Tuomo Hartonen
@tuomohartonen.bsky.social
I am developing machine learning approaches to harness the information contained in the electronic health records, national registries and biobanks for improved disease diagnosis and healthcare resource allocation. Based at FIMM, University of Helsinki.
Here we looked into more automated modelling of EHR data, and also diseases with no clinical risk models. Perhaps this information could be utilized in addition to models we already have? And also where we do not yet have established risk models. In my opinion, both approaches are valuable.
September 1, 2025 at 5:08 PM
Here we looked into more automated modelling of EHR data, and also diseases with no clinical risk models. Perhaps this information could be utilized in addition to models we already have? And also where we do not yet have established risk models. In my opinion, both approaches are valuable.
Absolutely, there are studies combining PGS with risk factors such blood pressure, smoking etc., often focusing on risk of one disease and thus going quite deep into that disease-specific risk factors. And we also mention this in the paper.
September 1, 2025 at 5:08 PM
Absolutely, there are studies combining PGS with risk factors such blood pressure, smoking etc., often focusing on risk of one disease and thus going quite deep into that disease-specific risk factors. And we also mention this in the paper.
I could have worded it better I think. The main point we had was to look whether these two sources of information are complementary in prediction models, and it seemed to be the case.
September 1, 2025 at 6:22 AM
I could have worded it better I think. The main point we had was to look whether these two sources of information are complementary in prediction models, and it seemed to be the case.
In case you want to check a quick run-through of the main results, I have written a blog post summarizing just these:
tuomohartonen.substack.com/p/medical-re...
tuomohartonen.substack.com/p/medical-re...
Medical records-based information and genetics are orthogonal predictors of disease risk
Instead of a review, this blog entry is a short summary about our own recently published paper.
tuomohartonen.substack.com
August 27, 2025 at 2:15 PM
In case you want to check a quick run-through of the main results, I have written a blog post summarizing just these:
tuomohartonen.substack.com/p/medical-re...
tuomohartonen.substack.com/p/medical-re...
This was a nice collaborative effort within the INTERVENE consortium which I had the chance to co-lead together with @detroiki.bsky.social and @andganna.bsky.social with strong support from our colleagues from University of Tartu, Estonia!
August 27, 2025 at 2:15 PM
This was a nice collaborative effort within the INTERVENE consortium which I had the chance to co-lead together with @detroiki.bsky.social and @andganna.bsky.social with strong support from our colleagues from University of Tartu, Estonia!
Reminder to come check our work on sex-specific effects in human plasma proteome today at 4pm. The poster is located just behind the break area #eshg2025
May 26, 2025 at 10:24 AM
Reminder to come check our work on sex-specific effects in human plasma proteome today at 4pm. The poster is located just behind the break area #eshg2025
A big plus is for sure that one gets to enjoy the views much more than when flying! This is me soaking in the last rays of the sun on the sundeck of the ferry to Stockholm 😎 I think for many trips within Europe, flight is not always the only viable option. 13/13.
May 24, 2025 at 11:28 AM
A big plus is for sure that one gets to enjoy the views much more than when flying! This is me soaking in the last rays of the sun on the sundeck of the ferry to Stockholm 😎 I think for many trips within Europe, flight is not always the only viable option. 13/13.
All in all, for this specific trip overland travel worked out very well without big downsides. Except that booking the trip was much more complicated, and Uni Helsinki travel agent only allows invoicing flight tickets directly from Uni, so overland travel costs one only gets back after the trip 12/n
May 24, 2025 at 11:22 AM
All in all, for this specific trip overland travel worked out very well without big downsides. Except that booking the trip was much more complicated, and Uni Helsinki travel agent only allows invoicing flight tickets directly from Uni, so overland travel costs one only gets back after the trip 12/n
Arrived to Milan just five minutes behind the original schedule!
In total the trip was around 2,700 km, and the overland travel saved ~70% of carbon emissions compared to flying. I also had to start the trip ~12 hours earlier and I missed the first morning session of the Milan conference. 11/n
In total the trip was around 2,700 km, and the overland travel saved ~70% of carbon emissions compared to flying. I also had to start the trip ~12 hours earlier and I missed the first morning session of the Milan conference. 11/n
May 24, 2025 at 11:12 AM
Arrived to Milan just five minutes behind the original schedule!
In total the trip was around 2,700 km, and the overland travel saved ~70% of carbon emissions compared to flying. I also had to start the trip ~12 hours earlier and I missed the first morning session of the Milan conference. 11/n
In total the trip was around 2,700 km, and the overland travel saved ~70% of carbon emissions compared to flying. I also had to start the trip ~12 hours earlier and I missed the first morning session of the Milan conference. 11/n