Deep Learning Published

Medical Image Segmentation

Multi-task Deep Learning for Automated Medical Image Analysis

Aydin Ayanzadeh et al.
Istanbul Technical University
2019 - 2021

Overview

This project developed advanced deep learning pipelines for cell segmentation in label-free optical microscopy images, achieving state-of-the-art accuracy in medical imaging applications.

The work addresses critical challenges in medical image analysis, including handling imbalanced datasets, achieving high dice scores for segmentation tasks, and developing robust pipelines that can generalize across different imaging conditions and cell types.

Key Features

Multi-task Learning

Simultaneous optimization of multiple segmentation objectives for improved accuracy.

Imbalanced Data Handling

Advanced techniques to handle class imbalance common in medical imaging datasets.

Label-Free Imaging

Works with phase-contrast microscopy without requiring fluorescent labels.

High Accuracy

Achieved 95%+ dice score on benchmark medical imaging datasets.

Technologies Used

PyTorch Python OpenCV U-Net ResNet CUDA NumPy Scikit-Image

Related Publications

Turkish Journal of Electrical Engineering and Computer Sciences (2021)

Improved cell segmentation using deep learning in label-free optical microscopy images.

View Paper

IEEE SIU Conference 2020

Deep Learning based Segmentation Pipeline for Label-Free Phase-Contrast Microscopy Images.

View Paper

TIPTEKNO 2019

Cell Segmentation of 2D Phase-Contrast Microscopy Images with Deep Learning Method.

View Paper

Interested in Medical AI Research?

I'm open to collaborations on medical imaging and computer vision projects.

Get in Touch