Keno Bressem
@k_bressem
Followers
220
Following
654
Media
5
Statuses
96
Radiologist interested in deep learning for radiology
Munich, Bavaria
Joined September 2020
🦙🩺 Proud to present #medAlpaca: Our latest research in fine-tuning Large Language Models for medical Q&A and dialogue applications. Leveraging 250k+ medical Q&A sets from our newly created dataset, the Medical Meadow. Get the data, models, and code here:
github.com
LLM finetuned for medical question answering. Contribute to kbressem/medAlpaca development by creating an account on GitHub.
3
7
29
RT @Radiology_AI: MRSegmentator accurately segmented 40 anatomical structures on #MRI and #CT across 3 external datasets .
0
1
0
RT @COMFORT_EU: Happy International Day of Radiology!🌐.Radiology has come far since the discovery of X-rays in 1895. Today, MRIs, ultrasoun….
0
3
0
RT @RadiologyEditor: 🛡️Privacy-preserving, open-source large language models show promise as an alternative to OpenAI’s GPT-4 for accurate….
0
4
0
RT @Fel_Busch: How will large language models transform structured reporting in radiology and beyond? In our comprehensive review just publ….
0
3
0
The study is now available as a preprint at: Many thanks to our global collaborators who made this study possible! Special shout out to @Fel_Busch for leading this initiative. #COMFORTproject #HorizonEU.
medrxiv.org
The successful implementation of artificial intelligence (AI) in healthcare is dependent upon the acceptance of this technology by key stakeholders, particularly patients, who are the primary...
1
3
4
Key findings from our study include:. * 72.9% prefer AI-physician collaboration.* Patients in poorer health were less supportive of AI use.* 61.8% worry that AI could reduce doctor-patient interaction.* 71.4% prefer healthcare organizations that use AI software
1
0
1
Our respondents spanned six continents:. 🌍 Europe (5,764 / 41.7%) .🌏 Asia (3,473 / 25.2%) .🌎 North America (2,284 / 16.5%) .🌎 South America (1,336 / 9.7%) .🌍 Africa (728 / 5.3%) .🌏 Oceania (221 / 1.6%).
1
0
0
With AI becoming increasingly crucial in healthcare, we wondered: how do patients feel about its use? To find out, we conducted a massive global survey, gathering insights from 13,806 patients across 74 hospitals in 43 countries as part of the #COMFORTproject.
1
5
11
RT @IsabellaWies: Exciting news! 🚀 We've just launched our innovative solution to one of the biggest #challenges in medical research: extra….
medrxiv.org
In clinical science and practice, text data, such as clinical letters or procedure reports, is stored in an unstructured way. This type of data is not a quantifiable resource for any kind of quanti...
0
5
0
RT @Radiology_AI: #DeepLearning model highly accurate at classifying cardiac implants chest xrays @Fel_Busch @HugoA….
0
3
0
RT @Radiology_AI: Chest Radiographs as Biological Clocks: Implications for Risk Stratification and Personalized Care .
0
1
0
RT @npjDigitalMed: An excellent reference point for the key elements of the @EU_Commission's ground breaking #ArtificialInteligence Act and….
0
7
0
RT @npjDigitalMed: New article! Read Navigating the European Union Artificial Intelligence Act for Healthcare in n….
nature.com
npj Digital Medicine - The European Union’s recently adopted Artificial Intelligence (AI) Act is the first comprehensive legal framework specifically on AI. This is particularly important for...
0
10
0
RT @Radiology_AI: @Fel_Busch @HugoAerts @k_bressem @LCAdamsRad present a #DeepLearning model to classify implants on chest xrays https://t.….
0
3
0
RT @Radiology_AI: #DeepLearning model highly accurate at classifying cardiac implants chest xrays @Fel_Busch @k_bre….
0
6
0
T5. If you want to understand the transformer architecture better, I recommend "The Illustrated Transformer" by @JayAlammar. #RadAIchat.
0
1
8
T5. A great resource for learning about BERT and AI in general is @fastdotai by @jeremyphoward. I highly recommend it to anyone starting out in AI (. #RadAIchat.
course.fast.ai
A free course designed for people with some coding experience, who want to learn how to apply deep learning and machine learning to practical problems.
0
0
7
T4. Keeping BERT models up to date with the latest clinical guidelines and research requires ongoing maintenance and retraining. #RadAIChat.
0
1
7
T4. BERT models can help classify text for common diseases, but will struggle with edge cases. #RadAIChat.
0
1
5