HS Analysis GmbH
Hybrid workshop on implementing custom deep learning models for image processing

HS Analysis ยท events

Hybrid workshop on implementing custom deep learning models for image processing

With the HSA Workshops, HS Analysis provides specialized training for engineers and medical professionals to learn artificial intelligence in image processing with a focus on deep learning applications. Engineers and Doctors can learn to collaborate with HSA KIT for an interactive, efficient learning experience. Upon workshop completion, participants will receive certification in deep learning implementation skills.

Course options include:

Artificial Intelligence

Deep Learning

Classification, Object Detection, Segmentation and Instance Segmentation

Hyperparameter Optimization

Metrics for Deep Learning Models

Image Analysis of Medical Data

Day 1

Time: 1:00 PM โ€“ 5:00 PM Berlin time Duration: 4 hours

Introduction

Basics of medical imaging, AI, ML, DL and DL technologies

Types of GTD for single deep learning technologies

Protocols for creating GTD (HSA GTD Protocol)

Required minimal amount and quality of GTD

Introduction to the HSA KIT module Annotator

Practical work and creation of GTD with HSA KIT

Introduction to hyperparameters with HSA KIT for: Classification

Object Detection

Segmentation

Instance Segmentation

Practical work and training of deep learning models with HSA KIT for Classification

Object Detection

Segmentation

Instance Segmentation

Summary

Discussion

Day 2

Time: 1:00 PM โ€“ 5:00 PM Berlin time Duration: 4 hours

Validation process theory

Basics of metrics

Basics of metrics for each deep learning category

Interpretation of metrics

Decision for the best deep learning model based on metrics

Decision for the best deep learning model based on visual results

Practical work on all trained deep learning models and decision for the best model for each deep learning technology

Automated image analysis with HSA KIT

Module selection

Project creation

Use of deep learning models

Settings for deep learning models

Processing of big data

Quantification results

Archiving

Reuse of existing projects

Foundation of objective, reproducible, and sustainable image analysis on medical images

Summary

Discussion