Michele D. Pierri MD PhD

The Death of the Grandmother in Proust's "À la recherche du temps perdu" - micheledpierri.com: statistics, data analysis and coding A literary analysis of the grandmother's death in Proust's 'In Search of Lost Time' as a pivotal moment of existential revelation, exploring medicine's limitations, symbolic imagery, and the narrator'...

Proust described the death of the grandmother with the precision of a case report, capturing symptoms, decline, and unspoken signs. When literature mirrors medicine, empathy becomes the diagnosis.
www.micheledpierri.com/2025/07/23/t...
#Proust #NarrativeMedicine #MedicalHumanities

We tested ChatGPT, Claude & Gemini with 70 cardiology questions, evaluated by practicing cardiologists. Want to know the winner? Find out! Though none are quite ready for clinical practice yet. Full study: www.mdpi.com/2673-3846/6/...
#AI #Cardiology

Advanced 3D imaging is revolutionizing aortic root segmentation, enabling better valve sizing, reducing complications, and allowing for personalized treatment plans. The cardiac revolution is happening now—are you part of it?
#CardiacImaging #TAVR #CardiacSurgery #Cardiology

U-Net is a CNN architecture for medical image segmentation. It compresses and expands data like a “U” to detect organs, tumors, vessels, and more. Widely used in radiology & surgery planning. #UNet #AI #MedTech #DeepLearning #MedicalImaging

Residual plots reveal key issues in linear regression: clustered points indicate non-linear relationships, cone-shaped patterns show heteroscedasticity, and points concentrated on one side suggest bias. #Statistics #DataScience

Dimensionality Reduction Techniques A comprehensive guide to dimensionality reduction techniques in machine learning, covering linear methods like PCA and non-linear approaches like t-SNE and UMAP. The article explores implementation st...

Remember folks: in data science, as in life, sometimes less is more. Unless we're talking about pizza. Or computing power. Or... ok maybe that saying needs some dimensionality reduction too. #DataScience #ML
www.micheledpierri.com/machine-lear...

statistician screaming...

Sample size estimation be like: theoretically needed: 1000 patients; budget allows for 100 patients; PI says: 'We'll just increase alpha to 0.1'; statistician screaming... #Statistics #DataScience #healthcare

Meta-analysis: When you combine 20 underpowered studies to create one slightly less underpowered conclusion. But hey, at least now it's peer-reviewed! 🤓
#Statistics #HealthInformatics

Missing data in clinical trials is like a group project where half the team ghosted you, but you still need to submit something meaningful.👻 MCAR, MAR, MNAR... or as I call them: 'Gone', 'Gone with an excuse', and 'Gone suspicious' #Statistics #DataScience

🤖 Ensemble models combine multiple simpler models to achieve better predictions, reducing bias & variance. Key types: bagging (like Random Forest) for robustness, boosting for complex patterns

Random Numbers in Python - micheledpierri.com: statistics, data analysis and coding A comprehensive guide to random number generation in Python, covering different libraries (random, NumPy, PyTorch, secrets, os), their applications in statistics and data science, differences between ...

www.micheledpierri.com/2025/02/23/r... #Python #DataScience

Rare side effect: occurs in 1 in 10,000 patients. Doctors: that's rare 🙂. Statisticians: that's an impressively large sample size 😮. Patients: WHAT DO YOU MEAN IT CAN HAPPEN 😱? #Statistics #Medicine

🤖 Support Vector Machines are masters at finding optimal boundaries between data classes. Their strength with high-dimensional data and resilience against outliers make them particularly valuable for analyzing medical datasets! #Statistics #DataAnalysis

I just made a donation to Wikipedia—it's such a vital resource!

Curse of Dimensionality explained: as the number of features increases, distances between data points become less meaningful, and models struggle to generalize. Solution? Feature selection, dimensionality reduction, and other metrics (Mahalanobis, Cosine. #DataScience #MachineLearning

t.co/nxvLSPLEPr

🔮 Did you know the Poisson distribution can predict chaos? From heart rhythms in ERs to call center surges, this mathematical crystal ball turns randomness into insights. One parameter to rule them all! #DataScience #Statistics

⚖️ Predictive models in medicine reflect the biases in their data. Can we trust a system that might amplify healthcare inequalities?

You’re absolutely right! My Monte Carlo simulation shows the mean path, not necessarily the most probable outcome. But hey, life isn’t about probabilities—it’s about embracing the chaos of the journey! Let’s agree that randomness keeps us humble.

🎲 Life is like Monte Carlo simulation: we keep generating random scenarios until we find the most probable outcome... but reality always has the final say! Here's what 1000 simulations look like 👇 #DataScience #Statistics

🔬 When doing multiple statistical tests, your Type I error risk adds up! That's where Bonferroni comes in: divide your α by the number of tests (α/n). Simple but powerful! #Statistics #DataScience

📖🌐I found this book in my library: before Google, there was the Internet Yellow Pages—a printed guide to the web. Efficient? Not really. Nostalgic? Absolutely. You search "pizza" and get 1 billion results these days. Back then? Just a single page saying: "Try this link." 🍕😂 #InternetHistory #90sTech

Bayes' Theorem - micheledpierri.com Master Bayesian statistics, conditional probability, and statistical inference with our comprehensive guide. Learn to apply Bayes' theorem using PyMC and MCMC sampling in Python, with real-world examp...

www.micheledpierri.com/elements-of-...

🤯 Mind-bending medical math: A 95% accurate test can still mean only an 8.7% chance of disease! Welcome to Bayes' theorem, where prior probabilities flip our intuition upside down. #Statistics #DataScience #Healthcare

🎨 Visualize your missing data! Use seaborn heatmaps or missingno matrices to spot patterns. A picture is worth a thousand NaNs. #DataScience #Statistics #Python

Udite, udite rustici - micheledpierri.com Explore Donizetti's comic opera "The Elixir of Love" through the aria "Udite, udite rustici". Discover how the charlatan Dulcamara embodies 18th and 19th-century medical quackery, offering a satirical...

🎭 Donizetti's "Elixir of Love" (1832) mocks medical charlatans through Dulcamara's character. His satirical portrayal of quack doctors still resonates today. 🎵 #Opera #Healthcare
youtu.be/9B5_JWwe1mM?...
www.micheledpierri.com/2024/09/17/u...

🏥 #HealthcareData In healthcare, data duplication isn't just annoying; it can lead to serious clinical errors and unnecessary treatments. Prioritize data quality in healthcare systems, implement robust patient matching algorithms, and always err on the side of caution when merging health records.

📊 Millions of medical records, billions of data points: are we creating better treatments or bigger datasets? Don’t lose sight of the patient behind the numbers.

Through Dr. Rieux's eyes, Camus's "The Plague" masterfully captures humanity's vulnerability to epidemics and the enduring importance of medical ethics, while serving as a powerful allegory for Nazi occupation

From emergency tracheotomies to battling syphilis outbreaks, Bulgakov's semi-autobiographical work highlights the courage and resilience required in healthcare. A must-read for aspiring doctors and history enthusiasts alike. #MedicalHistory #Bulgakov #Literature