Identifying primary tumor site of origin for liver metastases via a combination of handcrafted and deep learning features.

Liver is one of the most common sites for metastases, which can occur on account of primary tumors from multiple sites of origin. Identifying the primary site of origin (PSO) of a metastasis can help in guiding therapeutic options for liver metastases. In this pilot study, we hypothesized that computer extracted handcrafted (HC) histomorphometric features can be utilized to identify the PSO of liver metastases. Cellular features, including tumor nuclei morphological and graph features as well as cytoplasm texture features, were extracted by computer algorithms from 175 slides (114 patients). The study comprised three experiments: (1) comparing and (2) fusing a machine learning (ML) model trained with HC pathomic features and deep learning (DL)-based classifiers to predict site of origin; (3) identifying the section of the primary tumor from which metastases were derived. For experiment 1, we divided the cohort into training sets composed of primary and matched liver metastases [60 patients, 121 whole slide images (WSIs)], and a hold-out validation set (54 patients, 54 WSIs) composed solely of liver metastases of known site of origin. Using the extracted HC features of the training set, a combination of supervised machine classifiers and unsupervised clustering was applied to identify the PSO. A random forest classifier achieved areas under the curve (AUCs) of 0.83, 0.64, 0.82, and 0.64 in classifying the metastatic tumor from colon, esophagus, breast, and pancreas on the validation set. The top features related to nuclear and peri-nuclear shape and textural attributes. We also trained a DL network to serve as a direct comparison to our method. The DL model achieved AUCs for colon: 0.94, esophagus: 0.66, breast: 0.79, and pancreas: 0.67 in identifying PSO. A decision fusion-based strategy was deployed to fuse the trained ML and DL classifiers and achieved slightly better results than ML or DL classifier alone (colon: 0.93, esophagus: 0.68, breast: 0.81, and pancreas: 0.69). For the third experiment, WSI-level attention maps were also generated using a trained DL network to generate a composite feature similarity heat map between paired primaries and their associated metastases. Our experiments revealed that epithelium-rich and moderately differentiated tumor regions of primary tumors were quantitatively similar to paired metastatic tumors. Our findings suggest that a combination of HC and DL features could potentially help identify the PSO for liver metastases while at the same time also potentially identify the spatial sites of origin for the metastases within primary tumors.

A Probability Fusion Approach for Foot Placement Prediction in Complex Terrains.

Prediction of foot placement presents great potential in better assisting the walking of people with lower-limb disability in daily terrains. Previous researches mainly focus on foot placement prediction in level ground walking, however these methods cannot be applied to daily complex terrains including ramps, stairs, and level ground with obstacles. To predict foot placement in complex terrains, this paper presents a probability fusion approach for foot placement prediction in complex terrains which consists of two parts: model training and foot placement prediction. In the first part, a deep learning model is trained on augmented data to predict the probability distribution of preliminary foot placement. In the second part, environmental information and human walking constraints are used to calculate the feasible area, and finally the feasible area is fused with the probability distribution of preliminary foot placement to predict the foot placement in complex terrains. The proposed method can predict the foot placement of next step in complex terrains when heel-off is detected. Experiments (including structured terrains experiments and complex terrains experiments) show that the root mean square error (RMSE) of prediction is 8.19 ± 1.20 cm, which is less than 8% of the average stride length, and the landing feasible area accuracy (LFAA) of prediction is 95.11 ± 3.09%. Comparing with existing foot placement prediction studies, the method proposed in this paper achieves faster and more accurate prediction in complex terrains.

Local Adaptive Image Filtering Based on Recursive Dilation Segmentation.

This paper introduces a simple but effective image filtering method, namely, local adaptive image filtering (LAIF), based on an image segmentation method, i.e., recursive dilation segmentation (RDS). The algorithm is motivated by the observation that for the pixel to be smoothed, only the similar pixels nearby are utilized to obtain the filtering result. Relying on this observation, similar pixels are partitioned by RDS before applying a locally adaptive filter to smooth the image. More specifically, by directly taking the spatial information between adjacent pixels into consideration in a recursive dilation way, RDS is firstly proposed to partition the guided image into several regions, so that the pixels belonging to the same segmentation region share a similar property. Then, guided by the iterative segmented results, the input image can be easily filtered via a local adaptive filtering technique, which smooths each pixel by selectively averaging its local similar pixels. It is worth mentioning that RDS makes full use of multiple integrated information including pixel intensity, hue information, and especially spatial adjacent information, leading to more robust filtering results. In addition, the application of LAIF in the remote sensing field has achieved outstanding results, specifically in areas such as image dehazing, denoising, enhancement, and edge preservation, among others. Experimental results show that the proposed LAIF can be successfully applied to various filtering-based tasks with favorable performance against state-of-the-art methods.

Unsupervised Sim-to-Real Adaptation for Environmental Recognition in Assistive Walking.

Powered lower-limb prostheses with vision sensors are expected to restore amputees' mobility in various environments with supervised learning-based environmental recognition. Due to the sim-to-real gap, such as real-world unstructured terrains and the perspective and performance limitations of vision sensor, simulated data cannot meet the requirement for supervised learning. To mitigate this gap, this paper presents an unsupervised sim-to-real adaptation method to accurately classify five common real-world (level ground, stair ascent, stair descent, ramp ascent and ramp descent) and assist amputee's terrain-adaptive locomotion. In this study, augmented simulated environments are generated from a virtual camera perspective to better simulate the real world. Then, unsupervised domain adaptation is incorporated to train the proposed adaptation network consisting of a feature extractor and two classifiers is trained on simulated data and unlabeled real-world data to minimize domain shift between source domain (simulation) and target domain (real world). To interpret the classification mechanism visually, essential features of different terrains extracted by the network are visualized. The classification results in walking experiments indicate that the average accuracy on eight subjects reaches (98.06% ± 0.71 %) and (95.91% ± 1.09 %) in indoor and outdoor environments respectively, which is close to the result of supervised learning using both type of labeled data (98.37% and 97.05%). The promising results demonstrate that the proposed method is expected to realize accurate real-world environmental classification and successful sim-to-real transfer.

Computer-extracted features of nuclear morphology in hematoxylin and eosin images distinguish stage II and IV colon tumors.

We assessed the utility of quantitative features of colon cancer nuclei, extracted from digitized hematoxylin and eosin-stained whole slide images (WSIs), to distinguish between stage II and stage IV colon cancers. Our discovery cohort comprised 100 stage II and stage IV colon cancer cases sourced from the University Hospitals Cleveland Medical Center (UHCMC). We performed initial (independent) model validation on 51 (143) stage II and 79 (54) stage IV colon cancer cases from UHCMC (The Cancer Genome Atlas's Colon Adenocarcinoma, TCGA-COAD, cohort). Our approach comprised the following steps: (1) a fully convolutional deep neural network with VGG-18 architecture was trained to locate cancer on WSIs; (2) another deep-learning model based on Mask-RCNN with Resnet-50 architecture was used to segment all nuclei from within the identified cancer region; (3) a total of 26 641 quantitative morphometric features pertaining to nuclear shape, size, and texture were extracted from within and outside tumor nuclei; (4) a random forest classifier was trained to distinguish between stage II and stage IV colon cancers using the five most discriminatory features selected by the Wilcoxon rank-sum test. Our trained classifier using these top five features yielded an AUC of 0.81 and 0.78, respectively, on the held-out cases in the UHCMC and TCGA validation sets. For 197 TCGA-COAD cases, the Cox proportional hazards model yielded a hazard ratio of 2.20 (95% CI 1.24-3.88) with a concordance index of 0.71, using only the top five features for risk stratification of overall survival. The Kaplan-Meier estimate also showed statistically significant separation between the low-risk and high-risk patients, with a log-rank P value of 0.0097. Finally, unsupervised clustering of the top five features revealed that stage IV colon cancers with peritoneal spread were morphologically more similar to stage II colon cancers with no long-term metastases than to stage IV colon cancers with hematogenous spread. © 2022 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Dynamic contrast-enhanced CEST MRI using a low molecular weight dextran.

Natural and synthetic sugars have great potential for developing highly biocompatible and translatable chemical exchange saturation transfer (CEST) MRI contrast agents. In this study, we aimed to develop the smallest clinically available form of dextran, Dex1 (molecular weight, MW ~ 1 kDa), as a new CEST agent. We first characterized the CEST properties of Dex1 in vitro at 11.7 T and showed that the Dex1 had a detectable CEST signal at ~1.2 ppm, attributed to hydroxyl protons. In vivo CEST MRI studies were then carried out on C57BL6 mice bearing orthotopic GL261 brain tumors (n = 5) using a Bruker BioSpec 11.7 T MRI scanner. Both steady-state full Z-spectral images and single offset (1.2 ppm) dynamic dextran-enhanced (DDE) images were acquired before and after the intravenous injection of Dex1 (2 g/kg). The steady-state Z-spectral analysis showed a significantly higher CEST contrast enhancement in the tumor than in contralateral brain (∆MTRasym1.2 ppm  = 0.010 ± 0.006 versus 0.002 ± 0.008, P = 0.0069) at 20 min after the injection of Dex1. Pharmacokinetic analyses of DDE were performed using the area under the curve (AUC) in the first 10 min after Dex1 injection, revealing a significantly higher uptake of Dex1 in the tumor than in brain tissue for tumor-bearing mice (AUC[0-10 min] = 21.9 ± 4.2 versus 5.3 ± 6.4%·min, P = 0.0294). In contrast, no Dex1 uptake was foundling in the brains of non-tumor-bearing mice (AUC[0-10 min] = -1.59 ± 2.43%·min). Importantly, the CEST MRI findings were consistent with the measurements obtained using DCE MRI and fluorescence microscopy, demonstrating the potential of Dex1 as a highly translatable CEST MRI contrast agent for assessing tumor hemodynamics.

Molecular Imaging of Deoxycytidine Kinase Activity Using Deoxycytidine-Enhanced CEST MRI.

Deoxycytidine kinase (DCK) is a key enzyme for the activation of a broad spectrum of nucleoside-based chemotherapy drugs (e.g., gemcitabine); low DCK activity is one of the most important causes of cancer drug-resistance. Noninvasive imaging methods that can quantify DCK activity are invaluable for assessing tumor resistance and predicting treatment efficacy. Here we developed a "natural" MRI approach to detect DCK activity using its natural substrate deoxycytidine (dC) as the imaging probe, which can be detected directly by chemical exchange saturation transfer (CEST) MRI without any synthetic labeling. CEST MRI contrast of dC and its phosphorylated form, dCTP, successfully discriminated DCK activity in two mouse leukemia cell lines with different DCK expression. This dC-enhanced CEST MRI in xenograft leukemic cancer mouse models demonstrated that DCK(+) tumors have a distinctive dynamic CEST contrast enhancement and a significantly higher CEST contrast than DCK(-) tumors (AUC0-60 min = 0.47 ± 0.25 and 0.20 ± 0.13, respectively; P = 0.026, paired Student t test, n = 4) at 1 hour after the injection of dC. dC-enhanced CEST contrast also correlated well with tumor responses to gemcitabine treatment. This study demonstrates a novel MR molecular imaging approach for predicting cancer resistance using natural, nonradioactive, nonmetallic, and clinically available agents. This method has great potential for pursuing personalized chemotherapy by stratifying patients with different DCK activity. SIGNIFICANCE: A new molecular MRI method that detects deoxycytidine kinase activity using its natural substrate deoxycytidine has great translational potential for clinical assessment of tumor resistance and prediction of treatment efficacy.

Author Correction: Targeted gadofullerene for sensitive magnetic resonance imaging and risk-stratification of breast cancer.

In the original version of this Article, the penultimate sentence of the Abstract incorrectly read 'The dose of the contrast agent for effective molecular MRI is only slightly lower than that of ZD2-Cy5.5 (0.5 µmol kg-1) in fluorescence imaging.' The correct version states 'higher' in place of 'lower'. This error has been corrected in both the PDF and HTML versions of the Article.

Targeted gadofullerene for sensitive magnetic resonance imaging and risk-stratification of breast cancer.

Molecular imaging of cancer biomarkers is critical for non-invasive accurate cancer detection and risk-stratification in precision healthcare. A peptide-targeted tri-gadolinium nitride metallofullerene, ZD2-Gd3N@C80, is synthesised for sensitive molecular magnetic resonance imaging of extradomain-B fibronectin in aggressive tumours. ZD2-Gd3N@C80 has superior r 1 and r 2 relaxivities of 223.8 and 344.7 mM-1 s-1 (1.5 T), respectively. It generates prominent contrast enhancement in aggressive MDA-MB-231 triple negative breast cancer in mice at a low dose (1.7 µmol kg-1, 1 T), but not in oestrogen receptor-positive MCF-7 tumours. Strong tumour contrast enhancement is consistently observed in other triple negative breast cancer models, but not in low-risk slow-growing tumours. The dose of the contrast agent for effective molecular MRI is only slightly lower than that of ZD2-Cy5.5 (0.5 µmol kg-1) in fluorescence imaging. These results demonstrate that high-sensitivity molecular magnetic resonance imaging with ZD2-Gd3N@C80 may provide accurate detection and risk-stratification of high-risk tumours for precision healthcare of breast cancer.Molecular MRI is a powerful clinical tool for non-invasive detection of cancer biomarkers. Here, the authors develop a targeted peptide gadofullerene contrast agent that can sensitively distinguish between breast cancers of different aggressiveness.