ki:elements

Tomorrow, we will bring a talk to the 34th Alzheimer Europe Conference, in Geneva! 🇨🇭 Sofia Alexopoulou will present her work on assessing neuropsychiatric symptoms through speech analysis in early dementia patients. 💡 This presentation is based on data collected during the DeepSpA project, a partnership with Inria, Maastricht University, and Janssen (now Johnson & Johnson Innovative Medicine). Who else is in Geneva for this conference? 😎

🚨 New article published! 🚨 In this one, we analyzed the user experience of automated and semi-automated phone-based cognitive assessments. 📱 Participants rated the user experience of Mili Phone above averaged in all modalities. 🙌 Here’s the approach we followed for this study: 👇 ➜ 𝐒𝐚𝐦𝐩𝐥𝐞: 67 participants with early neurocognitive disorders (SCD and MCI). ➜ 𝐏𝐡𝐨𝐧𝐞-𝐛𝐚𝐬𝐞𝐝 𝐯𝐬 𝐟𝐚𝐜𝐞-𝐭𝐨 𝐟𝐚𝐜𝐞 𝐚𝐬𝐬𝐞𝐬𝐬𝐦𝐞𝐧𝐭𝐬: - Dutch 15-item Verbal Learning test (VLT). - 1-min Semantic Verbal Fluency Test (SVF). - Digit span from the WAIS-IV. - Free speech test. - Geriatric Depression Scale 15 items (GDS-15). - Subjective Cognitive Functioning scale (SCF). ➜ 𝐃𝐚𝐭𝐚 𝐜𝐨𝐥𝐥𝐞𝐜𝐭𝐢𝐨𝐧 𝐭𝐢𝐦𝐞𝐥𝐢𝐧𝐞: November 2019 to May 2022. ➜ 𝐌𝐚𝐢𝐧 𝐟𝐢𝐧𝐝𝐢𝐧𝐠𝐬: ✅ Above average user experiences.  ✅ Face-to-face ratings were comparable to the ones of the semi-automated phone-based assessment (except for the satisfaction and recommendation items, which were lower for the later). You can read the full article here (it’s open access!): https://lnkd.in/ef99NMgY This work is a collaboration with Maastricht University and The Janssen Pharmaceutical Companies of Johnson & Johnson. 🤝 Thank you to Alexandra König and Johannes Tröger, from ki:elements, for joining efforts with me on this one! 🙏

Features extracted from free speech can detect fatigue. ✅ In our latest article, we evaluated the efficacy of automated speech analysis in detecting fatigue in #MultipleSclerosis patients. 🔍 ➜ 297 participants performed a comprehensive speech assessment on Mili App, via a tablet. For validation purposes, our speech analysis focused only on three speech tasks: ▪ Positive storytelling. 😊 ▪ Negative storytelling. 🥺 ▪ Picture description. 🖼️ ➜ 𝐌𝐚𝐢𝐧 𝐟𝐢𝐧𝐝𝐢𝐧𝐠𝐬: ✅ Free speech patterns can be useful in detecting fatigue.  ✅ Speech ratio might be a specific marker of fatigue, particularly cognitive fatigue. ✅ Automated speech analysis of a free speech task offers an objective, valid and low-burden method for fatigue assessment. Full article available here: https://lnkd.in/e978ntyt This research is a collaboration with Technische Universität Dresden, Roche, University of Glasgow, Université Côte d'Azur, and The Philipp University of Marburg. 🤝 Thank you to our team at ki:elements, specially to Marcelo D., for all the great work on this paper! 🙌

This weekend, Felix Menne and I will be in Milan for #ECNP24. 💪 This time, ki:elements is represented with 3 posters: ➜ 𝐏𝐨𝐬𝐭𝐞𝐫 𝟏: ❝ Two-minute automated speech analysis: comparable performance to Beck Depression Inventory-II in distinguishing depressed and healthy individuals  ❞ This study aimed to identify speech features able to differentiate between persons with major depressive disorder and healthy ones, based on a symptom severity measure. ➜ 𝐏𝐨𝐬𝐭𝐞𝐫 𝟐: ❝ Speech pauses: biomarkers for differentiating depressed patients from healthy controls ❞  This project investigated which speech features differ between healthy participants and depressed patients, and studied the extent to which depression and its symptoms affect the speech of patients. ➜ 𝐏𝐨𝐬𝐭𝐞𝐫 𝟑: ❝ Automated analysis of linguistic features using verbal fluency tasks in psychiatric disorders: results from the PRONIA study ❞ This research focused on classifying diagnostic groups using automated linguistic features from VFTs in combination with classical cognitive measures. Thank you to our collaborators at Uniklinik RWTH Aachen, Carl von Ossietzky University of Oldenburg, and University of Cologne.

This weekend, ki:elements will bring 3 posters to #ECNP24 in Milano, Italy 🤌 Felix Menne will present our latest data on the role of (bio)markers in distinguishing depressed and healthy individuals. This data results from two of our recent projects: ▪ EmoCon, a collaboration with RWTH Aachen University. ▪ PsyApp, a project in partnership with Carl von Ossietzky University of Oldenburg. Millennia Chakraborty, from University of Cologne, will share results from our collaboration on the PRONIA study, which targets psychiatric disorders. Let’s go team! 🙌

This week, you can find us at ECTRIMS 2024 Congress, in Denmark 🇩🇰 Marcelo D., from ki:elements, will share our results on the automatic detection of fatigue in multiple sclerosis patients. Our collaborator, Riley Bove from University of California, San Francisco, will present two posters on our joint work in identifying key speech features for multiple sclerosis. This research had the invaluable contribution of Tjalf Ziemssen from Dresden International University (DIU) and it was supported by Roche (Intonate) and the National MS Society. 🙏 See you there? 😉

🎤 Breaking News from the Neuro-Frontlines! 🧠🗣️ 🤓 Attention, CNS enthusiasts! Our brand-new paper on speech analysis as a means of detecting fatigue in Multiple Sclerosis (MS) is here — and it's anything but dry reading! We all know that fatigue is the invisible (yet ever-present) symptom that challenges many people with MS day in and day out. Until now, we've had to rely on subjective reports and questionnaires. The universe called for objective measurement tools. Spoiler alert: We delivered! In an exciting collaboration with ki:elements and Roche we have initiated and conducted a pioneering study. The result? By analyzing spontaneous speech, we isolated parameters that open up some thrilling new avenues for assessing fatigue. In other words, we cracked the "language of tiredness," and it has quite a lot to say! 🥳 MSSPEECH Project: A Sea of Possibilities! Thanks to the support of the BMBF and in collaboration with KI Elements, we're continuing our course on MSSPEECH. We're developing tools that will hopefully enable speech analyses in clinical practice soon. And yes, we firmly believe in digital outcomes! Language can be an amazing conduit (pun intended 😉) as it encapsulates many different functions. Paper Summary in a Nutshell: 🧠 MS Basics: Multiple Sclerosis, a chronic neuroinflammatory disease, leads to demyelination and axonal degeneration in the central nervous system. Fatigue affects a major portion of MS patients, significantly impairing their quality of life. 📊 The Challenge: Despite its prevalence, measuring fatigue is tough and the underlying mechanisms are poorly understood. 🔍 Our Solution: Automated speech analysis as an objective means for detecting fatigue. 🤖 Our Method: Clinical assessment and a comprehensive speech protocol. Our support vector machine achieved an AUC of 0.74 in detecting fatigue — an impressive result! 📝 The Details: Specific spontaneous speech patterns are useful for fatigue detection. Notably, cognitive fatigue significantly correlates with a lower speech ratio in free speech (ρ = −0.283, p = 0.001). 🌟 Bottom Line: An automated speech analysis, based on a single narrative free speech task, offers an objective, ecologically valid, and low-burden method for fatigue assessment. The potential for clinical practice is promising. Curious? Take a look at our paper and discover the exciting possibilities of speech pattern analysis! 🌍📚 https://lnkd.in/eWGpRhur Share these insights with your colleagues, because science is way more fun when shared! 😄🤗 🧬 #NeuroScience #MSResearch #FatigueDetection #AI #SpeechAnalysis #DigitalHealth Marcelo D. Felix Dörr Simona Schäfer Nicklas Linz James Overell Helen Hayward-Koennecke, Dr.med. Anja Dillenseger Björn Tackenberg Technische Universität Dresden Universitätsklinikum Carl Gustav Carus Dresden ECTRIMS Deutsche Gesellschaft für Neurologie e.V. Dresden International University (DIU) CeTI Vodafone Tech Innovation Center Dresden

Innovative digital measurements are promising in the context of better and more sensitively monitoring diesease activity in chronic neurological disorders like Multiple Sclerosis. Thank you Tjalf Ziemssen, ki:elements and Tina Boggiano and their teams to discover new ways to utilize speech analyses in MS for medical practice. Curious to see the further evolution of this great piece of work !

Did you know that our Mili platform supports #Swahili? 🧐 I’m excited to be presenting at the conference "The Future of Dementia in Africa: Advancing Global Partnerships”. I will bring some early insights on dementia screening in Swahili from an ongoing study in Nairobi funded by Davos Alzheimer's Collaborative. 🌍 This is a joint effort between ki:elements, Karen Blackmon, PhD, ABPP-CN ‘s group (Aga Khan University), Irene Meier, and Vaibhav Narayan (both from Davos Alzheimer's Collaborative). If you are around, I am open for coffee invitations! 😎

Our ki:elements team met in person for the third time this year. 🙌 Me and half of the team work in our Saarbrücken office. But we have colleagues based in other parts of Germany and Europe, like France, Finland, and Portugal. For three days, everybody joined us at Saarbrücken. Barbecues, team dinners, cocktail hour, and a guided tour of the Völklingen Ironworks. 👌 See you again in December, guys!