Professor in biomedical engineering. Research in data analysis, image processing, statistics, deep learning. Tweets in Swedish and English

Sweden
Joined February 2011
For every like I will make the researcher more tired
8
71
30
1,349
257,774
Our multimodal post-covid study is now on biorxiv Lingering cognitive symptoms of post-COVID-19 condition (PCC) continue to challenge our understanding, highlighting the need for interdisciplinary research. We investigated whether individuals with PCC and fatigue differ from matched controls across neurocognitive, neuroimaging, and immunological measures, integrating these findings with qualitative phenomenological analyses through multimodal cross-analysis. Individuals with PCC and clinically significant fatigue (n=34), and age-, sex-, and education-matched controls recovered from COVID-19 without PCC (n=37) were included. Assessments encompassed physician evaluation, neuropsychological and functional tests, structural and functional MRI, immunological tests, and qualitative phenomenological analysis of interviews. Data were analyzed using established methods and an interdisciplinary cross-analysis to examine persistent cognitive symptoms in PCC. Group comparisons revealed subtle differences between individuals with PCC and controls, while neurocognitive performance and immunological markers remained within normative ranges. Modest differences in white matter, myelin related, and global brain measures were observed, alongside differences in associations between network connectivity and neurocognitive performance across groups, suggesting reduced network resilience rather than domain specific deficits. This pattern resonated with qualitative phenomenological findings: altered experiences of concentrating and difficulties to execute decisions and actions, despite preserved performance on standard neurocognitive tests. These findings suggest that cognitive symptoms in post-COVID-19 fatigue are associated with subtle neurobiological alterations not captured by standard cognitive assessments. In this highly fatigued sample, converging neuroimaging, neurocognitive and immunological findings point toward reduced network efficiency rather than overt cognitive impairment. Qualitative phenomenological analysis further highlighted altered modes of being in the world characterized by increased cognitive effort and difficulties executing decisions and actions. biorxiv.org/content/10.64898…
2
247
Anders Eklund retweeted
‘How are you so focused under pressure?’ Me in 1995:
249
1,616
217
11,324
615,977
Anders Eklund retweeted
I'm never reading IKEA instructions again
176
481
102
10,410
832,699
Are there tools for comparing how similar two Jupyter notebooks are?
1
1
296
Anders Eklund retweeted
if you write an ai statement like this im assuming you have no fucking clue whats going on
11
10
4
494
400,792
Inte konstigt man kämpar med att hitta finansiering
176
Pangram is absolutely crap. I have papers from 2016 and before that I obviously wrote myself and score as 100% AI generated. Do not trust AI to evaluate whether something is written by AI.
56
29
18
293
85,027
Hur kan en falafelrulle bara kosta 25 kr i Lund?
2
2
468
Working 5 years on a project and now we have the final results...not looked yet
1
7
303
Think I just broke my LLM I am a text-based AI, and that is beyond my capabilities.
1
220
a bride just gifted her husband a $12,000 GPU. and they say romance is dead.
62
107
47
1,064
158,339
Anders Eklund retweeted
Average time spent composing one email
15
532
51
8,536
146,938
🇪🇺 Europe competing with Elon Musk's flying car.
64
214
31
1,851
55,579
Thanks to all who helped with the course
11
290
Today @GoogleResearch is launching Deep Review — an agentic system designed to stress-test manuscripts, design docs, and complex proposals before submission or publication. After pilot partnerships analyzing over 15,000 research papers, we’re making Deep Review globally accessible to help researchers, engineers, and professionals bring structured critique to their daily workflows.
36
586
89
4,397
366,754
Блогер показал, что собеседование в 2026 году проходит уже без людей. Компания прислала на созвон ИИ-рекрутера, блогер в ответ подключил своего ИИ-агента и вышел из кадра. Нейросети ещё какое-то время обсуждали вакансию без людей. В конце обе зависли и начали бесконечно прощаться друг с другом.
271
3,202
286
35,968
8,652,217
Anders Eklund retweeted
Reading the grant application I wrote last year
1
3
298
Resultat från konferensen medicinteknikdagarna, suck
3
5
489
Reading the grant application I wrote last year
1
3
298
Anders Eklund retweeted
How professors spend their time
6
316
32
1,580
84,752