How can AI succeed in hospital care? On September 28, the AMPEL team will be at the Saxon Hospital Day to present the AMPEL-CDSS as a successful example. We are looking forward to stimulating conversations!

More information: https://khg-sachsen.de/kgs-engl/

#MedicalAI #AMPEL #Saxony #eHealth #CDSS

Key findings:
- age and sex had a significant effect on the 99th percentile of serum hs-cTnT concentration in both cohorts
- hospitalization is associated with higher hs-cTnT levels in in most age groups than in individuals in the general population, even in patients without diagnosed or suspected heart disease
- age- and sex-specific hs-cTnT reference values may be beneficial for hospitalized patients

3/3

#cTnT #Troponin #AMI #AMPEL #CDSS #SexBias #PatientSafety

Which project has the greatest clinical benefit? This question was posed to over 5000 employees of the University Hospital #Leipzig as part of the modernization initiative "Operation Zukunft" (Operation Future).

We are very pleased that we have landed in 1st place. The feedback from our colleagues really means a lot to us!

Our #AMPEL-CDSS connects laboratory findings with additional data, detects critical situations in daily ward routines, and assists clinicians with automated alarms.

#CDSS

Had a great week at Country Dance & Song Society Dance, Music, & Spice camp at Camp Cavell. Lots of English Country Dance and contra dance, including one session of techno contra. And sessions on #Bal folk dance. Wonderful music from musicians, and lots of talent at the daily gatherings. #CDSS #ECD

More #MELD content: We recently published a better #ML approach “AMELD” which outperforms all previous MELD scores

https://doi.org/10.1515/labmed-2022-0162

#LTx #Liver #AMELD #CDSS #AMPEL

Unadjusted Creatinine Hinders Females to Get New Liver

- Standard mathematical conversion of #creatinine disadvantages females in liver transplantation prioritization

- We investigated options to reduce this gender inequality

- We identified 72 of 650 (11%) females with an original #MELD score < 20, for whom the adjusted female-as-male MELD score would be >20, thus giving them a better chance to receive a #Liver #Transplant

https://doi.org/10.3390/jpm13060963

#LTx #SexBias #GenderBias #CDSS #AMPEL

I have created a "Medical AI" list of mastodon users to help colleagues get started more quickly #twitterMigration

You can add yourself here:
https://followlists.online/list/42dd6cc8-e347-d0b1-fcdc-72e9c254626c

#MedicalAI #ML #CDSS #CDS #eHealth #DigitalHealth #med #AI

Can Sepsis AI Save Lives?

Last summer, there were highly acclaimed publications of an #ML model called 'TREWS' in Nature Medicine. Sepsis alarms were integrated into a Clinical Decision Support System #CDSS, enabling earlier detection of #Sepsis. Due to the alerts, antibiotics were administered approximately 2 hours earlier, resulting in a reduction of mortality by 3.3% (absolute) and 18.7% (relative).

These are fantastic results... too fantastic to be true? #Review

1/10

Our recent study about troponin thresholds (https://doi.org/10.3390/jcm10235508) shows the large differences in age- and sex-specific reference ranges. Young women are systematically underdiagnosed. The study by Arslani & Tontsch et al (described before) now suggests that this may lead to increased mortality in young women.

Gender and age-specific adjustment of troponin thresholds, is easy to accomplish, has low costs, and is likely to save lives.

#SexBias #Troponin #AMI #CDS #CDSS #AMPEL

2/2

A major challenge in AI medicine is publicly available datasets #Opendata

We therefore publish our #ML data (as well as our code #Opensource), enabling a wide range of future research.

Included in the publication of #AMELD (see previous post), we released another #Dataset + an associated #Rstats package with a variety of integrated functions:

https://github.com/ampel-leipzig/ameld

(R: remotes::install_github("ampel-leipzig/sbcdata")

#Openscience #Opensource #MELD #LTx #Livertransplantation #AMPEL #CDSS

#ML outperforms classic scores. One of these is the end-stage liver disease prediction model for 90-day mortality (#MELD)

- We derived an ML model AMELD which outperforms MELD, MELD-Na, MELD 3.0, and MELD-Plus7

- How? AMELD extends the classic MELD predictors INR, bilirubin, and cystatin C / creatinine to include total protein, cholinesterase and #IL-6

https://doi.org/10.1515/labmed-2022-0162

Using more accurate AMELD prediction may improve #CDS for performing liver transplants

#AMPEL #CDSS #LTx #Liver

Today was the kick off meeting of #INTERPOLAR a use case of the medical informatics initiative in Germany #MII
The interoperable structures of the German university hospitals will be used to identify patients at risk for medical related problems and to perform an intervention with pharmacy visits. Our project is connected on several levels and looks forward to further collaboration. Let's harness the potential of #eHealth in clinics!

https://www.medizininformatik-initiative.de/en/about-initiative

#CDSS #datascience #riskscore

Today we also publish new data on the laboratory parameter serum creatinine and its use in staging of acute kidney injury:

- Women and elderly patients are systematically underdiagnosed

- A sex- and age-adapted approach is necessary to address this discrimination

Our aim is to automatically identify undetected medical conditions and integrate them into the clinicians' digital workflow.

https://doi.org/10.3390/jcm12031092

#CDSS #patientsafety #AKI #KDIGO #Troponin #Creatinine

2/2

Happy New Year!

Let’s start with our first publication 2023: https://doi.org/10.1515/cclm-2022-0883

Late follow-up measurements in patients with hyponatremia correlate with:
- Low sodium correction
- Discharge in hyponatremic states
- Inadequate fast correction
- Length of hospital stay

AMPEL aims to increase #PatientSafety by alerting these cases. Results of our prospective evaluation will be published soon.

#Hyponatremia #CDSS #OpenScience

#Introduction

The #AMPEL project operates a clinical decision support system in routine care and is continuously researching its application. We focus on increased #PatientSafety and improved diagnostics by harnessing the potential of #DigitalMedicine. In three German hospitals (>1500 beds), #AI models based mainly on laboratory data are applied in a randomized controlled setting. The project is funded by the German government #KHZG.

#CDSS #DataScience #OpenScience #UniLeipzig

As a #Physician, I started in #Anesthesiology and then found my way into #Science

With a highly motivated team we implement a clinical decision support system #CDSS based on #Labdata at the University Hospital #Leipzig (#AMPEL)

I am dedicated to sepsis prediction using #ML and #Rstats

As part of the German medical informatics initiative #MII, I am working on the interoperable use of #Labdata with #LOINC

I look forward to inspiring discussions about how we can leverage #Digitalmedicine