Diagnostic Assessment of Deep Learning Algorithms for Frozen Tissue Section Analysis in Women with Breast Cancer / Journal of the Korean Cancer Association, 대한암학회지

Purpose@#Assessing the metastasis status of the sentinel lymph nodes (SLNs) for hematoxylin and eosin–stained frozen tissue sections by pathologists is an essential but tedious and time-consuming task that contributes to accurate breast cancer staging. This study aimed to review a challenge competition (HeLP 2019) for the development of automated solutions for classifying the metastasis status of breast cancer patients. @*Materials and Methods@#A total of 524 digital slides were obtained from frozen SLN sections: 297 (56.7%) from Asan Medical Center (AMC) and 227 (43.4%) from Seoul National University Bundang Hospital (SNUBH), South Korea. The slides were divided into training, development, and validation sets, where the development set comprised slides from both institutions and training and validation set included slides from only AMC and SNUBH, respectively. The algorithms were assessed for area under the receiver operating characteristic curve (AUC) and measurement of the longest metastatic tumor diameter. The final total scores were calculated as the mean of the two metrics, and the three teams with AUC values greater than 0.500 were selected for review and analysis in this study. @*Results@#The top three teams showed AUC values of 0.891, 0.809, and 0.736 and major axis prediction scores of 0.525, 0.459, and 0.387 for the validation set. The major factor that lowered the diagnostic accuracy was micro-metastasis. @*Conclusion@#In this challenge competition, accurate deep learning algorithms were developed that can be helpful for making a diagnosis on intraoperative SLN biopsy. The clinical utility of this approach was evaluated by including an external validation set from SNUBH.

An Open Medical Platform to Share Source Code and Various Pre-Trained Weights for Models to Use in Deep Learning Research

Deep learning-based applications have great potential to enhance the quality of medical services. The power of deep learning depends on open databases and innovation. Radiologists can act as important mediators between deep learning and medicine by simultaneously playing pioneering and gatekeeping roles. The application of deep learning technology in medicine is sometimes restricted by ethical or legal issues, including patient privacy and confidentiality, data ownership, and limitations in patient agreement. In this paper, we present an open platform, MI2RLNet, for sharing source code and various pre-trained weights for models to use in downstream tasks, including education, application, and transfer learning, to encourage deep learning research in radiology. In addition, we describe how to use this open platform in the GitHub environment. Our source code and models may contribute to further deep learning research in radiology, which may facilitate applications in medicine and healthcare, especially in medical imaging, in the near future. All code is available at https://github.com/mi2rl/MI2RLNet.

Challenge for Diagnostic Assessment of Deep Learning Algorithm for Metastases Classification in Sentinel Lymph Nodes on Frozen Tissue Section Digital Slides in Women with Breast Cancer / Journal of the Korean Cancer Association, 대한암학회지

Purpose@#Assessing the status of metastasis in sentinel lymph nodes (SLNs) by pathologists is an essential task for the accurate staging of breast cancer. However, histopathological evaluation of sentinel lymph nodes by a pathologist is not easy and is a tedious and time-consuming task. The purpose of this study is to review a challenge competition (HeLP 2018) to develop automated solutions for the classification of metastases in hematoxylin and eosin–stained frozen tissue sections of SLNs in breast cancer patients. @*Materials and Methods@#A total of 297 digital slides were obtained from frozen SLN sections, which include post–neoadjuvant cases (n = 144, 48.5%) in Asan Medical Center, South Korea. The slides were divided into training, development, and validation sets. All of the imaging datasets have been manually segmented by expert pathologists. A total of 10 participants were allowed to use the Kakao challenge platform for six weeks with two P40 GPUs. The algorithms were assessed in terms of the AUC (area under receiver operating characteristic curve). @*Results@#The top three teams showed 0.986, 0.985, and 0.945 AUCs for the development set and 0.805, 0.776, and 0.765 AUCs for the validation set. Micrometastatic tumors, neoadjuvant systemic therapy, invasive lobular carcinoma, and histologic grade 3 were associated with lower diagnostic accuracy. @*Conclusion@#In a challenge competition, accurate deep learning algorithms have been developed, which can be helpful in making frozen diagnosis of intraoperative sentinel lymph node biopsy. Whether this approach has clinical utility will require evaluation in a clinical setting

Fully Automatic Segmentation of Acute Ischemic Lesions on Diffusion-Weighted Imaging Using Convolutional Neural Networks: Comparison with Conventional Algorithms

OBJECTIVE: To develop algorithms using convolutional neural networks (CNNs) for automatic segmentation of acute ischemic lesions on diffusion-weighted imaging (DWI) and compare them with conventional algorithms, including a thresholding-based segmentation. MATERIALS AND METHODS: Between September 2005 and August 2015, 429 patients presenting with acute cerebral ischemia (training:validation:test set = 246:89:94) were retrospectively enrolled in this study, which was performed under Institutional Review Board approval. Ground truth segmentations for acute ischemic lesions on DWI were manually drawn under the consensus of two expert radiologists. CNN algorithms were developed using two-dimensional U-Net with squeeze-and-excitation blocks (U-Net) and a DenseNet with squeeze-and-excitation blocks (DenseNet) with squeeze-and-excitation operations for automatic segmentation of acute ischemic lesions on DWI. The CNN algorithms were compared with conventional algorithms based on DWI and the apparent diffusion coefficient (ADC) signal intensity. The performances of the algorithms were assessed using the Dice index with 5-fold cross-validation. The Dice indices were analyzed according to infarct volumes ( 100), time intervals to DWI, and DWI protocols. RESULTS: The CNN algorithms were significantly superior to conventional algorithms (p < 0.001). Dice indices for the CNN algorithms were 0.85 for U-Net and DenseNet and 0.86 for an ensemble of U-Net and DenseNet, while the indices were 0.58 for ADC-b1000 and b1000-ADC and 0.52 for the commercial ADC algorithm. The Dice indices for small and large lesions, respectively, were 0.81 and 0.88 with U-Net, 0.80 and 0.88 with DenseNet, and 0.82 and 0.89 with the ensemble of U-Net and DenseNet. The CNN algorithms showed significant differences in Dice indices according to infarct volumes (p < 0.001). CONCLUSION: The CNN algorithm for automatic segmentation of acute ischemic lesions on DWI achieved Dice indices greater than or equal to 0.85 and showed superior performance to conventional algorithms.

Deep Learning in Medical Imaging: General Overview

The artificial neural network (ANN)–a machine learning technique inspired by the human neuronal synapse system–was introduced in the 1950s. However, the ANN was previously limited in its ability to solve actual problems, due to the vanishing gradient and overfitting problems with training of deep architecture, lack of computing power, and primarily the absence of sufficient data to train the computer system. Interest in this concept has lately resurfaced, due to the availability of big data, enhanced computing power with the current graphics processing units, and novel algorithms to train the deep neural network. Recent studies on this technology suggest its potentially to perform better than humans in some visual and auditory recognition tasks, which may portend its applications in medicine and healthcare, especially in medical imaging, in the foreseeable future. This review article offers perspectives on the history, development, and applications of deep learning technology, particularly regarding its applications in medical imaging.

Construction and Validation of Brain MRI Templates from a Korean Normal Elderly Population

OBJECTIVE: This study aimed to construct a Korean normal elderly brain template (KNE96) using Korean elderly individuals for use in brain MRI studies and to validate it. METHODS: We used high-resolution 3.0T T1 structural MR images from 96 Korean normal elderly individuals (M/F=48/48), aged 60 years or older (M=69.5+/-6.2 years, F=70.1+/-7.0 years), for constructing the KNE96 template. The KNE96 template was validated by comparing the registration-induced deformations between the KNE96 and ICBM152 templates using different MR images from 48 Korean normal elderly individuals (M/F=24/24), aged 60 years or older (M=71.5+/-5.9 years, F=72.8+/-5.1 years). We used the magnitude of displacement vectors (mag-displacement) and log of Jacobian determinants (log-Jacobian) to quantify the deformation produced during registration process to templates. RESULTS: The mag-displacement and log-Jacobian of the registration were much smaller using the KNE96 template than with the ICBM152 template in most brain regions. There was a prominent difference in the significant averaged differences (SADs) of the mag-displacement and log-Jacobian between the KNE96 and ICBM152 at the superior, medial, and middle frontal gyrus, the lingual, inferior, middle, and superior occipital gyrus, and the caudate and thalamus. CONCLUSION: This study suggests that templates constructed from Asian populations, such as the KNE96, may be more desirable than those from Caucasian populations, like the ICBM152, in computational neuroimaging studies that measure and compare anatomical features of the frontal and occipital lobe, thalamus and caudate.