Current image processing techniques have reached a high level of sophistication and allow the extraction of extensive information from medical imaging data. The problem is that most of these modern post processing techniques are not applied in a clinical setting, because the software developed by scientists is difficult to use, often does not run on operating systems used by clinicians, and requires extensive hardware resources. The integration of new post processing techniques by vendors often takes many years.
One solution to bring modern image processing into the clinic would be to take the medical data outside of the clinic and utilize a powerful cloud instance where all tools are installed and the clinician can upload the data via a simple graphical user interface. The problem however is that medical data contains sensitive information that cannot be easily removed, such as facial features in magnetic resonance imaging data of the brain. The goal is to create a platform-independent tool that is easy to use, reads standard medical DICOM data, anonymizes the data and uploads the data to a cloud instance to start the image processing in the cloud and sends back results.
Currently no software exists that combines these tasks.
Shu Ngo, Frederik Steyn, Anjali Henders, Fleur Garton, Naomi Wray, Anna Vinkhuyzen
Motor Neuron Disease (MND) is a rare and fatal neurodegenerative disease. Average life expectancy is just 2.5 years after diagnosis but this can range from 1 to 15 years. Symptoms vary among patients, but generally include muscle weakness, muscle twitching, and muscle cramping, difficulty with speaking and swallowing, breathing problems, and paralysis. Many patients experience mental health problems.
Despite no available cure, medical advances in care mean that quality of life can be extended through appropriate, individualised multi-disciplinary care. Due to the heterogeneous nature of the disease, timing of access and type of care needed varies from patient to patient.
MND is a neurodegenerative disease and patient’s care is overseen by a specialised neurologist. To improve care, neurologists need to collect individual-level information to personalise relevant allied health care required for each patient. There is a number of reasons a neurologist may fail to collect sufficient information from the patient: patients may not be able to report all relevant symptoms and disease progression to the neurologist during consultation; patients face physical challenges to visit the neurologist; neurologists have time restrictions during consultation hours; patients’ visits to the neurologists are infrequent; the neurologist not being able to structurally record relevant symptoms and disease progression.
Patient care can only be optimal when information about relevant symptoms and disease progression is sufficiently available to the neurologist. Currently, neurologists are lacking relevant information.
We propose an application with a multi-devise interace called ShowMndE. The application is a patient-centered solution to collect data regularly– anywhere, anytime. Use of the application avoids the typical hurdles that impact regular disease tracking (remote locations, impaired patient mobility, hospital visitation cost) but will ensure appropriate support is accessed at the right time.
Our vision is for an application that can be used to track an individual’s disease progression. Patients self-report on a fortnightly basis using validated questionnaires and symptom related questions. The self-report data are scored for various symptoms associated with MND so that any score (or rate of change) can alert the overseeing neurologist and/or help record between clinic visits.
In this manner, the application provides an interface for individual data-analyses and feedback for both the patient and the neurologist. Based on this information, the appropriate healthcare professional can be accessed to ensure quality of life is maximised. Realisation of this novel, precision medicine approach to MND could directly support a reduction of the burden of disease (suffering and premature death) and extend quality of life.
Currently there is no other solution than neurologists doing best they can to collect as much relevant information as possible.
Currently, there is fragmented presentation of perinatal data in the maternity sector, making it difficult for consumers to access and understand the information
Having searchable fields of maternity data, and infographics to aid in understanding of the data, would enable consumers to make informed choices about their place of birth and overall maternity care. We know the place of birth has the most impact on a woman’s outcome from birth. There is good evidence supporting the proposition that variances we see in maternity care are often unwarranted. Access to data that is easy to understand and easy to navigate will not just be of benefit to consumers, but also valuable for clinicians to improve the safety and quality of healthcare.
A searchable, user-friendly website that allows comparisons between maternity facilities and state or national averages, whilst also displaying the data as infographics.
There are no current solutions. Women often don’t know where to find the data, how to interpret it or what it means.
Biomedical higher degrees research (HDR) students at GRIDD need to meet ongoing candidature milestones, manage their project, interact with their supervisors, peers and collaborators, develop excellent written and oral communication skills, and develop networks and a broad range of professional skills that they can use during their candidature and in their future careers as biomedical/health professionals. A crucial aspect to succeeding as HDR candidates is a supportive environment, access to career development opportunities, strong mentoring and a vibrant HDR culture. Various mentoring opportunities exist for GRIDD HDR candidates (e.g. seminars/workshops, supervisor mentoring etc) and we would like to build on this through an e-mentoring approach that aims to both inform and inspire our HDR candidates, contributing to a vibrant HDR culture.
Our vision is a database of >100 mentoring tips (written tips, 20sec video tips from current/former students and supervisors, links to websites, TED talks, papers, university courses/workshops etc) that can be delivered to HDR candidates in an engaging format. Possible features:
- Random delivery of tips once, twice or three times a week, depending on HDR candidate preference
- Delivery of tips via email or via a pop up that appears on their monitor
- Delivery of tips in an engaging biomedical themed format – for example text appearing as a label on a flask, computer monitor etc.
- Inbuilt option to unsubscribe or change frequency of tips
- Inbuilt automatic simple survey to monitor project success – eg after 25, 50 and 100 tips have been delivered
- Inbuilt function to assess “clicks” or tip access – to monitor project
- Inbuilt functionality to allow additional tips to be added/modified by an administrator
- Inbuilt functionality to allow sub-sets of tips to be directed to certain subscribers – e.g. to allow future expansion to different users and tailored tips to be delivered (e.g. biomedical health versus allied health versus biostatistics versus physics etc)
Ad hoc emails from supervisors, HDR convenors, Griffith Graduate Research School, supervisors and other sources; HDR candidates seeking out mentoring opportunities themselves
Could a social robot be a new treatment tool in the health and wellbeing process?
I want to improve the ability of social robots to interact with humans so that they can support people taking part in healthcare programs.
Health and wellbeing programs that can support and motivate someone to achieve a health goal are in high demand. Health professionals use different technology tools to help improve the treatment process, such as web-based applications or wearables. A social robot that can have a conversation with someone about their health could become a new treatment delivery method to help encourage and motivate someone to achieve their health goal. However to do this, social robots need to be able to interact effectively with people in a conversation-based treatment program. This is important because people can respond to conversation topics and questions in unique and unpredictable ways.
A problem that is yet to be solved is the design of a personality or behaviour set for a social robot that can effectively interact and communicate with individuals, as well as encourage them to generate and sustain motivation to change in a treatment program session, especially beyond the initial novelty period of interacting with a social robot.
The ideal solution would be the design and establishment of a standard set of robot verbal and non-verbal communication skills (e.g. paraphrasing, gestures, eye contact) that the robot could perform in different health programs to help encourage individuals to reach their personal health and wellbeing goal.
A current solution involves the use of stock-standard communication behaviours. Other current solutions involve the use of extensive branching to help capture and respond to more predictable responses, or the use of a human operator that remotely controls the robots speech and behaviours.
banner image courtesy of Dr Nick Hamilton