André Forjaz

InterpolAI accurately generates synthetic images between real image pairs to help restore tissue damage and reduce stitching artifacts across different imaging modalities. Shows that deep learning can understand data to intelligently fill in the gaps.

Scientific research, working for everyone.

Thank you Rita.
Very proud of this work, co-led by PhD students Saurabh Joshi and André Forjaz.

Very excited about this paper that came online today! InterpolAI can predict missing image data from light microscopy to H&E to MRI for improved volumetric imaging and image analysis.

Hopkins on the Hill!

Johns Hopkins faculty will showcase the value of their federally-funded research on the Hill in Washington DC, talking to US representatives and their staff

June 11, 5:30-7:30pm, Rayburn House Office Bldg, DC

AI, health care, space exploration, public health, education, etc

Can a deep learning model know your data well enough to predict missing parts? InterpolAI accurately interpolates synthetic images between pairs of authentic images to “repair” tissue damage and reduce stitching artifacts across diverse modalities. @deniswirtz.bsky.social bit.ly/453o1i1

This work was supported by the National Cancer Institute (NCI). The NCI has long supported the integration of various expertises (here, oncology, computer science, and microscopy), to develop new understanding of cancer and new treatments.

Proposed cuts to NCI will cripple such efforts.

InterpolAI: deep learning-based optical flow interpolation and restoration of biomedical images for improved 3D tissue mapping Nature Methods - InterpolAI leverages optimal flow-based artificial intelligence to produce synthetic images between pairs of images for diverse three-dimensional image types. InterpolAI is more...

Dream Police alert: InterpolAI for enhanced 3D imaging (CODA, MRI, EM, confocal).

Read about and InterpolAI here: www.nature.com/articles/s41...

Proud of our new perspective describing the role of mathematical biology for immunotherapy www.nature.com/articles/s41... @natbiotech.nature.com

Preprint: www.biorxiv.org/content/10.1...

Excited to share our new paper, where we integrate 3D imaging and nuclear segmentation to profile cellular neighborhoods in human pancreas.

We demonstrate spatial immune patterns and microenvironment heterogeneity missed in 2D. A scalable pipeline for high-res tissue analysis.

Be on the lookout for a new paper from the lab:

"InterpolAI: Deep learning based optical flow interpolation and restoration of biomedical images for improved 3D tissue mapping"

InterpolAI interpolates and restores defective images for enhanced 3D imaging (3D histology, MRI, confocal, ssTEM)

This integrated method shows how misleading traditional 2D assessments of immune cells localization in tissues is misleading.

3D assessments are a must.

New paper from the lab: Integration of nuclear morphology and 3D imaging to profile cellular neighborhoods

Read it here: www.biorxiv.org/content/10.1...