iiinova — where your digital self keeps evolving

kwnsfk27.r.eu-west-1.awstrack.me

🧵 In Summary Prediction models are not “set and forget.” They must be constantly checked, updated, and re-evaluated—before and after clinical use. With thoughtful updating, we can keep models useful and safe over time. 📄 Read our commentary: kwnsfk27.r.eu-west-1.awstrack.me/L0/https:%2F...

000

🚨 What About After Implementation? Once a model is used in practice, it changes patient outcomes (e.g., high-risk patients get treated and avoid the outcome). This shifts the relationship between risk factors and outcomes. Updating without accounting for this is risky.

100

⚖️ Fairness Under Threat Temporal drift can impact some groups more than others. Even worse, updating models doesn’t guarantee fairness. It can reduce—or worsen—performance in minority groups. We need to monitor algorithmic fairness, not just accuracy.

💻 The Catch? Complexity Dynamic updates need: Continuous data More computing power Digital infrastructure in place Also, no method kept the original performance in later years. So, updating helps—but doesn’t fix everything.

100

In our commentary, we have expanded on the findings of the study, the impacts of temporal drift and the importance of addressing it.

100

A recent study looked at which methods could be effective at addressing this drift doi.org/10.1016/j.bp... They found that without updating, model performance deteriorated and dynamic updating might be able to address this.

doi.org

Redirecting

100

🎯 Why It Matters One form of drift—calibration drift—means predicted risks no longer match actual outcomes. 📉 High-risk patients might be missed. 📈 Low-risk patients might be overtreated. This can harm patients and erode trust in predictive tools.

100

🕰️ What is Temporal Drift? Data that clinical prediction models face when implemented can differ from the data they were trained on, even in the same setting. This can be due to changing patients, treatments or practices The result? The model’s predictions start to go wrong.

122

📊 Key Findings: Presence of psychosis prodrome: 85% in both groups Duration of prodrome: ~18 months in both groups Symptoms at first presentation: no differences Symptoms across the prodrome: no differences

100

Using electronic health records and natural language processing (NLP), we extracted 65 prodromal features (symptoms & substance use) from 1,545 FEP patients. We compared the presence, duration, first presentation and frequency of prodromal symptoms.

Could it be that those missed (DET-) don’t show early warning signs? We tested this by comparing DET- patients to those detected (DET+) by CHR-P services before psychosis onset. Do they share the same prodrome?

100

Early detection of psychosis through CHR-P (clinical high risk for psychosis) services can allow for preventive care — but most patients are not detected before their first episode (FEP). Why?

100

Why are so many people with psychosis missed before their first episode? We studied over 1,500 patients to find out whether those missed by early intervention services show different early signs. Here’s what we found. 🧵👇

Dr Matthew Nour @mattnour.bsky.social · 28/05/2025

Reposted by Dom Oliver

🚨 Job adverts LIVE! 🚨 We want to understand how the brain supports abstract cognition in psychosis, using cutting-edge neuroscience. Non-Clinical RA: lnkd.in/eva-Zzfj Clinical RA: lnkd.in/eSgdad_F Postdoc: lnkd.in/eid8gtbR Oxford University Due June 23

lnkd.in

This link will take you to a page that’s not on LinkedIn

186

Reposted by Dom Oliver

Dr. Sarah Thomas @sarahathomasphd.bsky.social · 04/06/2025

I am hiring a clinical research assistant for a project on adolescent cannabis use that will use MRI 🧠 and behavioral tasks 🎮 More information is available in the link below, and applicants can send a CV and cover letter to me. Please share—thank you! tinyurl.com/msn8h2ce

tinyurl.com

Thomas_TEDS Lab RA Position.pdf

01520

Dom Oliver @domapoliver.bsky.social · 30/05/2025

Thanks to everyone who stopped by to talk about how we are using data to improve mental healthcare

020

Reposted by Dom Oliver

NIHR Biomedical Research Centre: Oxford Health @oxhealthbrc.bsky.social · 30/05/2025

Huge thanks to all who joined our Health Research Showcase at Westgate Oxford yesterday! Great to see so many people engaging with our #research! Read more about the day: tinyurl.com/HRSevent2025

052

Reposted by Dom Oliver

NIHR Biomedical Research Centre: Oxford Health @oxhealthbrc.bsky.social · 29/05/2025

Our Health Research Showcase event is underway! Come join us at Westgate, Oxford and meet our health research experts and learn more about the work they do and how you can get involved! #BePartofResearch

043

Reposted by Dom Oliver

NIHR Biomedical Research Centre: Maudsley @nihrmaudsleybrc.bsky.social · 23/05/2025

Join our next Prediction Modelling presentation with Dr @domapoliver.bsky.social @kingsioppn.bsky.social on assessing bias in a transdiagnostic risk calculator for #psychosis. 🗓️Wednesday 28 May, 15:00 - 16:00, Online www.kcl.ac.uk/events/algor...

022

Dom Oliver @domapoliver.bsky.social · 23/05/2025

Giving a talk next week presenting some new work on algorithmic fairness. Online so come along if you can!

020

Reposted by Dom Oliver

Department of Psychiatry, University of Oxford @oxfordpsychiatry.bsky.social · 12/05/2025

This #MentalHealthAwarenessWeek we asked researchers what one fact they wished more people knew about brain and mental health. #ThisIsMyCommunity #MHAWeek2025 #mentalhealth

142

Reposted by Dom Oliver

NIHR Biomedical Research Centre: Oxford Health @oxhealthbrc.bsky.social · 12/05/2025

#MentalHealthAwarenessWeek 2025 promotes mental well-being and reduces stigma. The theme "Community" emphasises support networks. Our Researchers share key facts about brain and mental health. The campaign educates, encourages open conversations, and fosters understanding and empathy.

252

Reposted by Dom Oliver

NIHR Biomedical Research Centre: Oxford Health @oxhealthbrc.bsky.social · 12/05/2025

Today postdoctoral researcher Dr Dominic Oliver, who works on the Data Science Theme at our BRC, highlights how new predictive tools can help prevent early signs and symptoms of psychosis escalating. Read a review of the programme of research: tinyurl.com/y4nj6ha6 @domapoliver.bsky.social

tinyurl.com

021

Dom Oliver @domapoliver.bsky.social · 01/05/2025

Fantastic news Matthew! Massive, and well deserved, congratulations!!!