Publications

To promote collaboration and data sharing, below are recent machine learning publications.
Please submit your medical imaging machine learning publications using the form below.

View Publications

2022 | Self-supervised driven consistency training for annotation efficient histopathology image analysis

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: In this work, we propose two novel strategies for unsupervised representation learning in histology: (i) a self-supervised pretext task that harnesses the underlying multi-resolution contextual cues in histology whole-slide images to learn a powerful supervisory signal for unsupervised representation learning; (ii) a new teacher-student semi-supervised consistency paradigm that learns to effectively transfer the pretrained representations to downstream tasks based on prediction consistency with the task-specific unlabeled data. We carry out extensive validation experiments on three histopathology benchmark datasets across two classification and one regression based tasks.
Modality: Optical/Microscopy
Organ: Breast
Disease: Cancer
Click Here to Read the Full Publication

2022 | Self supervised contrastive learning for digital histopathology.

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: We apply a contrastive self-supervised method to digital histopathology. A large-scale study with 57 histopathology datasets without labels was conducted. Our study focuses on differences between natural-scene and histopathology images. We find combining multiple multi-organ datasets improves task performances. Using more pretraining data has diminishing returns after around 50,000 images.
Modality: Optical/Microscopy
Organ: Prostate, Breast, Colorectal
Disease: Cancer
Click Here to Read the Full Publication

2022 | Loss odyssey in medical image segmentation

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: In this paper, we present a comprehensive review of segmentation loss functions in an organized manner. We also conduct the first large-scale analysis of 20 general loss functions on four typical 3D segmentation tasks involving six public datasets from 10+ medical centres. Our code and segmentation results are publicly available and can serve as a loss function benchmark. We hope this work will also provide insights on new loss function development for the community.
Modality: CT
Organ: Liver, Pancreas
Disease: Cancer
Click Here to Read the Full Publication

2021 | AMINN: Autoencoder-Based Multiple Instance Neural Network Improves Outcome Prediction in Multifocal Liver Metastases

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: We built a multiple instance learning network with radiomic features of liver MRI to predict survival of multifocal colorectal cancer liver metastases patients. We empirically validated our hypothesis that incorporating imaging features of all lesions improves outcome prediction for multifocal cancer. We released our code at https://github.com/martellab-sri/AMINN.
Modality: MR
Organ: Liver
Disease: Cancer
Click Here to Read the Full Publication

bottom of publications

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium
  • All
  • CT
  • MR
  • Optical/Microscopy
  • PET
  • Ultrasound
  • X-Ray

2022 | Self-supervised driven consistency training for annotation efficient histopathology image analysis

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: In this work, we propose two novel strategies for unsupervised representation learning in histology: (i) a self-supervised pretext task that harnesses the underlying multi-resolution contextual cues in histology whole-slide images to learn a powerful supervisory signal for unsupervised representation learning; (ii) a new teacher-student semi-supervised consistency paradigm that learns to effectively transfer the pretrained representations to downstream tasks based on prediction consistency with the task-specific unlabeled data. We carry out extensive validation experiments on three histopathology benchmark datasets across two classification and one regression based tasks.
Modality: Optical/Microscopy
Organ: Breast
Disease: Cancer
Click Here to Read the Full Publication

2022 | Self supervised contrastive learning for digital histopathology.

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: We apply a contrastive self-supervised method to digital histopathology. A large-scale study with 57 histopathology datasets without labels was conducted. Our study focuses on differences between natural-scene and histopathology images. We find combining multiple multi-organ datasets improves task performances. Using more pretraining data has diminishing returns after around 50,000 images.
Modality: Optical/Microscopy
Organ: Prostate, Breast, Colorectal
Disease: Cancer
Click Here to Read the Full Publication

2022 | Loss odyssey in medical image segmentation

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: In this paper, we present a comprehensive review of segmentation loss functions in an organized manner. We also conduct the first large-scale analysis of 20 general loss functions on four typical 3D segmentation tasks involving six public datasets from 10+ medical centres. Our code and segmentation results are publicly available and can serve as a loss function benchmark. We hope this work will also provide insights on new loss function development for the community.
Modality: CT
Organ: Liver, Pancreas
Disease: Cancer
Click Here to Read the Full Publication

2021 | AMINN: Autoencoder-Based Multiple Instance Neural Network Improves Outcome Prediction in Multifocal Liver Metastases

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: We built a multiple instance learning network with radiomic features of liver MRI to predict survival of multifocal colorectal cancer liver metastases patients. We empirically validated our hypothesis that incorporating imaging features of all lesions improves outcome prediction for multifocal cancer. We released our code at https://github.com/martellab-sri/AMINN.
Modality: MR
Organ: Liver
Disease: Cancer
Click Here to Read the Full Publication

bottom of publications

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium
  • All
  • Brain
  • Breast
  • Cardiac
  • Head and Neck
  • Hematopathology
  • Kidney and Liver
  • Lung
  • Musculoskeletal
  • Prostate
  • Vascular

2022 | Self-supervised driven consistency training for annotation efficient histopathology image analysis

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: In this work, we propose two novel strategies for unsupervised representation learning in histology: (i) a self-supervised pretext task that harnesses the underlying multi-resolution contextual cues in histology whole-slide images to learn a powerful supervisory signal for unsupervised representation learning; (ii) a new teacher-student semi-supervised consistency paradigm that learns to effectively transfer the pretrained representations to downstream tasks based on prediction consistency with the task-specific unlabeled data. We carry out extensive validation experiments on three histopathology benchmark datasets across two classification and one regression based tasks.
Modality: Optical/Microscopy
Organ: Breast
Disease: Cancer
Click Here to Read the Full Publication

2022 | Self supervised contrastive learning for digital histopathology.

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: We apply a contrastive self-supervised method to digital histopathology. A large-scale study with 57 histopathology datasets without labels was conducted. Our study focuses on differences between natural-scene and histopathology images. We find combining multiple multi-organ datasets improves task performances. Using more pretraining data has diminishing returns after around 50,000 images.
Modality: Optical/Microscopy
Organ: Prostate, Breast, Colorectal
Disease: Cancer
Click Here to Read the Full Publication

2022 | Loss odyssey in medical image segmentation

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: In this paper, we present a comprehensive review of segmentation loss functions in an organized manner. We also conduct the first large-scale analysis of 20 general loss functions on four typical 3D segmentation tasks involving six public datasets from 10+ medical centres. Our code and segmentation results are publicly available and can serve as a loss function benchmark. We hope this work will also provide insights on new loss function development for the community.
Modality: CT
Organ: Liver, Pancreas
Disease: Cancer
Click Here to Read the Full Publication

2021 | AMINN: Autoencoder-Based Multiple Instance Neural Network Improves Outcome Prediction in Multifocal Liver Metastases

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: We built a multiple instance learning network with radiomic features of liver MRI to predict survival of multifocal colorectal cancer liver metastases patients. We empirically validated our hypothesis that incorporating imaging features of all lesions improves outcome prediction for multifocal cancer. We released our code at https://github.com/martellab-sri/AMINN.
Modality: MR
Organ: Liver
Disease: Cancer
Click Here to Read the Full Publication

bottom of publications

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium
  • All
  • Cancer
  • Heart Disease
  • Kidney and Liver Disease
  • Musculoskeletal Disease
  • Neurological Disease
  • Respiratory Disease
  • Stroke and Cardiovascular

2022 | Self-supervised driven consistency training for annotation efficient histopathology image analysis

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: In this work, we propose two novel strategies for unsupervised representation learning in histology: (i) a self-supervised pretext task that harnesses the underlying multi-resolution contextual cues in histology whole-slide images to learn a powerful supervisory signal for unsupervised representation learning; (ii) a new teacher-student semi-supervised consistency paradigm that learns to effectively transfer the pretrained representations to downstream tasks based on prediction consistency with the task-specific unlabeled data. We carry out extensive validation experiments on three histopathology benchmark datasets across two classification and one regression based tasks.
Modality: Optical/Microscopy
Organ: Breast
Disease: Cancer
Click Here to Read the Full Publication

2022 | Self supervised contrastive learning for digital histopathology.

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: We apply a contrastive self-supervised method to digital histopathology. A large-scale study with 57 histopathology datasets without labels was conducted. Our study focuses on differences between natural-scene and histopathology images. We find combining multiple multi-organ datasets improves task performances. Using more pretraining data has diminishing returns after around 50,000 images.
Modality: Optical/Microscopy
Organ: Prostate, Breast, Colorectal
Disease: Cancer
Click Here to Read the Full Publication

2022 | Loss odyssey in medical image segmentation

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: In this paper, we present a comprehensive review of segmentation loss functions in an organized manner. We also conduct the first large-scale analysis of 20 general loss functions on four typical 3D segmentation tasks involving six public datasets from 10+ medical centres. Our code and segmentation results are publicly available and can serve as a loss function benchmark. We hope this work will also provide insights on new loss function development for the community.
Modality: CT
Organ: Liver, Pancreas
Disease: Cancer
Click Here to Read the Full Publication

2021 | AMINN: Autoencoder-Based Multiple Instance Neural Network Improves Outcome Prediction in Multifocal Liver Metastases

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Short Description: We built a multiple instance learning network with radiomic features of liver MRI to predict survival of multifocal colorectal cancer liver metastases patients. We empirically validated our hypothesis that incorporating imaging features of all lesions improves outcome prediction for multifocal cancer. We released our code at https://github.com/martellab-sri/AMINN.
Modality: MR
Organ: Liver
Disease: Cancer
Click Here to Read the Full Publication

bottom of publications

150 150 MaLMIC - Machine Learning in Medical Imaging Consortium

Search Publications

(by publication title, year, modality, organ, or disease)

Search
Generic filters
Filter by Categories
Publications