Tag: data science

How Smartwatches Could Help People with Type 1 Diabetes 

Posted on by kerry.turcsi

JGI Seed Corn Funding Project Blog 2023/24: Miranda Armstrong

Introduction

Type 1 diabetes (T1D) requires consistent self-management, which places a large burden on those who live with it. We explored the role smartwatches could play in reducing that burden. 

Image contains photos of a continuous glucose monitor, smartwatch, closed loop algorithm and Insulin pump
Figure 1: Theoretical closed-loop system that uses smartwatch data in its algorithm. Closed-loop systems without smartwatch integration are the current state of the art of T1D technology. They begin to automate the T1D management process by using data from the eco-system of devices to predict future changes in blood glucose and change insulin dosage to counteract these changes.

Aims

The project aimed to collect and build a dataset that would allow for exploration into the potential of smartwatches in T1D management. This would include both data from the smartwatch and T1D technology the participants used, and user experience of using the smartwatch alongside their typical T1D management. To meet the aim, the following goals were set: 

  1. Collect data from participants, including from smartwatches and T1D devices, and in interviews and focus groups. 
  1. Clean, anonymise, and combine data from different sensors into a consistent format, and transcribe the interviews and focus groups. 
  1. Hold an online data challenge using a sample of the collected data to promote the dataset and highlight potential uses for it. 
  1. Release the dataset publicly to allow other researchers to use it as part of their work and therefore increase the value of the dataset. 

What was achieved

Two graphs. Top graph is Blood glucose, Insulin and Carbohydrates against Time. The bottom graph is Heart rate and steps against time
Figure 2: An example day of data from one participant, with some of the data available. The upper axis highlights the data available to current commercial closed-loop systems and the lower axis shows some smartwatch data from the same period. 

Data Collection

The project recruited 24 participants, and each were given a smartwatch or could use their own. Over six months, participants donated data from their smartwatches and type 1 diabetes (T1D) devices to create a dataset aimed at exploring the integration of smartwatch data into a closed-loop algorithm. This dataset reflects real-life conditions and participants used a range of T1D technology. Over 2000 days, the data that was donated had a high coverage from all the devices the participant used. During this time participants were involved in interviews and focus groups to discuss their opinions of the smartwatches and potential roles they could see in T1D management. A total of 62 interviews and 11 focus groups were completed across the study period. 

Data Processing

We processed a large amount of data to prepare it for public use. The smartwatch and T1D data were cleaned and anonymised (so no one involved in the study could be identified) and then organised into two formats. One was an easy-to-use dataset for researchers to test their algorithms, and the other kept the data in its original form for deeper exploration. We also transcribed and anonymised the interviews and focus groups so other researchers could analyse them to understand the participants’ experience of using the smartwatch. 

Initial Findings

Initial engagement with the interviews and focus group data has highlighted several potential uses for smartwatches in T1D management. These include as a device to display data quickly and discretely to the user, as an interface with T1D technology for easier access, and as a data source to inform management decision making around activity. There are also design implications highlighted in this analysis. These include utilising automation to provide benefit without increasing user burden, allowing customisation to accommodate the wide range of user preferences and usage patterns to promote uptake, and flexibility to allow these systems to adapt to changing user needs and ensure use of the device into the future. 

Future Plans

The data challenge and the public release of the dataset are scheduled for later this year.  We plan to run the competition from mid-September to the end of November 2024, with £1600 in prize vouchers available across entries. If you would like to hear more details about the competition, please leave your details in this form. The whole dataset will then be published after the competition of the data competition.  

Additionally, we will conduct our own analysis on the data that has been collected. This will expand our initial findings that highlight where and how a smartwatch could be used to improve T1D management. It will also test if adding smartwatch data can improve the prediction of blood glucose, by factoring in information on activity. This could be utilised in closed-loop systems (Figure 1) and would allow them to factor activity into their algorithms. For example, if the user was to go for a walk, this system could detect that activity and then predict the drop in blood glucose levels it would cause and so reduce insulin delivery to counteract this drop. Such a system would improve T1D management and reduce the burden placed on those managing it. 

Contact details

Sam James: sam.james@bristol.ac.uk 

Miranda Armstrong: Miranda.armstrong@bristol.ac.uk 

Zahraa Abdallah: zahraa.abdallah@bristol.ac.uk 

Ask JGI Student Experience Profiles: Rachael Laidlaw

Posted on by kerry.turcsi

Rachael Laidlaw (Ask-JGI Data Science Support 2023-24) 

I first came into contact with the Jean Golding Institute last year at The Alan Turing Institute’s annual AI UK conference in London, and then again in the early stages of the DataFace project in collaboration with Cheltenham Science Festival. This meant that before I officially joined the team back in October, I already knew what a lovely group of people I’d be getting involved with! Having nice colleagues, however, was not my only motivation for applying to be an Ask-JGI student. On top of that, I’d decided that whilst starting out in my ecological computer-vision PhD niche, I didn’t want to forget all of the statistical skills that I’d developed back in my MSc degree. Plus, it sounded really fun to keep myself on my toes by exercising my mind tackling a variety of data-oriented requests from across the university’s many departments. 

Rachael Laidlaw in centre with two JGI staff members to the left and one JGI staff member to the right pointing towards a Data pin board at the JGI stall
Rachael Laidlaw (centre), second-year PhD student in Interactive Artificial Intelligence, and other JGI staff members at the JGI stall

During the course of my academic life, I’ve taken the plunge of changing disciplines twice, moving from pure mathematics to applied statistics and then again to computer science, and I liked the idea of supporting others to potentially do the same thing as they looked to enhance their work by delving into data. Through Ask-JGI, I kept my weeks interesting by having something other than my own research to sometimes switch my focus to, and it felt very fulfilling to be able to offer useful technical advice to those who were in the same position that I myself had been in not so long ago too! I therefore got stuck in with anything and everything, from training CNNs for rainfall forecasting or performing statistical tests to compare the antibiotic resistance of different bacteria, to modelling the outcomes of university spinouts or advising on the ethical considerations and potential bias present when designing and deploying a questionnaire-based study. And, of course, by exposing myself to these problems (alongside additional outreach initiatives and showcase events), I also learned a lot along the way, both from my own exploration and from the rest of the team’s insights. 

One especially exciting query revolved around automating the process of identifying from images which particular underground printing presses had been used to produce various historical political pamphlets, based on imperfections in the script. This piqued my interest immediately as it drew parallels with my PhD project, highlighting the copious amount of uses of computer vision and how it can save us time by speeding up traditionally manual processes: from the monitoring of animal biodiversity to carrying out detective work on old written records. 

All in all, this year has broadened my horizons by giving me great consultancy-style work experience through the opportunity to share my expertise and help a wide range of researchers. I would absolutely encourage other curious PhD students to apply and see what they can both give to and gain from the role! 

Children of the 90s and Synthetic Health Data 

Posted on by kerry.turcsi

JGI Seed Corn Funding Project Blog 2023/24: Mark Mumme, Eleanor Walsh, Dan Smith, Huw Day and Debbie Johnson

What is Children of the 90s? 

Children of the 90s (Co90s) is a multi-generational population-based study following the health and development of nearly 15,000 families living around Bristol, whose children were born in 1991 and 1992. 

Co90s initially recruited its participants during the early stages of the mum’s pregnancy and captures information prospectively, at key time points, using self-reported questionnaires, interviews, clinics and electronic health records (EHR). 

The Co90s supports about 20 project teams using NHS data at any one time.  

What is Synthetic Data? 

At its most basic, synthetic data is information generated artificially rather than recorded directly from real-world events. It is essentially a computer-generated version of the data that doesn’t contain any real data and preserving privacy and confidentiality. 

Privacy vs Fidelity

Generating synthetic data is frequently a balancing act between fidelity and privacy (Figure 1). 

“Fidelity”: how well does the synthetic data represent the real-world data?  

“Privacy”: can personal information be deduced from the synthetic data? 

Blue line with an arrowhead at each end. Left side is High privacy, low fidelity and the right side is low privacy, high fidelity
Figure 1: Privacy versus fidelity

Why synthetic NHS data: 

EHR data are incredibly valuable and rich data sources, yet there are significant difficulties to accessing this data, including financial costs and the time taken to complete multiple application forms and have these approved. 

Because the authentic NHS data is so difficult to access, it is also not unusual for researchers to have never worked with, or possibly even seen, this type of data before. They often face a learning curve to understand how the data is structured, what variables are present in the data and how those variables relate to each other. 

The journey for a project to travel (Figure 2) just to get NHS data typically goes through the following stages: 

Multiple coloured boxed with each stage a project has to go through to get NHS data from initial grant application to data access
Figure 2: The stages a project goes through to get NHS data

Each of these stages can take several months and are usually sequential. It not unheard of for projects to run out of time and/or money due to these lengthy timescales. 

Current synthetic NHS data: 

Recently, the NHS has released synthetic Hospital Episode Statistics (HES) data (available here; https://digital.nhs.uk/services/artificial-data) which is, unfortunately, quite limited for practical purposes. This is because a very simple approach was adopted; each variable is randomly generated independently from all others. While it is possible to infer broadly accurate descriptive statistics for single variables (e.g., age or sex), it is impossible to infer relations between variables (e.g., how the number of cancer diagnoses increases with age). In the terms introduced above, it has high privacy but low fidelity. As shown in the heatmap, Figure 3, we observe practically no association between diagnosis and treatment because synthetic NHS data is randomly generated variable-by-variable.  

Heat map with disease groupings on the right side and different treatments on the bottom
Figure 3: Heatmap displaying the relations between disease groupings (right side) and treatment (bottom) from the synthetic NHS data. The colour shadings represent the number of patients (e.g., the darker the shading, the higher the number). The similarity in shading within each diagnosis row shows that treatment and diagnosis were largely independent in this synthetic dataset.

What do researchers want from synthetic data? 

We developed an anonymous survey and asked 230 researchers experienced with EHR data, what would be important to them when considering using synthetic EHR data. Out of the 24 responding most were epidemiologists at fellow or professor level. Researchers were then invited to an online discussion group to expand on insights from the survey. Seven researchers attended.   

Most researchers had a more than 3 years of experience using EHRs both within and outside of cohort studies. Although few had much knowledge of synthetic EHR data, many had heard of synthetic EHR data and were interested in its application, particularly as a tool for training and learning about EHRs generally. 

The most important issues to researchers (Figure 4) were consistent patient details and having all the additional diagnosis & treatment codes rather than just the main ones: 

Horizontal bar chart showing different desirable quantities in synthetic EHR against the number of responses.
Figure 4: What researchers look for in synthetic EHRs

The most important utility for these researchers was to test/develop code and understand broad structure of the data, as shown below (Figure 5): 

Chart showing the priorities of researchers on a scale from first to last choice
Figure 5: Priorities of researchers when using synthetic data

This was reflected in their main concerns about maintaining the utility of the data in the synthetic version by producing high level of accuracy and attention to detail. 

During the discussion it was recognised that EHRs are “messy” and synthetic data should emulate this, providing an opportunity to prepare for real EHRs. 

Visual showing discussion points about emulate "messy" real data
Emulate “messy” real data discussion visual

Being able to prepare for the use of real EHRs was the main use case for synthetic data. No one suggested using the synthetic data as the analysis dataset in place of the real data.   

Visual showing factors to consider in relation to preparation for using real EHR data
Preparation for using real EHR data visual

It was suggested, in both survey responses and the discussion group, that any synthetic data should be bespoke to the requirements of each project. Further, it was observed that each research project only ever used a portion of the complete dataset, therefore synthetic data should be minimized also.  

“I think any synthetic data set based on any of the electronic health records should be stripped back to the key things that people use, because then the task of making it a synthetic copy [is] less.” (online participant) 

Summary

Following the survey and discussion with some researchers familiar with EHRs a few key points came through: 

  • Training – using synthetic data to understand how EHRs work, and to develop code. 
  • Fidelity is important – using synthetic data as way for researchers to experience using EHRs (e.g. the real data flaws, linkage errors, duplicates). 
  • Cost – the synthetic data set, and associated training, must be low cost and easily accessible.  

Next Steps

There is a demand for a synthetic data set with a higher level of fidelity than is currently available, and particularly there is a need for data which is much more consistent over time. 

The Co90s is well placed to respond to this demand, and will look to: 

Ask JGI Student Experience Profiles: Mike Nsubuga

Posted on by kerry.turcsi

Mike Nsubuga (Ask-JGI Data Science Support 2023-24) 

Embarking on a New Path 

Mike Nsubuga
Mike Nsubuga, first year PhD Student in Computational Biology and Bioinformatics

In the early days at Bristol, even before I began my PhD, I stumbled upon something extraordinary. AskJGI, a university initiative that provides data science support to researchers from all disciplines, caught my attention through a recruitment advert circulated by my PhD supervisor for support data scientists.

My journey started with hesitation. As a brand-new PhD student, who had just relocated to the UK, I questioned whether I was ready or suitable for such a role. Despite my reservations, my supervisor saw potential in me and encouraged me to seize this opportunity. Yielding to their encouragement, I applied, not fully realizing then how this decision would profoundly shape both my academic and professional paths. 

A World of Opportunities 

Joining AskJGI opened a door to a dynamic world brimming with ideas and innovations. My background in bioinformatics and computational biology meant that working on biomedical queries was particularly rewarding. These projects varied from analyzing protein expression data to studying infectious diseases, allowing me to use data science in meaningful ways. 

Among the initiatives I was involved in was developing models to predict protein production efficiency in cells from their genetic sequences. Our goal was clear yet impactful: to identify patterns in genetic sequences that indicate protein production efficiency. We employed advanced data analysis and machine learning techniques to achieve effective predictions. 

Additionally, I contributed to a project analyzing the severity of dengue infections by using statistical models to identify key biological markers. We pinpointed certain markers as critical for distinguishing between mild and severe cases of the infection. 

These projects showcased the transformative power of data science in understanding and potentially managing diseases, directly impacting public health strategies. 

Making Science Accessible: Community Engagement at City Hall

A highlight of my tenure with AskJGI was participating in Data Science Week at bustling Bristol City Hall. The event was not merely a showcase of data science but an opportunity to demystify complex concepts for the public. Engaging in lively discussions and simplifying intricate algorithms for curious visitors was incredibly fulfilling, especially seeing their excitement as they understood the concepts that are often discussed in our professional circles. 

Audience sitting in City Hall. Some audience members are raising there hand. There is a projector and a speaker at the front of the hall
AI and the Future of Society event as part of Bristol Data Week 2024

Fostering Connections and Gaining Insights 

AskJGI enhanced my technical skills and broadened my understanding of the academic landscape at the University of Bristol. The connections I forged were invaluable, sparking collaborations that would have been unthinkable in the more isolated environment of my earlier academic career. Reflecting on my transformative journey with AskJGI, I am convinced more than ever of the importance of interdisciplinary collaboration and the critical role of data science in tackling complex challenges. I encourage any researcher at the University of Bristol who is uncertain about their next step to explore what AskJGI has to offer. For PhD students looking to get involved, it represents not just a learning opportunity but a chance to make a significant societal impact. 

24/25 Ask-JGI Application

Unlocking big web archives: a tool to learn about new economic activities over space and time

Posted on by kerry.turcsi

JGI Seed Corn Funding Project Blog 2022/23: Emmanouil Tranos 

Where do websites go to die? Well, fortunately they don’t always die even if their owners stop caring about them. Their ‘immortality’ can be attributed to organisations known as web archives, whose mission is to preserve online content. There are quite a few web archives today with different characteristics – e.g. focusing on specific topics vs. archiving the whole web – but the Internet Archive is the oldest one. Even if you are not familiar with it directly, you might have come across the Wayback Machine, which is a graphical user interface to access webpages archived by the internet archive.  

Although it might be fun to check the aesthetics of a website from the internet’s early days – especially considering the current 1990s revival – one might question the utility of such archives. But some archived websites are more useful than others. Imagine accessing archived websites from businesses located in a specific neighborhood and analysing the textual descriptions of the services and products these firms offer as they appear on their websites. Imagine being able to geolocate these websites by using information available in the text. Image doing this over time. And, image doing this programmatically for a large array of websites. Well, our past research did that and, therefore, serves as a proof-of-concept for the utility of web archives in understanding the geography of economic activities. Our models were successful in utilising a well-curated by The British Library and the UK Web Archive data set to understand how a well-known tech cluster – that is Shoreditch in London – evolved over time. Importantly, we were able to do this at a much higher level of detail in terms of the descriptions of the types of economic activities than if we had used more traditional business data.  

The JGI project provided the opportunity to start looking forward. Our proof-of-concept research was useful in validating the value of such a research trajectory and revealing the evolving mechanisms of economic activities as we only focused on the 2000-2012 period. The next question is how to use this research framework in a current context.  

Before I explain the challenge in doing this, let me tell you about the value of being able to do this. Our current understanding of the typologies of economic activities is based on a system called Standard Industrial Classification (SIC) codes. Briefly, businesses need to choose the SIC code that describes best what they do. Useful as they may be, SIC codes have not been updated since 2007 and, therefore, cannot capture new and evolving economic activities. In addition, there is built-in ambiguity in SIC codes as quite a few of them are defined as “… not elsewhere classified” or “… other than …”. Having a flexible system that can easily provide granular and up-to-date classifications of economic activities within a city or a region can be very useful to a wide range of organisations including local authorities, chambers of commerce and sector-specific support organisations.  

The main challenge of building such a tool is data in terms of finding, accessing, filtering and modelling relevant data. Our JGI seedcorm project together with Rui Zhu and Giulia Occhini allowed us to pave the path for such a research project. Thanks to the Common Crawl, another web archive which offers all its crawled data openly every two months, we have all the data we need. The problem is that we have much more data than what we need as the Common Crawl crawls and scrapes the whole web providing a couple of hundred of terabyte of data every two months. And that is in compressed format! So, only accessing these data can be challenging set aside building a workflow which can do all the steps I mentioned above and – importantly – keep on doing these steps every few months once new data dumps become available.  

Although we are nowhere close to completing such a big project, the JGI seedcorn funding allowed us to test some of the code and the data infrastructure needed to complete such a task. We are now developing funding proposals for such a large research programme and although a risky endeavour, we are confident that we can find the needle in the haystack and build a dynamic system of typologies of economic activities at a level of detail higher than current official and traditional data offer, which is based on open data and reproducible workflows.  

Emmanouil Tranos 

Professor of  Quantitative Human Geography | Fellow at the Alan Turing Institute  

e.tranos@bristol.ac.uk | @EmmanouilTranos | etranos.info | LinkedIn