Tissue Classification Research Articles

DORIS: A diffusion MRI-based 10 tissue class deep learning segmentation algorithm tailored to improve anatomically-constrained tractography.

Modern tractography algorithms such as anatomically-constrained tractography (ACT) are based on segmentation maps of white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF). These maps are generally estimated from a T1-weighted (T1w) image and then registered in diffusion weighted images (DWI) space. Registration of T1w to diffusion space and partial volume estimation are challenging and rarely voxel-perfect. Diffusion-based segmentation would, thus, potentially allow not to have higher quality anatomical priors injected in the tractography process. On the other hand, even if FA-based tractography is possible without T1 registration, the literature shows that this technique suffers from multiple issues such as holes in the tracking mask and a high proportion of generated broken and anatomically implausible streamlines. Therefore, there is an important need for a tissue segmentation algorithm that works directly in the native diffusion space. We propose DORIS, a DWI-based deep learning segmentation algorithm. DORIS outputs 10 different tissue classes including WM, GM, CSF, ventricles, and 6 other subcortical structures (putamen, pallidum, hippocampus, caudate, amygdala, and thalamus). DORIS was trained and validated on a wide range of subjects, including 1,000 individuals from 22 to 90 years old from clinical and research DWI acquisitions, from 5 public databases. In the absence of a "true" ground truth in diffusion space, DORIS used a silver standard strategy from Freesurfer output registered onto the DWI. This strategy is extensively evaluated and discussed in the current study. Segmentation maps provided by DORIS are quantitatively compared to Freesurfer and FSL-fast and the impacts on tractography are evaluated. Overall, we show that DORIS is fast, accurate, and reproducible and that DORIS-based tractograms produce bundles with a longer mean length and fewer anatomically implausible streamlines.

Deep learning-based tumor microenvironment segmentation is predictive of tumor mutations and patient survival in non-small-cell lung cancer

BackgroundDespite the fact that tumor microenvironment (TME) and gene mutations are the main determinants of progression of the deadliest cancer in the world – lung cancer, their interrelations are not well understood. Digital pathology data provides a unique insight into the spatial composition of the TME. Various spatial metrics and machine learning approaches were proposed for prediction of either patient survival or gene mutations from this data. Still, these approaches are limited in the scope of analyzed features and in their explainability, and as such fail to transfer to clinical practice.MethodsHere, we generated 23,199 image patches from 26 hematoxylin-and-eosin (H&E)-stained lung cancer tissue sections and annotated them into 9 different tissue classes. Using this dataset, we trained a deep neural network ARA-CNN. Next, we applied the trained network to segment 467 lung cancer H&E images from The Cancer Genome Atlas (TCGA) database. We used the segmented images to compute human-interpretable features reflecting the heterogeneous composition of the TME, and successfully utilized them to predict patient survival and cancer gene mutations.ResultsWe achieved per-class AUC ranging from 0.72 to 0.99 for classifying tissue types in lung cancer with ARA-CNN. Machine learning models trained on the proposed human-interpretable features achieved a c-index of 0.723 in the task of survival prediction and AUC up to 73.5% for PDGFRB in the task of mutation classification.ConclusionsWe presented a framework that accurately predicted survival and gene mutations in lung adenocarcinoma patients based on human-interpretable features extracted from H&E slides. Our approach can provide important insights for designing novel cancer treatments, by linking the spatial structure of the TME in lung adenocarcinoma to gene mutations and patient survival. It can also expand our understanding of the effects that the TME has on tumor evolutionary processes. Our approach can be generalized to different cancer types to inform precision medicine strategies.

Computer-Assisted Differentiation between Colon-Mesocolon and Retroperitoneum Using Hyperspectral Imaging (HSI) Technology.

Complete mesocolic excision (CME), which involves the adequate resection of the tumor-bearing colonic segment with “en bloc” removal of its mesocolon along embryological fascial planes is associated with superior oncological outcomes. However, CME presents a higher complication rate compared to non-CME resections due to a higher risk of vascular injury. Hyperspectral imaging (HSI) is a contrast-free optical imaging technology, which facilitates the quantitative imaging of physiological tissue parameters and the visualization of anatomical structures. This study evaluates the accuracy of HSI combined with deep learning (DL) to differentiate the colon and its mesenteric tissue from retroperitoneal tissue. In an animal study including 20 pig models, intraoperative hyperspectral images of the sigmoid colon, sigmoid mesentery, and retroperitoneum were recorded. A convolutional neural network (CNN) was trained to distinguish the two tissue classes using HSI data, validated with a leave-one-out cross-validation process. The overall recognition sensitivity of the tissues to be preserved (retroperitoneum) and the tissues to be resected (colon and mesentery) was 79.0 ± 21.0% and 86.0 ± 16.0%, respectively. Automatic classification based on HSI and CNNs is a promising tool to automatically, non-invasively, and objectively differentiate the colon and its mesentery from retroperitoneal tissue.

Differentiation of glioblastoma tissues using spontaneous Raman scattering with dimensionality reduction and data classification.

The neurosurgery of intracranial tumors is often complicated by the difficulty of distinguishing tumor center, infiltration area, and normal tissue. The current standard for intraoperative navigation is fluorescent diagnostics with a fluorescent agent. This approach can be further enhanced by measuring the Raman spectrum of the tissue, which would provide additional information on its composition even in the absence of fluorescence. However, for the Raman spectra to be immediately helpful for a neurosurgeon, they must be additionally processed. In this work, we analyzed the Raman spectra of human brain glioblastoma multiforme tissue samples obtained during the surgery and investigated several approaches to dimensionality reduction and data classificatin to distinguish different types of tissues. In our study two approaches to Raman spectra dimensionality reduction were approbated and as a result we formulated new technique combining both of them: feature filtering based on the selection of those shifts which correspond to the biochemical components providing the statistically significant differences between groups of examined tissues (center of glioblastoma multiforme, tissues from infiltration area and normally appeared white matter) and principal component analysis. We applied the support vector machine to classify tissues after dimensionality reduction of registered Raman spectra. The accuracy of the classification of malignant tissues (tumor edge and center) and normal ones using the principal component analysis alone was 83% with sensitivity of 96% and specificity of 44%. With a combined technique of dimensionality reduction we obtained 83% accuracy with 77% sensitivity and 92% specificity of tumor tissues classification.

Model-based characterization platform of fiber optic extended-wavelength diffuse reflectance spectroscopy for identification of neurovascular bundles.

.Significance: Fiber-optic extended-wavelength diffuse reflectance spectroscopy (EWDRS) using both visible/near-infrared and shortwave-infrared detectors enables improved detection of spectral absorbances arising from lipids, water, and collagen and has demonstrated promise in a variety of applications, including detection of nerves and neurovascular bundles (NVB). Development of future applications of EWDRS for nerve detection could benefit from the use of model-based analyses including Monte Carlo (MC) simulations and evaluation of agreement between model systems and empirical measurements.Aim: The aim of this work is to characterize agreement between EWDRS measurements and simulations and inform future applications of model-based studies of nerve-detecting applications.Approach: A model-based platform consisting of an ex vivo microsurgical nerve dissection model, unique two-layer optical phantoms, and MC model simulations of fiber-optic EWDRS spectroscopic measurements were used to characterize EWDRS and compare agreement across models. In addition, MC simulations of an EWDRS measurement scenario are performed to provide a representative example of future analyses.Results: EWDRS studies performed in the common chicken thigh femoral nerve microsurgical dissection model indicate similar spectral features for classification of NVB versus adjacent tissues as reported in porcine models and human subjects. A comparison of measurements from unique EWDRS issue mimicking optical phantoms and MC simulations indicates high agreement between the two in homogeneous and two-layer optical phantoms, as well as in dissected tissues. Finally, MC simulations of measurement over a simulated NVB indicate the potential of future applications for measurement of nerve plexus.Conclusions: Characterization of agreement between fiber-optic EWDRS measurements and MC simulations demonstrates strong agreement across a variety of tissues and optical phantoms, offering promise for further use to guide the continued development of EWDRS for translational applications.

A deep learning approach for automatic detection, segmentation and classification of breast lesions from thermal images

PurposeThis study investigates implementation of deep learning (DL) approaches to breast tumor recognition based on thermal images. We propose to utilize Mask R-CNN technique on images by first assigning bounding boxes and then creating a border for each tumor volume to differentiate it from adjacent tissues and structures. In this manner, thermal images can be handled by a single DL model to successfully perform detection, classification, and segmentation of normal and abnormal breast tissues. MethodsThis study employs Mask R-CNN technique along with transfer learning models to accurately delineate breast volumes from adjacent tissues and structures. It is a novel study that uses a single DL model to carry out three steps of breast tumor diagnosis, namely detection, segmentation and classification of normal and abnormal tissues based on thermal images. ResultsTwo network architectures (ResNet-50 and ResNet-101) were trained for 60 epochs and were then evaluated based on their classification and segmentation performances. The testing process resulted in higher classification success for ResNet-50 backbone pretrained on COCO images (%97.1 accuracy). Detection and segmentation performances of this model were also higher with mAP of 0.921 and overlap score of 0.868. ConclusionsOur results indicate that the classification and segmentation performances of Mask R-CNN method on ResNet-50 architecture are better than the data reported in the literature for thermal breast image studies. It should be emphasized that our approach employs a single DL model that successfully performs detection, classification, and segmentation procedures for diagnosing normal and abnormal breast tissues.

Real-time Tracking and Classification of Tumor and Nontumor Tissue in Upper Gastrointestinal Cancers Using Diffuse Reflectance Spectroscopy for Resection Margin Assessment

Cancers of the upper gastrointestinal tract remain a major contributor to the global cancer burden. The accurate mapping of tumor margins is of particular importance for curative cancer resection and improvement in overall survival. Current mapping techniques preclude a full resection margin assessment in real time. To evaluate whether diffuse reflectance spectroscopy (DRS) on gastric and esophageal cancer specimens can differentiate tissue types and provide real-time feedback to the operator. This was a prospective ex vivo validation study. Patients undergoing esophageal or gastric cancer resection were prospectively recruited into the study between July 2020 and July 2021 at Hammersmith Hospital in London, United Kingdom. Tissue specimens were included for patients undergoing elective surgery for either esophageal carcinoma (adenocarcinoma or squamous cell carcinoma) or gastric adenocarcinoma. A handheld DRS probe and tracking system was used on freshly resected ex vivo tissue to obtain spectral data. Binary classification, following histopathological validation, was performed using 4 supervised machine learning classifiers. Data were divided into training and testing sets using a stratified 5-fold cross-validation method. Machine learning classifiers were evaluated in terms of sensitivity, specificity, overall accuracy, and the area under the curve. Of 34 included patients, 22 (65%) were male, and the median (range) age was 68 (35-89) years. A total of 14 097 mean spectra for normal and cancerous tissue were collected. For normal vs cancer tissue, the machine learning classifier achieved a mean (SD) overall diagnostic accuracy of 93.86% (0.66) for stomach tissue and 96.22% (0.50) for esophageal tissue and achieved a mean (SD) sensitivity and specificity of 91.31% (1.5) and 95.13% (0.8), respectively, for stomach tissue and of 94.60% (0.9) and 97.28% (0.6) for esophagus tissue. Real-time tissue tracking and classification was achieved and presented live on screen. This study provides ex vivo validation of the DRS technology for real-time differentiation of gastric and esophageal cancer from healthy tissue using machine learning with high accuracy. As such, it is a step toward the development of a real-time in vivo tumor mapping tool for esophageal and gastric cancers that can aid decision-making of resection margins intraoperatively.

Explaining and Evaluating Deep Tissue Classification by Visualizing Activations of Most Relevant Intermediate Layers

Abstract Deep Learning-based tissue classification may support pathologists in analyzing digitized whole slide images. However, in such critical tasks, only approaches that can be validated by medical experts in advance to deployment, are suitable. We present an approach that contributes to making automated tissue classification more transparent. We step beyond broadly used visualizations for last layers of a convolutional neural network by identifying most relevant intermediate layers applying Grad-CAM. A visual evaluation by a pathologist shows that these layers assign relevance, where important morphological structures are present in case of correct class decisions. We introduce a tool that can be easily used by medical experts for such validation purposes for any convolutional neural network and any layer. Visual explanations for intermediate layers provide insights into a neural network’s decision for histopathological tissue classification. In future research also the context of the input data must be considered.

Robust Colon Tissue Cartography with Semi-Supervision

Abstract We explore the task of tissue classification for colon cancer histology in a low label regime comparing a semi-supervised and a supervised learning strategy in a series of experiments. Further, we investigate the model robustness w.r.t. distribution shifts in the unlabeled data and domain shifts across different scanners to prove their practicality in a histology context. By utilizing unlabeled data in addition to nl = 1000 labeled tiles per class, we yield a substantial increase in accuracy from 89.9% to 91.4%.

PDBL: Improving Histopathological Tissue Classification With Plug-and-Play Pyramidal Deep-Broad Learning.

Histopathological tissue classification is a simpler way to achieve semantic segmentation for the whole slide images, which can alleviate the requirement of pixel-level dense annotations. Existing works mostly leverage the popular CNN classification backbones in computer vision to achieve histopathological tissue classification. In this paper, we propose a super lightweight plug-and-play module, named Pyramidal Deep-Broad Learning (PDBL), for any well-trained classification backbone to improve the classification performance without a re-training burden. For each patch, we construct a multi-resolution image pyramid to obtain the pyramidal contextual information. For each level in the pyramid, we extract the multi-scale deep-broad features by our proposed Deep-Broad block (DB-block). We equip PDBL in three popular classification backbones, ShuffLeNetV2, EfficientNetb0, and ResNet50 to evaluate the effectiveness and efficiency of our proposed module on two datasets (Kather Multiclass Dataset and the LC25000 Dataset). Experimental results demonstrate the proposed PDBL can steadily improve the tissue-level classification performance for any CNN backbones, especially for the lightweight models when given a small among of training samples (less than 10%). It greatly saves the computational resources and annotation efforts. The source code is available at: https://github.com/linjiatai/PDBL.

Detection and Analysis of Human Cells Based on Artificial Neural Network.

The detection and classification of histopathological cell images is a hot topic in current research. Medical images are an important research direction and are widely used in computer-aided diagnosis, biological research, and other fields. A neural network model based on deep learning is also common in medical image analysis and automatic detection and classification of tissue and cell images. Current medical cell detection methods generally do not consider that the yield is affected by other factors in the topological region, which leads to inevitable errors in the accuracy and generalization of the algorithm; at the same time, the current medical cell imaging methods are too simple to predict the classification markers, which affect the accuracy of cell image classification. This study introduces the concepts of two kinds of neural networks and then constructs a cell recognition model based on the convolution neural network principle and staining principle. In the experimental part, we developed three groups of experiments using the same equation as the experiment and tested the best cell recognition model proposed in this study.

Brain Tumor Classification Using Machine Learning Mixed Approach

Abstract: In this paper, we propose an effective method using Machine learning for the classification of brain tumor tissues. For successful treatment correct and early detection of brain tumors is essential. Here proposed system is using Convolutional Neural Network for feature extraction and classification. In feature extraction, we reduce the number of features in the dataset by creating new features from the existing ones. Here we recognize the types of tissues using CNN. The pooling layer is used to reduce the spatial resolution of the feature maps. This layer brings down the number of parameters needed for image processing. This paper is focused on helping the radiologist and physician to have a second opinion on the diagnosis. These systems help specialists to perform tumor detection very easily. This study aims to diagnose brain tumors using MRI images.

Breast Tumor Detection and Classification Based on Microwave Imaging

Limitations caused by traditional breast cancer detection and screening techniques have encouraged researchers to investigate alternative solutions. This study examines the use of a microwave-based approach for tumor detection in breast tissue and related tumor type classification using matched-filtering. Radar-like confocal microwave imaging (CMI) method constructs the foundation of such tumor detection approach. In particular, a microwave pulse is first transmitted, then back-scattered pulses are collected. All major reflective sites in the breast tissue are detected by repeating this procedure on a microwave pulse transmission-reception grid, aligning captured signals in-time to focus on a particular region in the breast tissue and superimposing such time-shifted signals to improve signal-to-clutter level. In the observed signals, clutter is originated by the heterogeneity of the breast tissue while signal is originated by a tumor site as a function of its water content.
 All calculations, in the study, were performed computationally in terms of a 3D Finite-Difference Time-Domain (FDTD) simulation models. For the antenna system, two cross-polarized resistively loaded bow-ties antennas were used in the computational model, and the tumor site was modeled using five different size and morphologies. Matched-filtering, on the other hand, was performed matching such obtained observations with that of a homogenous breast tissue, namely clutter-free model. Performance of the proposed approach was tested for two different antenna array resolutions, and it was observed that this parameter is important for successful detection and classification of a tumor-site in a realistic heterogenous breast tissue model.

Virtual CT Myelography: A Patch-Based Machine Learning Model to Improve Intraspinal Soft Tissue Visualization on Unenhanced Dual-Energy Lumbar Spine CT

Background: Distinguishing between the spinal cord and cerebrospinal fluid (CSF) non-invasively on CT is challenging due to their similar mass densities. We hypothesize that patch-based machine learning applied to dual-energy CT can accurately distinguish CSF from neural or other tissues based on the center voxel and neighboring voxels. Methods: 88 regions of interest (ROIs) from 12 patients’ dual-energy (100 and 140 kVp) lumbar spine CT exams were manually labeled by a neuroradiologist as one of 4 major tissue types (water, fat, bone, and nonspecific soft tissue). Four-class classifier convolutional neural networks were trained, validated, and tested on thousands of nonoverlapping patches extracted from 82 ROIs among 11 CT exams, with each patch representing pixel values (at low and high energies) of small, rectangular, 3D CT volumes. Different patch sizes were evaluated, ranging from 3 × 3 × 3 × 2 to 7 × 7 × 7 × 2. A final ensemble model incorporating all patch sizes was tested on patches extracted from six ROIs in a holdout patient. Results: Individual models showed overall test accuracies ranging from 99.8% for 3 × 3 × 3 × 2 patches (N = 19,423) to 98.1% for 7 × 7 × 7 × 2 patches (N = 1298). The final ensemble model showed 99.4% test classification accuracy, with sensitivities and specificities of 90% and 99.6%, respectively, for the water class and 98.6% and 100% for the soft tissue class. Conclusions: Convolutional neural networks utilizing local low-level features on dual-energy spine CT can yield accurate tissue classification and enhance the visualization of intraspinal neural tissue.

Development and Evaluation of a Novel Deep-Learning-Based Framework for the Classification of Renal Histopathology Images.

Kidney cancer has several types, with renal cell carcinoma (RCC) being the most prevalent and severe type, accounting for more than 85% of adult patients. The manual analysis of whole slide images (WSI) of renal tissues is the primary tool for RCC diagnosis and prognosis. However, the manual identification of RCC is time-consuming and prone to inter-subject variability. In this paper, we aim to distinguish between benign tissue and malignant RCC tumors and identify the tumor subtypes to support medical therapy management. We propose a novel multiscale weakly-supervised deep learning approach for RCC subtyping. Our system starts by applying the RGB-histogram specification stain normalization on the whole slide images to eliminate the effect of the color variations on the system performance. Then, we follow the multiple instance learning approach by dividing the input data into multiple overlapping patches to maintain the tissue connectivity. Finally, we train three multiscale convolutional neural networks (CNNs) and apply decision fusion to their predicted results to obtain the final classification decision. Our dataset comprises four classes of renal tissues: non-RCC renal parenchyma, non-RCC fat tissues, clear cell RCC (ccRCC), and clear cell papillary RCC (ccpRCC). The developed system demonstrates a high classification accuracy and sensitivity on the RCC biopsy samples at the slide level. Following a leave-one-subject-out cross-validation approach, the developed RCC subtype classification system achieves an overall classification accuracy of 93.0% ± 4.9%, a sensitivity of 91.3% ± 10.7%, and a high classification specificity of 95.6% ± 5.2%, in distinguishing ccRCC from ccpRCC or non-RCC tissues. Furthermore, our method outperformed the state-of-the-art Resnet-50 model.

Electrode–brain interface fractional order modelling for brain tissue classification in SEEG

Epileptic patients with drug resistant epilepsy can undergo invasive stereoelectroencephalography (SEEG) examination in order to identify the epileptic zone to be removed with surgery. The correct classification of the brain tissue the electrodes are inserted in is decisive for the final clinical decision. The objective of this paper is to explore differences in brain tissue via the modelling of the brain electrode interface from SEEG recordings. The proposed model is based on the physical properties of the electrode–brain interface already presented in literature, and considers the voltage of three consecutive SEEG contacts. An identification algorithm is proposed considering constraints from the physical model and prior knowledge of the values of the components and a non-parametric study. Parameter models are identified using the data of 19 different patients, and validated by visual inspection and cross-correlation tests. Differences in the resistance of white and grey matter are observed using the coefficients of the identified models. From this coefficients, an accuracy of 73±6% is achieved for heterogeneous tissue classification of triplets of contacts. The information obtained from the study of the electrode-brain interface using typical SEEG signals proposed in this paper can be used to estimate tissue composition between triplets of consecutive SEEG contacts.

Learning Representations with Contrastive Self-Supervised Learning for Histopathology Applications

Unsupervised learning has made substantial progress over the last few years, especially by means of contrastive self-supervised learning. The dominating dataset for benchmarking self-supervised learning has been ImageNet, for which recent methods are approaching the performance achieved by fully supervised training. The ImageNet dataset is however largely object-centric, and it is not clear yet what potential those methods have on widely different datasets and tasks that are not object-centric, such as in digital pathology. While self-supervised learning has started to be explored within this area with encouraging results, there is reason to look closer at how this setting differs from natural images and ImageNet. In this paper we make an in-depth analysis of contrastive learning for histopathology, pin-pointing how the contrastive objective will behave differently due to the characteristics of histopathology data. Using SimCLR and H&E stained images as a representative setting for contrastive self-supervised learning in histopathology, we bring forward a number of considerations, such as view generation for the contrastive objective and hyper-parameter tuning. In a large battery of experiments, we analyze how the downstream performance in tissue classification will be affected by these considerations. The results point to how contrastive learning can reduce the annotation effort within digital pathology, but that the specific dataset characteristics need to be considered. To take full advantage of the contrastive learning objective, different calibrations of view generation and hyper-parameters are required. Our results pave the way for realizing the full potential of self-supervised learning for histopathology applications. Code and trained models are available at <a href='https://github.com/k-stacke/ssl-pathology'>https://github.com/k-stacke/ssl-pathology</a>

Nonlinearly scaled prior image-controlled frequency split for high-frequency metal artifact reduction in computed tomography.

This paper introduces a new approach for the dedicated reduction of high-frequency metal artifacts, which applies a nonlinear scaling (NLS) transfer function on the high-frequency projection domain to reduce artifacts, while preserving edge information and anatomic detail by incorporating prior image information. An NLS function is applied to suppress high-frequency streak artifacts, but to restrict the correction to metal projections only, scaling is performed in the sinogram domain. Anatomic information should be preserved and is excluded from scaling by incorporating a prior image from tissue classification. The corrected high-frequency sinogram is reconstructed and combined with the low-frequency component of a normalized metal artifact reduction (NMAR) image. Scans of different anthropomorphic phantoms were acquired (unilateral hip, bilateral hip, dental implants, and embolization coil). Multiple regions of interest (ROIs) were drawn around the metal implants and hounsfield unit (HU) deviations were analyzed. Clinical data sets including single image slices of dental fillings, a bilateral hip implant, spinal fixation screws, and an aneurysm coil were reconstructed and assessed. The prior image-controlled NLS can remove streak artifacts while preserving anatomic detail within the bone and soft tissue. The qualitative analysis of clinical cases showed a tremendous enhancement within dental fillings and neuro coils, and a significant enhancement within spinal screws or hip implants. The phantom scan measurements support this observation. In all phantom setups, the NLS-corrected result showed lowest HU derivation and the best visualization of thedata. The prior image-controlled NLS provides a method to reduce high-frequency streaks in metal-corrupted computed tomography (CT)data.

Quantitative classification of invasive and noninvasive breast cancer using dynamic magnetic resonance imaging of the mammary gland.

ObjectivesBreast cancers are classified as invasive or noninvasive based on histopathological findings. Although time-intensity curve (TIC) analysis using magnetic resonance imaging (MRI) can differentiate benign from malignant disease, its diagnostic ability to quantitatively distinguish between invasive and noninvasive breast cancers has not been determined. In this study, we evaluated the ability of TIC analysis of dynamic MRI data (MRI-TIC) to distinguish between invasive and noninvasive breast cancers.Material and MethodsWe collected and analyzed data for 429 cases of epithelial invasive and noninvasive breast carcinomas. TIC features were extracted in washout areas suggestive of malignancy.ResultsThe graph determining the positive diagnosis rate for invasive and noninvasive cases revealed that the cut-off θi/ni value was 21.6° (invasive: θw > 21.6°, noninvasive: θw ≤ 21.6°). Tissues were classified as invasive or noninvasive using this cut-off value, and each result was compared with the histopathological diagnosis. Using this method, the accuracy of tissue classification by MRI-TIC was 88.6% (380/429), which was higher than that using ultrasound (73.4%, 315/429).ConclusionMRI-TIC is effective for the classification of invasive vs. noninvasive breast cancer.

A diagnostic index based on pseudo-continuous arterial spin labeling and T1-mapping improves efficacy in discriminating Alzheimer's disease from normal cognition.

BackgroundPseudo-continuous arterial spin labeling (pCASL) is widely used to quantify cerebral blood flow (CBF) abnormalities in patients with Alzheimer’s disease (AD). T1-mapping techniques assess microstructural characteristics in various pathologic changes, but their application in AD remains in the exploratory stage. We hypothesized that combining quantitative CBF and T1 values would generate diagnostic results with higher accuracy than using either method alone in discriminating AD patients from cognitively normal control (NC) subjects.Materials and methodsA total of 45 patients diagnosed with AD and 33 NC subjects were enrolled, and cognitive assessment was performed for each participant according to the Chinese version of the Mini-Mental State Examination (MMSE). T1-weighted magnetization-prepared 2 rapid acquisition gradient echo (MP2RAGE) and pCASL sequence were scanned on a 3T MR scanner. A brain morphometric analysis was integrated into prototype sequence, providing tissue classification and morphometric segmentation results. Quantitative CBF and T1 values of each brain region were automatically generated inline after data acquisition. Independent samples t-test was used to compare regional CBF and T1 values controlled by false discovery rate correction (corrected p < 0.01). The model with combined CBF and T1 values was compared with the individual index by performing receiver operating characteristic curves analysis. The associations between the MMSE score and CBF and T1 values of the brain were investigated using partial correlations.ResultsCerebral blood flow of the right caudate nucleus (RCc) and left hippocampus (LHc) was significantly lower in the AD group compared with the NC group, while the T1 values of the right caudate nucleus (RCt) and left hippocampus (LHt) increased in the AD group. Prediction accuracies of 73.1, 77.2, 75.9, and 81.3% were achieved for each of the above parameters, respectively. In distinguishing patients from controls using the corresponding optimized cut-off values, most combinations of parameters were elevated (area under curve = 0.775–0.894). The highest area under curve value was 0.944, by combining RCc, LHc, RCt, and LHt.ConclusionIn this preliminary study, the combined model based on pCASL and T1-mapping improved the diagnostic performance of discriminating AD and NC groups. T1-mapping may become a competitive technique for quantitatively measuring pathologic changes in the brain.