iiinova — where your digital self keeps evolving

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 02/04/2026

I just realized we'll probably be getting an updated version of this photo 🥰 #artemis #bluemarble

Blue marble photo of Earth taken from Apollo 17

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 25/12/2025

Merry Christmas and happy holidays to all! May you all be with/find a person who GETS you 🥰 #science #stirlingengine #christmas 🧪

511615

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 20/02/2025

Awesome, can't wait to complete the set! 🥰

Fox and Cat, in autumn and winter, framed on a wall

000

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 24/12/2024

The classics never get old. A very happy holiday season to all. 😊

030

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 25/11/2024

Field of fluorescent mushrooms: LNCaP cell line, IF stain. Nuclei in blue, nuclear membrane in red, cytoplasm in green :) #PathSky #MedSky #science 🧪

1141

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 20/11/2024

Quad immunofluorescent stain in liver tissue. Nuclei in pink, nuclear membrane in green, cell membrane in red and sinusoids in yellow 😍 #PathSky #science 🧪

Quadruple immunofluorescent stain of liver tissue.

1172

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 18/11/2024

Dataset is exported to stats analysis environment, cells of interest are reconstructed as datapoints and distances calculated:

Tissue reconstructed as datapoints in analysis environment

250

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 18/11/2024

Cells of interest identified, X-Y coordinates marked:

Cells of interest detected and characterised

120

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 18/11/2024

Cells detected and analysed for IF marker intensity, this overlay is for all markers combined:

Cell detection and marker intensity measurement

120

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 18/11/2024

Same image, with only markers of interest. Blue is cell nuclei, purple and green are the two cell markers we're interested in:

Prostatic adenocarcinoma, needle biopsy, markers of interest

120

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 18/11/2024

Inflammation in prostate adenocarcinoma, hyperplex IF. Calculating distances between cells of interest: #pathsky #science 🧪

+50 marker hyperplex IF of prostatic adenocarcinoma needle biopsy

1203

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 16/11/2024

I have been following @davekellett.bsky.social 's work for over 15 years now and could not recommend DRIVE highly enough. A must read for any sci-fi fan!

0187

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 16/11/2024

Post a picture of yourself to show the newbies what they're dealing with 😁

040

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 15/11/2024

Intro to computational pathology: Image analysis of IF stains generates vast datasets. Beyond simple positivity, these data - when imported into a statistical computing environment - enable digital reconstruction and analysis of entire biopsies. #pathsky #science 🧪

Cells detected and identified with deep learning algorithm

0122

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 11/11/2024

Then the computer looks at every marker in everycell, using yet another deep learning algorithm, identifying each cell. We then look at how these cells are interacting with each other, and how specific treatments affect these cells overall. #pathsky #science #medsky 🧪

Marker detection, scoring and cellular phenotyping

160

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 11/11/2024

Each cell in the image is detected by yet another deep learning algorithm, which looks at the shape of each cell nucleus and decides if it is malignant or benign. Red cells are malignant, blue are benign, light blue are immune cells and green are stromal cell nuclei. #pathsky #science #medsky 🧪

Nuclei detection and morphological classification

100

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 11/11/2024

The slide is then digitized with a scanner, and a deep-learning algorithm (trained by yours truly) then detects and annotates different types of tissue. Here, red is tumour, light blue is benign, green is stroma and yellow is empty areas. We used to do this by hand 😭 #pathsky #science #medsky 🧪

AI detection and segmentation of different tissue types

100

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 11/11/2024

How deep learning assists scientists in characterising tissue, a story in a few parts: Tumour tissue is stained with a large number of fluorescent antibodies, each detecting a specific protein that assists in naming (phenotyping) each cell. Each blue blob is one cell. #pathsky #science #medsky 🧪

260

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 10/11/2024

Makes me feel like the #predator, not going to lie 😁

010

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 10/11/2024

Science can be beautiful: +25-antibody hyperplex immunofluorescence, allows us to visualize a lot of proteins in tumour tissue all at once. You can then run deep learning-based image analysis routines to examine how different kinds of immune cells respond to cancer. #pathsky #science #medsky

3548

Bora Gurel, MD, Ph.D @boragurel.bsky.social · 22/10/2024

Our lab does a lot of IF, which often result in some pretty stunning images:

Immunofluorescence image of a double-stained prostate cancer sample

130