SmartPaint: a tool for interactive segmentation of medical volume images

Arterial occlusion duration affects the cuff-induced hyperemic response in skeletal muscle BOLD perfusion imaging as shown in young healthy subjects

OBJECTIVE

Dynamic BOLD MRI with cuff compression, inducing ischemia and post-occlusive hyperemia in skeletal muscle, has been pointed out as a potential diagnostic tool to assess peripheral limb perfusion. The objective was to explore the robustness of this technique and its sensitivity to the occlusion duration.

MATERIALS AND METHODS

BOLD images were acquired at 3 T in 14 healthy volunteers. [Formula: see text]-imaging with 5- and 1.5-min occlusions were acquired and several semi-quantitative BOLD parameters were derived from ROI-based [Formula: see text]-time curves. Differences in parameters from the two different occlusion durations were evaluated in the gastrocnemius and soleus muscles using non-parametrical tests. Intra- and inter-scan repeatability were evaluated with coefficient of variation.

RESULTS

Longer occlusion duration resulted in an increased hyperemic signal effect yielding significantly different values (p < 0.05) in gastrocnemius for all parameters describing the hyperemic response, and in soleus for two of these parameters. Specifically, 5-min occlusion yielded steeper hyperemic upslope in gastrocnemius (41.0%; p < 0.05) and soleus (59.7%; p = 0.03), shorter time to half peak in gastrocnemius (46.9%; p = 0.00008) and soleus (33.5%; p = 0.0003), and shorter time to peak in gastrocnemius (13.5%; p = 0.02). Coefficients of variation were lower than percentage differences that were found significant.

DISCUSSION

Findings show that the occlusion duration indeed influences the hyperemic response and thus should play a part in future methodological developments.

White matter hyperintensity mediating gait disorders in iNPH patients via neurofilament light chain

Purpose

This study aimed to analyze the differences in regional white matter hyperintensities (WMH) volume and cerebrospinal fluid biomarker levels between idiopathic normal pressure hydrocephalus (iNPH) patients with or without gait disorder.

Methods

Forty-eight iNPH patients undergoing bypass surgery and 20 normal senile individuals were included. The LST toolkit was used to segment all MRI fluid attenuation inversion images and quantify the WMH volume in each brain region. Cerebrospinal fluid was collected from all individuals and measured for concentrations of Aβ, t-tau, p-tau, and neurofilament light chain (NfL). Patients with iNPH were followed up for 1 year and divided categorized into a gait disorder improvement group and no improvement group according to the 3 m round-trip test time parameter improvement by more than 10%.

Results

We found that WMH in all areas of iNPH patients was higher than that in the control group. CSF levels of Aβ, t-tau, and p-tau were lower than those in the control group, while NfL levels were higher than those in the control group. The gait (+) group NfL level was higher than that in gait (−), and there were no statistical differences in Aβ, t-tau, and p-tau levels. The gait (+) group of frontal and parietal lobe WMH volume PVH above the gait (−) group. The mediating effect model analysis showed that PVH might affect the gait disorder of iNPH patients through NfL. A 1-year follow-up of the patients after the bypass surgery found that 24 of the 35 patients in the gait (+) group had improvements, while 11 had no significant improvements. The comparison of CSF marker levels between the two groups showed that the CSF NfL level in the improved group was lower than that in the non-improved group. The WMH volume and PVH in the frontal–parietal lobe of the improved group were lower than those of the non-improved group.

Conclusion

iNPH patients have more serious frontoparietal and periventricular white matter lesions, and WMH volume in the frontoparietal may mediate the occurrence of gait disorder in iNPH patients through the increase of NfL level.

White matter changes should not exclude patients with idiopathic normal pressure hydrocephalus from shunt surgery

Introduction

White matter changes (WMC) on brain imaging can be classified as deep white matter hyperintensities (DWMH) or periventricular hyperintensities (PVH) and are frequently seen in patients with idiopathic normal pressure hydrocephalus (iNPH). Contradictory results have been reported on whether preoperative WMC are associated with outcome after shunt surgery in iNPH patients. The aim of this study was to investigate any association between DWMH and PVH and shunt outcome in patients with iNPH, using magnetic resonance volumetry.

Methods

A total of 253 iNPH patients operated with shunt surgery and clinically assessed before and 12 months after surgery were included. All patients were investigated preoperatively with magnetic resonance imaging of the brain. The volumes of DWMH and PVH were quantified on fluid-attenuated inversion recovery images using an in-house semi-automatic volumetric segmentation software (SmartPaint). Shunt outcome was defined as the difference in symptom score between post- and preoperative investigations, measured on the iNPH scale, and shunt response was defined as improvement with ≥ 5 points.

Results

One year after shunt surgery, 51% of the patients were improved on the iNPH scale. When defining improvement as ≥ 5 points on the iNPH scale, there was no significant difference in preoperative volume of WMC between shunt responders and non-responders. If outcome was determined by a continuous variable, a larger volume of PVH was negatively associated with postoperative change in the total iNPH scale (p < 0.05) and negatively associated with improvement in gait (p < 0.01) after adjusting for age, sex, waiting time for surgery, preoperative level of symptoms, Evans’ index, and disproportionately enlarged subarachnoid space hydrocephalus. The volume of DWMH was not associated with shunt outcome.

Conclusions

An association between outcome after shunt surgery and volume of PVH was seen, but there was no difference between shunt responders and non-responders in the volumes of DWMH and PVH. We conclude that preoperative assessment of WMC should not be used to exclude patients with iNPH from shunt surgery.

IRIS—Intelligent Rapid Interactive Segmentation for Measuring Liver Cyst Volumes in Autosomal Dominant Polycystic Kidney Disease

Purpose: To develop and integrate interactive features with automatic methods for accurate liver cyst segmentation in patients with autosomal dominant polycystic kidney and liver disease (ADPKD). Methods: SmartClick and antiSmartClick were developed using iterative region growth guided by spatial and intensity connections and were integrated with automated level set (LS) segmentation and graphical user interface, forming an intelligent rapid interactive segmentation (IRIS) tool. IRIS and LS segmentations of liver cysts on T2 weighted images of patients with ADPKD (n = 17) were compared with manual segmentation as ground truth (GT). Results: Compared to manual GT, IRIS reduced the segmentation time by more than 10-fold. Compared to automated LS, IRIS reduced the mean liver cyst volume error from 42.22% to 13.44% (p < 0.001). IRIS segmentation agreed well with manual GT (79% dice score and 99% intraclass correlation coefficient). Conclusion: IRIS is feasible for fast, accurate liver cyst segmentation in patients with ADPKD.

Integration of innovative educational technologies in anatomy teaching: new normal in anatomy education

COVID-19 pandemic has created a lot of turmoil in medical teaching, the magnitude of impact is many folds in the subject of anatomy, as it is practical based. A major challenge for anatomy teachers is to replicate the experience of practical exposures. These exposures range from cadaveric dissection to demonstration of bones, museum specimens, and histology slides, where they will have interactive communication with students, and thus help in the enhancement of communication and clinical skills among them. In recent days, anatomy teachers throughout the globe started using various advanced technology to make the teaching-learning session more interesting. In pre-pandemic era, usage of such advancements in information and communication technology was a 'choice'. But pandemic has changed the situation drastically, what was a 'choice' earlier is now an 'obligation.' Presently although infection rate is low, vaccination rate is high, most of the medical schools re-opened for usual offline teaching, still body donation is all time low making the situation 'back to square one'. Keeping such unprecedented situations in mind, we need to incorporate various innovative educational technologies in day-to-day teaching-learning methodologies.

A Reperfusion BOLD-MRI Tissue Perfusion Protocol Reliably Differentiate Patients with Peripheral Arterial Occlusive Disease from Healthy Controls

There is no established technique that directly quantifies lower limb tissue perfusion. Blood oxygenation level-dependent magnetic resonance imaging (BOLD-MRI) is an MRI technique that can determine skeletal muscle perfusion. BOLD-MRI relies on magnetic differences of oxygenated and deoxygenated hemoglobin, and regional changes in oxy/deoxyhemoglobin ratio can be recorded by T2* weighted MRI sequences. We aimed to test whether BOLD-MRI can differentiate lower limb tissue perfusion in peripheral arterial occlusive disease (PAOD) patients and healthy controls. Twenty-two PAOD patients and ten healthy elderly volunteers underwent lower limb BOLD-MRI. Reactive hyperemia was provoked by transient cuff compression and images of the gastrocnemius and soleus muscles were continuously acquired at rest, during ischemia and reperfusion. Key BOLD parameters were baseline T2* absolute value and time to T2* peak value after cuff deflation (TTP). Correlations between imaging parameters and ankle-brachial index (ABI) was investigated. The mean TTP was considerably prolonged in PAOD patients compared to healthy controls (m. gastrocnemius: 111 ± 46 versus 48 ± 22 s, p = 0.000253; m. soleus: 100 ± 42 versus 41 ± 30 s, p = 0.000216). Both gastrocnemius and soleus TTP values correlated strongly with ABI (−0.82 and −0.78, p < 0.01). BOLD-MRI during reactive hyperemia differentiated most PAOD patients from healthy controls. TTP was the most decisive parameter and strongly correlated with the ABI.

Generative Adversarial Networks in Medical Image Processing

BACKGROUND

The emergence of generative adversarial networks (GANs) has provided new technology and framework for the application of medical images. Specifically, a GAN requires little to no labeled data to obtain high-quality data that can be generated through competition between the generator and discriminator networks. Therefore, GANs are rapidly proving to be a state-of-the-art foundation, achieving enhanced performances in various medical applications.

METHODS

In this article, we introduce the principles of GANs and their various variants, deep convolutional GAN, conditional GAN, Wasserstein GAN, Info-GAN, boundary equilibrium GAN, and cycle-GAN.

RESULTS

All various GANs have found success in medical imaging tasks, including medical image enhancement, segmentation, classification, reconstruction, and synthesis. Furthermore, we summarize the data processing methods and evaluation indicators. Finally, we note the limitations of existing methods and the existing challenges that need to be addressed in this field.

CONCLUSION

Although GANs are in the initial stage of development in medical image processing, it will have a great prospect in the future.

TissueWand, a Rapid Histopathology Annotation Tool

Background:

Recent advancements in machine learning (ML) bring great possibilities for the development of tools to assist with diagnostic tasks within histopathology. However, these approaches typically require a large amount of ground truth training data in the form of image annotations made by human experts. As such annotation work is a very time-consuming task, there is a great need for tools that can assist in this process, saving time while not sacrificing annotation quality.

Methods:

In an iterative design process, we developed TissueWand – an interactive tool designed for efficient annotation of gigapixel-sized histopathological images, not being constrained to a predefined annotation task.

Results:

Several findings regarding appropriate interaction concepts were made, where a key design component was semi-automation based on rapid interaction feedback in a local region. In a user study, the resulting tool was shown to cause substantial speed-up compared to manual work while maintaining quality.

Conclusions:

The TissueWand tool shows promise to replace manual methods for early stages of dataset curation where no task-specific ML model yet exists to aid the effort.

The SGLT2 Inhibitor Dapagliflozin Reduces Liver Fat but Does Not Affect Tissue Insulin Sensitivity: A Randomized, Double-Blind, Placebo-Controlled Study With 8-Week Treatment in Type 2 Diabetes Patients

OBJECTIVE

The aim of this study was to investigate tissue-specific effects of dapagliflozin on insulin sensitivity and liver and body fat in patients with type 2 diabetes.

RESEARCH DESIGN AND METHODS

This randomized, double-blind, parallel group, placebo-controlled study recruited 32 patients with type 2 diabetes. Enrolled patients were to have HbA_1c 6.5-10.5% (48-91 mmol/mol) and ≥3 months of stable treatment with metformin, dipeptidyl peptidase 4 inhibitor, or their combination. Patients were randomized 1:1 to receive 10 mg dapagliflozin or placebo daily for 8 weeks. Before and after the intervention, tissue insulin sensitivity was measured using [¹⁸F]-fluorodeoxyglucose and positron emission tomography during hyperinsulinemic-euglycemic clamp. Liver proton density fat fraction (PDFF) and adipose tissue volumes were assessed using MRI, and blood biomarkers were analyzed.

RESULTS

After 8 weeks, glycemic control was improved by dapagliflozin (placebo-corrected change in HbA_1c -0.39%, P < 0.01), but whole-body glucose uptake was not increased (P = 0.90). Tissue-specific insulin-stimulated glucose uptake did not change in skeletal muscle, liver, myocardium, or white and brown adipose tissue, and endogenous glucose production remained unaffected. However, there were significant placebo-corrected decreases in liver PDFF (-3.74%, P < 0.01), liver volume (-0.10 L, P < 0.05), visceral adipose tissue volume (-0.35 L, P < 0.01), interleukin-6 (-1.87 pg/mL, P < 0.05), and N-terminal prohormone of brain natriuretic peptide (-96 ng/L, P = 0.03).

CONCLUSIONS

In this study, 8 weeks of treatment with dapagliflozin reduced liver PDFF and the volume of visceral adipose tissue in obese patients with type 2 diabetes. Although glycemic control was improved, no effect on tissue-level insulin sensitivity was observed.

Fully convolutional networks for automated segmentation of abdominal adipose tissue depots in multicenter water-fat MRI

PURPOSE

An approach for the automated segmentation of visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) in multicenter water-fat MRI scans of the abdomen was investigated, using 2 different neural network architectures.

METHODS

The 2 fully convolutional network architectures U-Net and V-Net were trained, evaluated, and compared using the water-fat MRI data. Data of the study Tellus with 90 scans from a single center was used for a 10-fold cross-validation in which the most successful configuration for both networks was determined. These configurations were then tested on 20 scans of the multicenter study beta-cell function in JUvenile Diabetes and Obesity (BetaJudo), which involved a different study population and scanning device.

RESULTS

The U-Net outperformed the used implementation of the V-Net in both cross-validation and testing. In cross-validation, the U-Net reached average dice scores of 0.988 (VAT) and 0.992 (SAT). The average of the absolute quantification errors amount to 0.67% (VAT) and 0.39% (SAT). On the multicenter test data, the U-Net performs only slightly worse, with average dice scores of 0.970 (VAT) and 0.987 (SAT) and quantification errors of 2.80% (VAT) and 1.65% (SAT).

CONCLUSION

The segmentations generated by the U-Net allow for reliable quantification and could therefore be viable for high-quality automated measurements of VAT and SAT in large-scale studies with minimal need for human intervention. The high performance on the multicenter test data furthermore shows the robustness of this approach for data of different patient demographics and imaging centers, as long as a consistent imaging protocol is used.

Effects of free omega-3 carboxylic acids and fenofibrate on liver fat content in patients with hypertriglyceridemia and non-alcoholic fatty liver disease: A double-blind, randomized, placebo-controlled study

BACKGROUND

Treatment with omega-3 fatty acids and fenofibrates reduces serum triglyceride levels, but few studies have compared the effect of these agents on liver fat.

OBJECTIVE

The aim of the EFFECT I trial (NCT02354976) was to determine the effects of free omega-3 carboxylic acids (OM-3CA) and fenofibrate on liver fat in overweight or obese individuals with non-alcoholic fatty liver disease and hypertriglyceridemia.

METHODS

Seventy-eight patients were randomized to receive oral doses of 4 g OM-3CA (n = 25), 200 mg fenofibrate (n = 27), or placebo (n = 26) for 12 weeks in a double-blind, parallel-group study. Liver proton density fat fraction (PDFF) and volume, pancreas volume, and adipose tissue volumes were assessed by magnetic resonance imaging.

RESULTS

Changes in liver PDFF at 12 weeks were not significantly different across treatment groups (relative changes from baseline: placebo, +4%; OM-3CA, -2%; and fenofibrate, +17%). The common PNPLA3 genetic polymorphism (I148M) did not significantly influence the effects of OM-3CA or fenofibrate on liver PDFF. Fenofibrate treatment significantly increased liver and pancreas volumes vs placebo treatment, and the changes in liver and pancreas volumes were positively correlated (rho 0.45, P = .02). Total liver fat volume increased significantly in patients using fenofibrate vs OM-3CA (+23% vs -3%, P = .04). Compared with OM-3CA, fenofibrate increased total liver fat and liver volume. Serum triglycerides decreased with OM-3CA (-26%, P = .02) and fenofibrate (-38%, P < .001) vs placebo. In contrast to OM-3CA, fenofibrate reduced plasma docosahexaenoic acid levels and increased plasma acetylcarnitine and butyrylcarnitine levels, estimated delta-9 desaturase activity and the concentration of urine F2-isoprostanes.

CONCLUSIONS

OM-3CA and fenofibrate reduced serum triglycerides but did not reduce liver fat. Fenofibrate increased total liver volume and total liver fat volume vs OM-3CA, indicating a complex effect of fenofibrate on human hepatic lipid metabolism.

Liver segmentation: indications, techniques and future directions

OBJECTIVES

Liver volumetry has emerged as an important tool in clinical practice. Liver volume is assessed primarily via organ segmentation of computed tomography (CT) and magnetic resonance imaging (MRI) images. The goal of this paper is to provide an accessible overview of liver segmentation targeted at radiologists and other healthcare professionals.

METHODS

Using images from CT and MRI, this paper reviews the indications for liver segmentation, technical approaches used in segmentation software and the developing roles of liver segmentation in clinical practice.

RESULTS

Liver segmentation for volumetric assessment is indicated prior to major hepatectomy, portal vein embolisation, associating liver partition and portal vein ligation for staged hepatectomy (ALPPS) and transplant. Segmentation software can be categorised according to amount of user input involved: manual, semi-automated and fully automated. Manual segmentation is considered the "gold standard" in clinical practice and research, but is tedious and time-consuming. Increasingly automated segmentation approaches are more robust, but may suffer from certain segmentation pitfalls. Emerging applications of segmentation include surgical planning and integration with MRI-based biomarkers.

CONCLUSIONS

Liver segmentation has multiple clinical applications and is expanding in scope. Clinicians can employ semi-automated or fully automated segmentation options to more efficiently integrate volumetry into clinical practice.

TEACHING POINTS

• Liver volume is assessed via organ segmentation on CT and MRI examinations. • Liver segmentation is used for volume assessment prior to major hepatic procedures. • Segmentation approaches may be categorised according to the amount of user input involved. • Emerging applications include surgical planning and integration with MRI-based biomarkers.

Accurate Lungs Segmentation on CT Chest Images by Adaptive Appearance-Guided Shape Modeling

To accurately segment pathological and healthy lungs for reliable computer-aided disease diagnostics, a stack of chest CT scans is modeled as a sample of a spatially inhomogeneous joint 3D Markov-Gibbs random field (MGRF) of voxel-wise lung and chest CT image signals (intensities). The proposed learnable MGRF integrates two visual appearance sub-models with an adaptive lung shape submodel. The first-order appearance submodel accounts for both the original CT image and its Gaussian scale space (GSS) filtered version to specify local and global signal properties, respectively. Each empirical marginal probability distribution of signals is closely approximated with a linear combination of discrete Gaussians (LCDG), containing two positive dominant and multiple sign-alternate subordinate DGs. The approximation is separated into two LCDGs to describe individually the lungs and their background, i.e., all other chest tissues. The second-order appearance submodel quantifies conditional pairwise intensity dependencies in the nearest voxel 26-neighborhood in both the original and GSS-filtered images. The shape submodel is built for a set of training data and is adapted during segmentation using both the lung and chest appearances. The accuracy of the proposed segmentation framework is quantitatively assessed using two public databases (ISBI VESSEL12 challenge and MICCAI LOLA11 challenge) and our own database with, respectively, 20, 55, and 30 CT images of various lung pathologies acquired with different scanners and protocols. Quantitative assessment of our framework in terms of Dice similarity coefficients, 95-percentile bidirectional Hausdorff distances, and percentage volume differences confirms the high accuracy of our model on both our database (98.4_±1.0%, 2.2_±1.0 mm, 0.42_±0.10%) and the VESSEL12 database (99.0_±0.5%, 2.1_±1.6 mm, 0.39_±0.20%), respectively. Similarly, the accuracy of our approach is further verified via a blind evaluation by the organizers of the LOLA11 competition, where an average overlap of 98.0% with the expert's segmentation is yielded on all 55 subjects with our framework being ranked first among all the state-of-the-art techniques compared.

Dense breast tissue in postmenopausal women is associated with a pro-inflammatory microenvironment in vivo

Inflammation is one of the hallmarks of carcinogenesis. High mammographic density has been associated with increased risk of breast cancer but the mechanisms behind are poorly understood. We evaluated whether breasts with different mammographic densities exhibited differences in the inflammatory microenvironment. Postmenopausal women attending the mammography-screening program were assessed having extreme dense, n = 20, or entirely fatty breasts (nondense), n = 19, on their regular mammograms. Thereafter, the women were invited for magnetic resonance imaging (MRI), microdialysis for the collection of extracellular molecules in situ and a core tissue biopsy for research purposes. On the MRI, lean tissue fraction (LTF) was calculated for a continuous measurement of breast density. LTF confirmed the selection from the mammograms and gave a continuous measurement of breast density. Microdialysis revealed significantly increased extracellular in vivo levels of IL-6, IL-8, vascular endothelial growth factor, and CCL5 in dense breast tissue as compared with nondense breasts. Moreover, the ratio IL-1Ra/IL-1β was decreased in dense breasts. No differences were found in levels of IL-1β, IL-1Ra, CCL2, leptin, adiponectin, or leptin:adiponectin ratio between the two breast tissue types. Significant positive correlations between LTF and the pro-inflammatory cytokines as well as between the cytokines were detected. Stainings of the core biopsies exhibited increased levels of immune cells in dense breast tissue. Our data show that dense breast tissue in postmenopausal women is associated with a pro-inflammatory microenvironment and, if confirmed in a larger cohort, suggests novel targets for prevention therapies for women with dense breast tissue.