Motor Disorders , Plastic Surgery Procedures , Electroencephalography , Humans , Motion Sickness/etiology , Motion Sickness/therapy , Problem SolvingABSTRACT
BACKGROUND: Implementation science seeks to enable change, underpinned by theories and frameworks such as the Consolidated Framework for Implementation Research (CFIR). Yet academia and frontline healthcare improvement remain largely siloed, with limited integration of implementation science methods into frontline improvement where the drivers include pragmatic, rapid change. Using the CIFR lens, we aimed to explore how pragmatic and complex healthcare improvement and implementation science can be integrated. METHODS: Our research involved the investigation of a case study that was undertaking the implementation of an improvement intervention at a large public health service. Our research involved qualitative data collection methods of semi-structured interviews and non-participant observations of the implementation team delivering the intervention. Thematic analysis identified key themes from the qualitative data. We examined our themes through the lens of CFIR to gain in-depth understanding of how the CFIR components operated in a 'real-world' context. RESULTS: The key themes emerging from our research outlined that leadership, context and process are the key components that dominate and affect the implementation process. Leadership which cultivates connections with front line clinicians, fosters engagement and trust. Navigating context was facilitated by 'bottom-up' governance. Multi-disciplinary and cross-sector capability were key processes that supported pragmatic and agile responses in a changing complex environment. Process reflected the theoretically-informed, and iterative implementation approach. Mapping CFIR domains and constructs, with these themes demonstrated close alignment with the CFIR. The findings bring further depth to CFIR. Our research demonstrates that leadership which has a focus on patient need as a key motivator to engage clinicians, which applies and ensures iterative processes which leverage contextual factors can achieve successful, sustained implementation and healthcare improvement outcomes. CONCLUSIONS: Our longitudinal study highlights insights that strengthen alignment between implementation science and pragmatic frontline healthcare improvement. We identify opportunities to enhance the relevance of CFIR in the 'real-world' setting through the interconnected nature of our themes. Our study demonstrates actionable knowledge to enhance the integration of implementation science in healthcare improvement.
Subject(s)
Delivery of Health Care , Implementation Science , Delivery of Health Care/methods , Humans , Leadership , Longitudinal Studies , Qualitative ResearchABSTRACT
BACKGROUND: Despite increasing investments in academic health science centres (AHSCs) in Australia and an expectation that they will serve as vehicles for knowledge translation and exchange, there is limited empirical evidence on whether and how they deliver impact. The aim of this study was to examine and compare the early development of four Australian AHSCs to explore how they are enacting their impact-focused role. METHODS: A descriptive qualitative methodology was employed across four AHSCs located in diverse health system settings in urban and regional locations across Australia. Data were collected via semi-structured interviews with 15 academic, industry and executive board members of participating AHSCs. The analysis combined inductive and deductive elements, with inductive categories mapped to deductive themes corresponding to the study aims. RESULTS: AHSCs in Australia are in an emergent state of development and are following different pathways. Whilst varied approaches to support research translation are apparent, there is a dominant focus on structure and governance, as opposed to action-oriented roles and processes to deliver strategic goals. Balancing collaboration and competition between partners presents a challenge, as does identifying appropriate ways to evaluate impact. CONCLUSION: The early stage of development of AHSCs in Australia presents an important opportunity for formative learning and evaluation to optimise their enactment of knowledge mobilisation processes for impact.
Subject(s)
Knowledge , Organizations , Australia , Health Facilities , Humans , Qualitative ResearchABSTRACT
This paper uses the recent glottal flow model for iterative adaptive inverse filtering to analyze recordings from dysfunctional speakers, namely those with larynx-related impairment such as laryngectomy. The analytical model allows extraction of the voice source spectrum, described by a compact set of parameters. This single model is used to visualize and better understand speech production characteristics across impaired and nonimpaired voices. The analysis reveals some discriminative aspects of the source model which map to a physiological class description of those impairments. Furthermore, being based on analysis of source parameters only, it is complementary to any existing techniques of vocal-tract or phonetic analysis. The results indicate the potential for future automated speech reconstruction systems that adapt to the method of reconstruction required, as well as being useful for mainstream speech systems, such as ASR, in which front-end analysis can direct back-end models to suit characteristics of impaired speech.
Subject(s)
Voice Quality , Voice , Humans , Speech Acoustics , Glottis/surgery , Glottis/physiology , Voice/physiology , AcousticsABSTRACT
This paper presents an inception-based deep neural network for detecting lung diseases using respiratory sound input. Recordings of respiratory sound collected from patients are first transformed into spectrograms where both spectral and temporal information are well represented, in a process referred to as front-end feature extraction. These spectrograms are then fed into the proposed network, in a process referred to as back-end classification, for detecting whether patients suffer from lung-related diseases. Our experiments, conducted over the ICBHI benchmark metadataset of respiratory sound, achieve competitive ICBHI scores of 0.53/0.45 and 0.87/0.85 regarding respiratory anomaly and disease detection, respectively.
Subject(s)
Lung Diseases , Humans , Lung , Lung Diseases/diagnosis , Neural Networks, Computer , Respiratory SoundsABSTRACT
OBJECTIVES: We draw on institutional theory to explore the roles and actions of innovation teams and how this influences their behaviour and capabilities as 'institutional entrepreneurs (IEs)', in particular the extent to which they are both 'willing' and 'able' to facilitate transformational change in healthcare through service redesign. DESIGN: A longitudinal qualitative study that applied a 'researcher in residence' as an ethnographic approach. SETTING: The development and implementation of two innovation projects within a single public hospital setting in an Australian state jurisdiction. PARTICIPANTS: Two innovation teams, with members including senior research fellows, PhD scholars and front-line clinicians (19 participants and 47 interviews). RESULTS: Despite being from the same hospital, the two innovation teams occupied contrasting subject positions with one facilitating transformational improvements in service delivery, while the other sought more conservative improvements. Cast as 'IEs' we show how one team took steps to build legitimacy for their interventions enabling spread and scale in improvements and how, in the other case, failure to build legitimacy resulted in unintended consequences which undermined the sustainability of the improvements achieved. CONCLUSIONS: Adopting an institutional approach provided insight into the 'willingness' and 'ability' to facilitate transformational change in healthcare through service redesign. The manner in which innovation teams operate from different subject positions influences the structural and normative legitimacy afforded to their activities. Specifically, we observed that those with the most power (organisational or professional) to bring about transformational change can be the least willing to do so in ways which challenge current practice. Those most willing to challenge the status quo (more peripheral organisation members or professionals) can be least able to deliver transformation. Better understanding of these insights can inform healthcare leaders in supporting innovation team efforts, considering their subject position.
Subject(s)
Delivery of Health Care , Entrepreneurship , Australia , Hospitals, Public , Humans , Qualitative ResearchABSTRACT
Attention-based convolutional neural network (CNN) models are increasingly being adopted for speaker and language recognition (SR/LR) tasks. These include time, frequency, spatial and channel attention, which can focus on useful time frames, frequency bands, regions or channels while extracting features. However, these traditional attention methods lack the exploration of complex information and multi-scale long-range speech feature interactions, which can benefit SR/LR tasks. To address these issues, this paper firstly proposes mixed-order attention (MOA) for low frame-level speech features to gain the finest grain multi-order information at higher resolution. We then combine that with a non-local attention (NLA) mechanism and a dilated residual structure to balance fine grained local detail with convolution from multi-scale long-range time/frequency regions in feature space. The proposed dilated mixed-order non-local attention network (D-MONA) exploits the detail available from the first and the second-order feature attention analysis, but achieves this over a much wider context than purely local attention. Experiments are conducted on three datasets, including two SR tasks of Voxceleb and CN-celeb, and one LR task, NIST LRE 07. For SR, D-MONA improves on ResNet-34 results by at least 29% and 15% for Voxceleb1 and CN-celeb respectively. For the LR task, a large improvement is achieved over ResNet-34 of 21% for the challenging 3s utterance condition, 59% for the 10s condition and 67% for the 30s condition. It also outperforms the state-of-the-art deep bottleneck feature-DNN (DBF-DNN) x-vector system at all scales.
Subject(s)
Natural Language Processing , Neural Networks, Computer , Speech Recognition SoftwareABSTRACT
This paper presents and explores a robust deep learning framework for auscultation analysis. This aims to classify anomalies in respiratory cycles and detect diseases, from respiratory sound recordings. The framework begins with front-end feature extraction that transforms input sound into a spectrogram representation. Then, a back-end deep learning network is used to classify the spectrogram features into categories of respiratory anomaly cycles or diseases. Experiments, conducted over the ICBHI benchmark dataset of respiratory sounds, confirm three main contributions towards respiratory-sound analysis. Firstly, we carry out an extensive exploration of the effect of spectrogram types, spectral-time resolution, overlapping/non-overlapping windows, and data augmentation on final prediction accuracy. This leads us to propose a novel deep learning system, built on the proposed framework, which outperforms current state-of-the-art methods. Finally, we apply a Teacher-Student scheme to achieve a trade-off between model performance and model complexity which holds promise for building real-time applications.
Subject(s)
Lung Diseases , Neural Networks, Computer , Auscultation , Humans , Lung , Lung Diseases/diagnosis , Respiratory SoundsABSTRACT
Objective: To identify processes to engage stakeholders in healthcare Simulation Modeling (SM), and the impacts of this engagement on model design, model implementation, and stakeholder participants. To investigate how engagement process may lead to specific impacts. Data Sources: English-language articles on health SM engaging stakeholders in the MEDLINE, EMBASE, Scopus, Web of Science and Business Source Complete databases published from inception to February 2020. Study Design: A systematic review of the literature based on a priori protocol and reported according to PRISMA guidelines. Extraction Methods: Eligible articles were SM studies with a health outcome which engaged stakeholders in model design. Data were extracted using a data extraction form adapted to be specific for stakeholder engagement in SM studies. Data were analyzed using summary statistics, deductive and inductive content analysis, and narrative synthesis. Principal Findings: Thirty-two articles met inclusion criteria. Processes used to engage stakeholders in healthcare SM are heterogenous and often based on intuition rather than clear methodological frameworks. These processes most commonly involve stakeholders across multiple SM stages via discussion/dialogue, interviews, workshops and meetings. Key reported impacts of stakeholder engagement included improved model quality/accuracy, implementation, and stakeholder decision-making. However, for all but four studies, these reports represented author perceptions rather than formal evaluations incorporating stakeholder perspectives. Possible process enablers of impact included the use of models as "boundary objects" and structured facilitation via storytelling to promote effective communication and mutual understanding between stakeholders and modelers. Conclusions: There is a large gap in the current literature of formal evaluation of SM stakeholder engagement, and a lack of consensus about the processes required for effective SM stakeholder engagement. The adoption and clear reporting of structured engagement and process evaluation methodologies/frameworks are required to advance the field and produce evidence of impact.
ABSTRACT
BACKGROUND: Despite recent significant progress in the development of automatic sleep staging methods, building a good model still remains a big challenge for sleep studies with a small cohort due to the data-variability and data-inefficiency issues. This work presents a deep transfer learning approach to overcome these issues and enable transferring knowledge from a large dataset to a small cohort for automatic sleep staging. METHODS: We start from a generic end-to-end deep learning framework for sequence-to-sequence sleep staging and derive two networks as the means for transfer learning. The networks are first trained in the source domain (i.e. the large database). The pretrained networks are then finetuned in the target domain (i.e. the small cohort) to complete knowledge transfer. We employ the Montreal Archive of Sleep Studies (MASS) database consisting of 200 subjects as the source domain and study deep transfer learning on three different target domains: the Sleep Cassette subset and the Sleep Telemetry subset of the Sleep-EDF Expanded database, and the Surrey-cEEGrid database. The target domains are purposely adopted to cover different degrees of data mismatch to the source domains. RESULTS: Our experimental results show significant performance improvement on automatic sleep staging on the target domains achieved with the proposed deep transfer learning approach. CONCLUSIONS: These results suggest the efficacy of the proposed approach in addressing the above-mentioned data-variability and data-inefficiency issues. SIGNIFICANCE: As a consequence, it would enable one to improve the quality of automatic sleep staging models when the amount of data is relatively small.11The source code and the pretrained models are published at https://github.com/pquochuy/sleep_transfer_learning.
Subject(s)
Neural Networks, Computer , Sleep Stages , Humans , Machine Learning , Polysomnography , SleepABSTRACT
This paper presents a robust deep learning framework developed to detect respiratory diseases from recordings of respiratory sounds. The complete detection process firstly involves front end feature extraction where recordings are transformed into spectrograms that convey both spectral and temporal information. Then a back-end deep learning model classifies the features into classes of respiratory disease or anomaly. Experiments, conducted over the ICBHI benchmark dataset of respiratory sounds, evaluate the ability of the framework to classify sounds. Two main contributions are made in this paper. Firstly, we provide an extensive analysis of how factors such as respiratory cycle length, time resolution, and network architecture, affect final prediction accuracy. Secondly, a novel deep learning based framework is proposed for detection of respiratory diseases and shown to perform extremely well compared to state of the art methods.
Subject(s)
Deep Learning , Neural Networks, Computer , Humans , Respiratory SoundsABSTRACT
OBJECTIVES: To explore patient characteristics recorded at the initial consultation associated with a poor response to non-surgical multidisciplinary management of knee osteoarthritis (KOA) in tertiary care. DESIGN: Prospective multisite longitudinal study. SETTING: Advanced practice physiotherapist-led multidisciplinary orthopaedic service within eight tertiary hospitals. PARTICIPANTS: 238 patients with KOA. PRIMARY AND SECONDARY OUTCOME MEASURES: Standardised measures were recorded in all patients prior to them receiving non-surgical multidisciplinary management in a tertiary hospital service across multiple sites. These measures were examined for their relationship with a poor response to management 6 months after the initial consultation using a 15-point Global Rating of Change measure (poor response (scores -7 to +1)/positive response (scores+2 to+7)). Generalised linear models with binomial family and logit link were used to examine which patient characteristics yielded the strongest relationship with a poor response to management as estimated by the OR (95% CI). RESULTS: Overall, 114 out of 238 (47.9%) participants recorded a poor response. The odds of a poor response decreased with higher patient expectations of benefit (OR 0.74 (0.63 to 0.87) per 1/10 point score increase) and higher self-reported knee function (OR 0.67 (0.51 to 0.89) per 10/100 point score increase) (p<0.01). The odds of a poor response increased with a greater degree of varus frontal knee alignment (OR 1.35 (1.03 to 1.78) per 5° increase in varus angle) and a severe (compared with mild) radiological rating of medial compartment degenerative change (OR 3.11 (1.04 to 9.3)) (p<0.05). CONCLUSIONS: These characteristics may need to be considered in patients presenting for non-surgical multidisciplinary management of KOA in tertiary care. Measurement of these patient characteristics may potentially better inform patient-centred management and flag the need for judicious monitoring of outcome for some patients to avoid unproductive care.
Subject(s)
Osteoarthritis, Knee , Physical Therapists , Humans , Knee Joint/diagnostic imaging , Longitudinal Studies , Osteoarthritis, Knee/therapy , Prospective StudiesABSTRACT
Given the pace of technological advancement and government mandates for healthcare and system transformation, there is an imperative for change. Health systems are highly complex in their design, networks and interacting components, and experience demonstrates that change is very challenging to enact, sustain and scale. Policy-makers, academics and clinicians all need better insight into the nature of this complexity and an understanding of the evidence-base that can support healthcare improvement (HCI), or quality improvement, interventions and make them more effective in driving change. The evidence base demonstrates the vital role of clinical engagement and leadership in HCI, and it is imperative that clinicians engage to improve front-line healthcare. The literature on HCI is vast, applies different and inconsistent terminology and encompasses often loosely defined and overlapping concepts. An increasingly broad range of disciplines has contributed to the available evidence base, but often discipline-specific perspectives frame these contributions. Available literature can also be overly driven by the generation of theoretical concepts and the advancement of academic understanding. It does not necessarily primarily provide focussed and pragmatic insights to guide and inform frontline practice. We aim to address these issues by summarising theories, frameworks, models and success factors for improvement in complex health systems to assist clinicians and others to engage and lead change. We integrate the field of HCI into the learning health system highlighting the key role of the clinician. We seek to inform stakeholders; clinicians and managers to guide the planning, enacting, sustaining and scaling of HCI.