Deep Learning Diagnosis and Classification of Rotator Cuff Tears on Shoulder MRI.

BACKGROUND: Detection of rotator cuff tears, a common cause of shoulder disability, can be time-consuming and subject to reader variability. Deep learning (DL) has the potential to increase radiologist accuracy and consistency. PURPOSE: The aim of this study was to develop a prototype DL model for detection and classification of rotator cuff tears on shoulder magnetic resonance imaging into no tear, partial-thickness tear, or full-thickness tear. MATERIALS AND METHODS: This Health Insurance Portability and Accountability Act-compliant, institutional review board-approved study included a total of 11,925 noncontrast shoulder magnetic resonance imaging scans from 2 institutions, with 11,405 for development and 520 dedicated for final testing. A DL ensemble algorithm was developed that used 4 series as input from each examination: fluid-sensitive sequences in 3 planes and a sagittal oblique T1-weighted sequence. Radiology reports served as ground truth for training with categories of no tear, partial tear, or full-thickness tear. A multireader study was conducted for the test set ground truth, which was determined by the majority vote of 3 readers per case. The ensemble comprised 4 parallel 3D ResNet50 convolutional neural network architectures trained via transfer learning and then adapted to the targeted domain. The final tear-type prediction was determined as the class with the highest probability, after averaging the class probabilities of the 4 individual models. RESULTS: The AUC overall for supraspinatus, infraspinatus, and subscapularis tendon tears was 0.93, 0.89, and 0.90, respectively. The model performed best for full-thickness supraspinatus, infraspinatus, and subscapularis tears with AUCs of 0.98, 0.99, and 0.95, respectively. Multisequence input demonstrated higher AUCs than single-sequence input for infraspinatus and subscapularis tendon tears, whereas coronal oblique fluid-sensitive and multisequence input showed similar AUCs for supraspinatus tendon tears. Model accuracy for tear types and overall accuracy were similar to that of the clinical readers. CONCLUSIONS: Deep learning diagnosis of rotator cuff tears is feasible with excellent diagnostic performance, particularly for full-thickness tears, with model accuracy similar to subspecialty-trained musculoskeletal radiologists.

Repeatability of Multiparametric Prostate MRI Radiomics Features.

In this study we assessed the repeatability of radiomics features on small prostate tumors using test-retest Multiparametric Magnetic Resonance Imaging (mpMRI). The premise of radiomics is that quantitative image-based features can serve as biomarkers for detecting and characterizing disease. For such biomarkers to be useful, repeatability is a basic requirement, meaning its value must remain stable between two scans, if the conditions remain stable. We investigated repeatability of radiomics features under various preprocessing and extraction configurations including various image normalization schemes, different image pre-filtering, and different bin widths for image discretization. Although we found many radiomics features and preprocessing combinations with high repeatability (Intraclass Correlation Coefficient > 0.85), our results indicate that overall the repeatability is highly sensitive to the processing parameters. Neither image normalization, using a variety of approaches, nor the use of pre-filtering options resulted in consistent improvements in repeatability. We urge caution when interpreting radiomics features and advise paying close attention to the processing configuration details of reported results. Furthermore, we advocate reporting all processing details in radiomics studies and strongly recommend the use of open source implementations.

Evaluation of deep learning methods for parotid gland segmentation from CT images.

The segmentation of organs at risk is a crucial and time-consuming step in radiotherapy planning. Good automatic methods can significantly reduce the time clinicians have to spend on this task. Due to its variability in shape and low contrast to surrounding structures, segmenting the parotid gland is challenging. Motivated by the recent success of deep learning, we study the use of two-dimensional (2-D), 2-D ensemble, and three-dimensional (3-D) U-Nets for segmentation. The mean Dice similarity to ground truth is ∼ 0.83 for all three models. A patch-based approach for class balancing seems promising for false-positive reduction. The 2-D ensemble and 3-D U-Net are applied to the test data of the 2015 MICCAI challenge on head and neck autosegmentation. Both deep learning methods generalize well onto independent data (Dice 0.865 and 0.88) and are superior to a selection of model- and atlas-based methods with respect to the Dice coefficient. Since appropriate reference annotations are essential for training but often difficult and expensive to obtain, it is important to know how many samples are needed for training. We evaluate the performance after training with different-sized training sets and observe no significant increase in the Dice coefficient for more than 250 training cases.

An annotated test-retest collection of prostate multiparametric MRI.

Multiparametric Magnetic Resonance Imaging (mpMRI) is widely used for characterizing prostate cancer. Standard of care use of mpMRI in clinic relies on visual interpretation of the images by an expert. mpMRI is also increasingly used as a quantitative imaging biomarker of the disease. Little is known about repeatability of such quantitative measurements, and no test-retest datasets have been available publicly to support investigation of the technical characteristics of the MRI-based quantification in the prostate. Here we present an mpMRI dataset consisting of baseline and repeat prostate MRI exams for 15 subjects, manually annotated to define regions corresponding to lesions and anatomical structures, and accompanied by region-based measurements. This dataset aims to support further investigation of the repeatability of mpMRI-derived quantitative prostate measurements, study of the robustness and reliability of the automated analysis approaches, and to support development and validation of new image analysis techniques. The manuscript can also serve as an example of the use of DICOM for standardized encoding of the image annotation and quantification results.

Visualization of Vascular and Parenchymal Regeneration after 70% Partial Hepatectomy in Normal Mice.

A modified silicone injection procedure was used for visualization of the hepatic vascular tree. This procedure consisted of in-vivo injection of the silicone compound, via a 26 G catheter, into the portal or hepatic vein. After silicone injection, organs were explanted and prepared for ex-vivo micro-CT (µCT) scanning. The silicone injection procedure is technically challenging. Achieving a successful outcome requires extensive microsurgical experience from the surgeon. One of the challenges of this procedure involves determining the adequate perfusion rate for the silicone compound. The perfusion rate for the silicone compound needs to be defined based on the hemodynamic of the vascular system of interest. Inappropriate perfusion rate can lead to an incomplete perfusion, artificial dilation and rupturing of vascular trees. The 3D reconstruction of the vascular system was based on CT scans and was achieved using preclinical software such as HepaVision. The quality of the reconstructed vascular tree was directly related to the quality of silicone perfusion. Subsequently computed vascular parameters indicative of vascular growth, such as total vascular volume, were calculated based on the vascular reconstructions. Contrasting the vascular tree with silicone allowed for subsequent histological work-up of the specimen after µCT scanning. The specimen can be subjected to serial sectioning, histological analysis and whole slide scanning, and thereafter to 3D reconstruction of the vascular trees based on histological images. This is the prerequisite for the detection of molecular events and their distribution with respect to the vascular tree. This modified silicone injection procedure can also be used to visualize and reconstruct the vascular systems of other organs. This technique has the potential to be extensively applied to studies concerning vascular anatomy and growth in various animal and disease models.

Zonated quantification of steatosis in an entire mouse liver.

Many physiological processes and pathological conditions in livers are spatially heterogeneous, forming patterns at the lobular length scale or varying across the organ. Steatosis, a common liver disease characterized by lipids accumulating in hepatocytes, exhibits heterogeneity at both these spatial scales. The main goal of the present study was to provide a method for zonated quantification of the steatosis patterns found in an entire mouse liver. As an example application, the results were employed in a pharmacokinetics simulation. For the analysis, an automatic detection of the lipid vacuoles was used in multiple slides of histological serial sections covering an entire mouse liver. Lobuli were determined semi-automatically and zones were defined within the lobuli. Subsequently, the lipid content of each zone was computed. The steatosis patterns were found to be predominantly periportal, with a notable organ-scale heterogeneity. The analysis provides a quantitative description of the extent of steatosis in unprecedented detail. The resulting steatosis patterns were successfully used as a perturbation to the liver as part of an exemplary whole-body pharmacokinetics simulation for the antitussive drug dextromethorphan. The zonated quantification is also applicable to other pathological conditions that can be detected in histological images. Besides being a descriptive research tool, this quantification could perspectively complement diagnosis based on visual assessment of histological images.

Rodent models and imaging techniques to study liver regeneration.

The liver has the unique capability of regeneration from various injuries. Different animal models and in vitro methods are used for studying the processes and mechanisms of liver regeneration. Animal models were established either by administration of hepatotoxic chemicals or by surgical approach. The administration of hepatotoxic chemicals results in the death of liver cells and in subsequent hepatic regeneration and tissue repair. Surgery includes partial hepatectomy and portal vein occlusion or diversion: hepatectomy leads to compensatory regeneration of the remnant liver lobe, whereas portal vein occlusion leads to atrophy of the ipsilateral lobe and to compensatory regeneration of the contralateral lobe. Adaptation of modern radiological imaging technologies to the small size of rodents made the visualization of rodent intrahepatic vascular anatomy possible. Advanced knowledge of the detailed intrahepatic 3D anatomy enabled the establishment of refined surgical techniques. The same technology allows the visualization of hepatic vascular regeneration. The development of modern histological image analysis tools improved the quantitative assessment of hepatic regeneration. Novel image analysis tools enable us to quantify reliably and reproducibly the proliferative rate of hepatocytes using whole-slide scans, thus reducing the sampling error. In this review, the refined rodent models and the newly developed imaging technology to study liver regeneration are summarized. This summary helps to integrate the current knowledge of liver regeneration and promises an enormous increase in hepatological knowledge in the near future.

Registration of histological whole slide images guided by vessel structures.

INTRODUCTION: The registration of histological whole slide images is an important prerequisite for modern histological image analysis. A partial reconstruction of the original volume allows e.g. colocalization analysis of tissue parameters or high-detail reconstructions of anatomical structures in 3D. METHODS: In this paper, we present an automatic staining-invariant registration method, and as part of that, introduce a novel vessel-based rigid registration algorithm using a custom similarity measure. The method is based on an iterative best-fit matching of prominent vessel structures. RESULTS: We evaluated our method on a sophisticated synthetic dataset as well as on real histological whole slide images. Based on labeled vessel structures we compared the relative differences for corresponding structures. The average positional error was close to 0, the median for the size change factor was 1, and the median overlap was 0.77. CONCLUSION: The results show that our approach is very robust and creates high quality reconstructions. The key element for the resulting quality is our novel rigid registration algorithm.

Object-based analysis of CT images for automatic detection and segmentation of hypodense liver lesions.

PURPOSE: Hypodense liver lesions are commonly detected in CT, so their segmentation and characterization are essential for diagnosis and treatment. Methods for automatic detection and segmentation of liver lesions were developed to support this task. METHODS: The detection algorithm uses an object-based image analysis approach, allowing for effectively integrating domain knowledge and reasoning processes into the detection logic. The method is intended to succeed in cases typically difficult for computer-aided detection systems, especially low contrast of hypodense lesions relative to healthy tissue. The detection stage is followed by a dedicated segmentation algorithm needed to synthesize 3D segmentations for all true-positive findings. RESULTS: The automated method provides an overall detection rate of 77.8% with a precision of 0.53 and performs better than other related methods. The final lesion segmentation delivers appropriate quality in 89% of the detected cases, as evaluated by two radiologists. CONCLUSIONS: A new automated liver lesion detection algorithm employs the strengths of an object-based image analysis approach. The combination of automated detection and segmentation provides promising results with potential to improve diagnostic liver lesion evaluation.