Deep Learning in Medical Imaging Lab | HAYIT GREENSPAN's LAB

Welcome to the Deep Learning in Medical Imaging Lab

Research Focus Areas:


We focus on Machine Learning in Medical Image Analysis. In the past three years we have been focusing on Deep Learning and we bring these tools to the many important and challenging medical tasks: from image augmentation- to enable the physicians to visualize the image better and to detect earlier; to image segmentation, to separate out important areas within the image; to image quantification – to provide quantitative measurements, thus enabling more advanced quantitative analysis and tracking over time; to image labeling and classification – to support the radiologist and oncologist decision making process. Projects in the Lab span many focus areas, as listed below. In each focus area we collaborate with clinical partners, who support us in the identification of the medical challenge, provide us with the data examples – and join us in the exploration of computational tools which we develop. The work in the Lab is funded by both Israeli as well as International grants. Including: the Israeli Science Foundation (ISF), as well as awards from INTEL and GOOGLE.

Research Domains

  • MRI Brain Image Analysis (affiliated with the Sagol school of Brain research and Neuroscience): Augmenting resolution in MRI; Neurological diseases, Multiple-Sclerosis; CT lesion analysis;
  • CT Abdomen analysis: In particular, Liver lesions – detection, segmentation, characterization in time (Collaborations with Radiology Dept, Stanford & Head of CT Abdomen Unit at Sheba medical center, Israel)
  • Mammmography - microcalcification analysis
  • Xray or Radiograph Analysis: In particular, Chest analysis. Identification of diseases with the lungs, such as enlarged heart, liquid, air and more.
  • Uterine-cervix image analysis and indexing for cancer screening;
  • Medical image retrieval from large archives (PACS)

With Support From:


Longitudinal Multiple Sclerosis Lesion Segmentation using Multi-View Convolutional Neural Networks
Automatic segmentation of Multiple Sclerosis (MS) lesions is a challenging task due to their variability in shape, size, location and texture in Magnetic Resonance (MR) images. A reliable, automatic segmentation method can help diagnosis and patient follow-up while reducing the time consuming need of manual segmentation. In this paper, we present a fully automated method for MS lesion segmentation. The proposed method uses MR intensities and White Matter (WM) priors for extraction of candidate lesion voxels and uses Convolutional Neural Networks for false positive reduction. Our networks process longitudinal data, a novel contribution in the domain of MS lesion analysis. The method was tested on the ISBI 2015 dataset and obtained state-of-the-art Dice results with the performance level of a trained human rater. Keywords
Fully convolutional network for liver segmentation and lesion detection
In this work we explore a fully convolutional network (FCN) for the task of liver segmentation and liver metastases detection in computed tomography (CT) examinations. FCN has proven to be a very powerful tool for semantic segmentation. We explore the FCN performance on a relatively small dataset and compare it to patch based CNN and sparsity based classification schemes. Our data contains CT examinations from 20 patients with overall 68 lesions and 43 livers marked in one slice and 20 different patients with a full 3D liver segmentation. We ran 3-fold cross-validation and results indicate superiority of the FCN over all other methods tested. Using our fully automatic algorithm we achieved true positive rate of 0.86 and 0.6 false positive per case which are very promising and clinically relevant results.
MIDL 2018 Panel discussion about radiology and artificial intelligence
An interesting panel discussion about the influence of artificial intelligence on the radiology profession now and in the future took place on Friday,July 6th. Both the industry and academy were represented in the panel: Bram van Ginneken (Radboud University, Nijmegen), Heinrich von Busch (Siemens Healthineers), Tim Salimans (Aidence / Open AI), Graham Taylor (University of Guelph), Hayit Greenspan (Tel Aviv University), Ron Summers (NIH) , Mathias Prokop (Radboud University, Nijmegen)

Read a summary of the panel discussion on

News Item 2
How data science can help doctors fight cancer
Fully convolutional network for liver segmentation and lesion detection
.
Current Phd's
Alon Baram
Current Grad Students
Michal Heker
Rula Amer
Yana Podolsky
Eytan Kats
Alumni
Idit Diamant (PhD)
Gali Zimmerman Moreno (PhD)
Ariel Birenbaum
Guy Alexandroni
Evgeny Bal
Anat Shkolyar (2016) Ofer Geva (2016) Marina Asherov (2016) Yaron Anavi (2015) Moran Shalhon (2015) Roey Mechrez (2014) Avi Ben Cohen (2014) Amir Alush (2009) Arnaldo Mayer (2009) Orly Zyitia (2009) Shiri Gordon (2009) Sigal Trattner (2009) Uri Avni (2009) Shelly Lotenberg (2008) Oren Freifeld (2007) Jeremie Dreyfuss (2007) Omer Rotem (2007) Hila Dvir (2007) Yulian Wolf (2007) Yoad Bar-Shean (2007) Gali Zimmerman Moreno(2006) Adi Pinhas (2005) Amit Ruf (2005) Oron Shechner (2004) Allon Shahar (2004) Guy Dvir (2003) Arnaldo Mayer (2002) Gal Meister (2002) Gal Oz (2002) Sigal Trattner (2002) Shiri Gordon (2002)