Starfysh integrates spatial transcriptomic and histologic data to reveal heterogeneous tumor-immune hubs.

Spatially resolved gene expression profiling provides insight into tissue organization and cell-cell crosstalk; however, sequencing-based spatial transcriptomics (ST) lacks single-cell resolution. Current ST analysis methods require single-cell RNA sequencing data as a reference for rigorous interpretation of cell states, mostly do not use associated histology images and are not capable of inferring shared neighborhoods across multiple tissues. Here we present Starfysh, a computational toolbox using a deep generative model that incorporates archetypal analysis and any known cell type markers to characterize known or new tissue-specific cell states without a single-cell reference. Starfysh improves the characterization of spatial dynamics in complex tissues using histology images and enables the comparison of niches as spatial hubs across tissues. Integrative analysis of primary estrogen receptor (ER)-positive breast cancer, triple-negative breast cancer (TNBC) and metaplastic breast cancer (MBC) tissues led to the identification of spatial hubs with patient- and disease-specific cell type compositions and revealed metabolic reprogramming shaping immunosuppressive hubs in aggressive MBC.

Hyperspectral imaging benchmark based on machine learning for intraoperative brain tumour detection.

Brain surgery is one of the most common and effective treatments for brain tumour. However, neurosurgeons face the challenge of determining the boundaries of the tumour to achieve maximum resection, while avoiding damage to normal tissue that may cause neurological sequelae to patients. Hyperspectral (HS) imaging (HSI) has shown remarkable results as a diagnostic tool for tumour detection in different medical applications. In this work, we demonstrate, with a robust k-fold cross-validation approach, that HSI combined with the proposed processing framework is a promising intraoperative tool for in-vivo identification and delineation of brain tumours, including both primary (high-grade and low-grade) and secondary tumours. Analysis of the in-vivo brain database, consisting of 61 HS images from 34 different patients, achieve a highest median macro F1-Score result of 70.2 ± 7.9% on the test set using both spectral and spatial information. Here, we provide a benchmark based on machine learning for further developments in the field of in-vivo brain tumour detection and delineation using hyperspectral imaging to be used as a real-time decision support tool during neurosurgical workflows.

New Primers for Detection and Differentiation between Leishmania viannia and L. leishmania Subgenera by Polymerase Chain Reaction.

Background: Leishmania is the parasitic protozoan responsible for leishmaniases, a disease that can cause a range of cutaneous, mucosal, and visceral infections. Two subgenera L. Viannia and L. Leishmania are known to infect humans in the tropics and subtropics of the Americas. The aim of the present study was to develop a new pair of primers for the two subgenera and test in clinical samples. Methods: We designed two new pairs of primers for a PCR method from two conserved genes, cysteine proteinase B (cpb) and N-acetylglucosamine-6-phosfate deacetylase-like protein (nagA), as specific markers for those two respective subgenera. Primers were tested with 16 microscopical positive clinical samples from the Amazon region of Ecuador obtained in 2010-2020 period. Results: The cpb presented a band of 172 bp and the nagA a band of 300 bp, thus clearly differentiating L. viannia from L. leishmania. Additionally, primers identified and differentiated the clinical samples in the two subgenera. Conclusion: The new primers targeting different two genes and standardized in a PCR assay could identified and differentiated Leishmania parasites at subgenus level. This protocol could be used for Leishmania genus identification and diagnosis at the subgenus level and for determining the parasite's geographical distribution where different Leishmania subgenera are found in the same area.

Clustering and graph mining techniques for classification of complex structural variations in cancer genomes.

For many years, a major question in cancer genomics has been the identification of those variations that can have a functional role in cancer, and distinguish from the majority of genomic changes that have no functional consequences. This is particularly challenging when considering complex chromosomal rearrangements, often composed of multiple DNA breaks, resulting in difficulties in classifying and interpreting them functionally. Despite recent efforts towards classifying structural variants (SVs), more robust statistical frames are needed to better classify these variants and isolate those that derive from specific molecular mechanisms. We present a new statistical approach to analyze SVs patterns from 2392 tumor samples from the Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium and identify significant recurrence, which can inform relevant mechanisms involved in the biology of tumors. The method is based on recursive KDE clustering of 152,926 SVs, randomization methods, graph mining techniques and statistical measures. The proposed methodology was able not only to identify complex patterns across different cancer types but also to prove them as not random occurrences. Furthermore, a new class of pattern that was not previously described has been identified.

VNIR-NIR hyperspectral imaging fusion targeting intraoperative brain cancer detection.

Currently, intraoperative guidance tools used for brain tumor resection assistance during surgery have several limitations. Hyperspectral (HS) imaging is arising as a novel imaging technique that could offer new capabilities to delineate brain tumor tissue in surgical-time. However, the HS acquisition systems have some limitations regarding spatial and spectral resolution depending on the spectral range to be captured. Image fusion techniques combine information from different sensors to obtain an HS cube with improved spatial and spectral resolution. This paper describes the contributions to HS image fusion using two push-broom HS cameras, covering the visual and near-infrared (VNIR) [400-1000 nm] and near-infrared (NIR) [900-1700 nm] spectral ranges, which are integrated into an intraoperative HS acquisition system developed to delineate brain tumor tissue during neurosurgical procedures. Both HS images were registered using intensity-based and feature-based techniques with different geometric transformations to perform the HS image fusion, obtaining an HS cube with wide spectral range [435-1638 nm]. Four HS datasets were captured to verify the image registration and the fusion process. Moreover, segmentation and classification methods were evaluated to compare the performance results between the use of the VNIR and NIR data, independently, with respect to the fused data. The results reveal that the proposed methodology for fusing VNIR-NIR data improves the classification results up to 21% of accuracy with respect to the use of each data modality independently, depending on the targeted classification problem.

Enabling Genomics Pipelines in Commodity Personal Computers With Flash Storage.

Analysis of a patient's genomics data is the first step toward precision medicine. Such analyses are performed on expensive enterprise-class server machines because input data sets are large, and the intermediate data structures are even larger (TB-size) and require random accesses. We present a general method to perform a specific genomics problem, mutation detection, on a cheap commodity personal computer (PC) with a small amount of DRAM. We construct and access large histograms of k-mers efficiently on external storage (SSDs) and apply our technique to a state-of-the-art reference-free genomics algorithm, SMUFIN, to create SMUFIN-F. We show that on two PCs, SMUFIN-F can achieve the same throughput at only one third (36%) the hardware cost and half (45%) the energy compared to SMUFIN on an enterprise-class server. To the best of our knowledge, SMUFIN-F is the first reference-free system that can detect somatic mutations on commodity PCs for whole human genomes. We believe our technique should apply to other k-mer or n-gram-based algorithms.

Most Relevant Spectral Bands Identification for Brain Cancer Detection Using Hyperspectral Imaging.

Hyperspectral imaging (HSI) is a non-ionizing and non-contact imaging technique capable of obtaining more information than conventional RGB (red green blue) imaging. In the medical field, HSI has commonly been investigated due to its great potential for diagnostic and surgical guidance purposes. However, the large amount of information provided by HSI normally contains redundant or non-relevant information, and it is extremely important to identify the most relevant wavelengths for a certain application in order to improve the accuracy of the predictions and reduce the execution time of the classification algorithm. Additionally, some wavelengths can contain noise and removing such bands can improve the classification stage. The work presented in this paper aims to identify such relevant spectral ranges in the visual-and-near-infrared (VNIR) region for an accurate detection of brain cancer using in vivo hyperspectral images. A methodology based on optimization algorithms has been proposed for this task, identifying the relevant wavelengths to achieve the best accuracy in the classification results obtained by a supervised classifier (support vector machines), and employing the lowest possible number of spectral bands. The results demonstrate that the proposed methodology based on the genetic algorithm optimization slightly improves the accuracy of the tumor identification in ~5%, using only 48 bands, with respect to the reference results obtained with 128 bands, offering the possibility of developing customized acquisition sensors that could provide real-time HS imaging. The most relevant spectral ranges found comprise between 440.5-465.96 nm, 498.71-509.62 nm, 556.91-575.1 nm, 593.29-615.12 nm, 636.94-666.05 nm, 698.79-731.53 nm and 884.32-902.51 nm.

Deep Learning-Based Framework for In Vivo Identification of Glioblastoma Tumor using Hyperspectral Images of Human Brain.

The main goal of brain cancer surgery is to perform an accurate resection of the tumor, preserving as much normal brain tissue as possible for the patient. The development of a non-contact and label-free method to provide reliable support for tumor resection in real-time during neurosurgical procedures is a current clinical need. Hyperspectral imaging is a non-contact, non-ionizing, and label-free imaging modality that can assist surgeons during this challenging task without using any contrast agent. In this work, we present a deep learning-based framework for processing hyperspectral images of in vivo human brain tissue. The proposed framework was evaluated by our human image database, which includes 26 in vivo hyperspectral cubes from 16 different patients, among which 258,810 pixels were labeled. The proposed framework is able to generate a thematic map where the parenchymal area of the brain is delineated and the location of the tumor is identified, providing guidance to the operating surgeon for a successful and precise tumor resection. The deep learning pipeline achieves an overall accuracy of 80% for multiclass classification, improving the results obtained with traditional support vector machine (SVM)-based approaches. In addition, an aid visualization system is presented, where the final thematic map can be adjusted by the operating surgeon to find the optimal classification threshold for the current situation during the surgical procedure.

Revalidation of the RACE scale after its regional implementation in Catalonia: a triage tool for large vessel occlusion.

BACKGROUND AND PURPOSE: Our aim was to revalidate the RACE scale, a prehospital tool that aims to identify patients with large vessel occlusion (LVO), after its region-wide implementation in Catalonia, and to analyze geographical differences in access to endovascular treatment (EVT). METHODS: We used data from the prospective CICAT registry (Stroke Code Catalan registry) that includes all stroke code activations. The RACE score evaluated by emergency medical services, time metrics, final diagnosis, presence of LVO, and type of revascularization treatment were registered. Sensitivity, specificity, and area under the curve (AUC) for the RACE cut-off value ≥5 for identification of both LVO and eligibility for EVT were calculated. We compared the rate of EVT and time to EVT of patients transferred from referral centers compared with those directly presenting to comprehensive stroke centers (CSC). RESULTS: The RACE scale was evaluated in the field in 1822 patients, showing a strong correlation with the subsequent in-hospital evaluation of the National Institute of Health Stroke Scale evaluated at hospital (r=0.74, P<0.001). A RACE score ≥5 detected LVO with a sensitivity 0.84 and specificity 0.60 (AUC 0.77). Patients with RACE ≥5 harbored a LVO and received EVT more frequently than RACE <5 patients (LVO 35% vs 6%; EVT 20% vs 6%; all P<0.001). Direct admission at a CSC was independently associated with higher odds of receiving EVT compared with admission at a referral center (OR 2.40; 95% CI 1.66 to 3.46), and symtoms onset to groin puncture was 133 min shorter. CONCLUSIONS: This large validation study confirms RACE accuracy to identify stroke patients eligible for EVT, and provides evidence of geographical imbalances in the access to EVT to the detriment of patients located in remote areas.

Predictive Value of Modifications of the Prehospital Rapid Arterial Occlusion Evaluation Scale for Large Vessel Occlusion in Patients with Acute Stroke.

BACKGROUND: Prehospital clinical scales to identify patients with acute stroke with a large vessel occlusion (LVO) and direct them to an endovascular-capable stroke center are needed. We evaluated whether simplification of the Rapid Arterial oCclusion Evaluation (RACE) scale, a 5-item scale previously validated in the field, could maintain its high performance to identify patients with LVO. METHODS: Using the original prospective validation cohort of the RACE scale, 7 simpler versions of the RACE scale were designed and retrospectively recalculated for each patient. National Institutes of Health Stroke Scale score and proximal LVO were evaluated in hospital. Receiver operating characteristic analysis was performed to test performance of the simplified versions to identify LVO. For each version, the threshold with sensitivity closest to the original scale (85%) was used, and the variation in specificity and correct classification were assessed. RESULTS: The study included 341 patients with suspected stroke; 20% had LVO. The 7 simpler versions of the RACE scale had slightly lower area under the curve for detecting LVO because of lower specificity at the chosen sensitivity level. Correct classification rate decreased 9% if facial palsy was simplified or if eye or gaze deviation was removed, and decreased 4.5% if the aphasia or agnosia cortical sign was removed. CONCLUSIONS: We recommend the original RACE scale for prehospital assessment of patients with suspected stroke for its ease of use and its high performance to predict the presence of a LVO. The use of simplified versions would reduce its predictive value.

Total Cerebral Blood Flow in Patients with Cardioembolic Stroke: Is It Clinically Meaningful?

Chronic hypoperfusion may hinder the washout of emboli coming from the heart and facilitate the formation of intra-cavitary thrombi. We investigated whether a decreased total cerebral blood flow (tCBF) resulted in recurrence of stroke and other vascular events in consecutive patients with cardioembolic stroke. We excluded patients with extra-cranial carotid or vertebral stenosis. The recorded tCBF was the sum of blood flow in both the carotid and vertebral extra-cranial arteries as measured with ultrasonography. Patients were followed up to assess stroke recurrence, vascular events and mortality. We also recorded demographic data, vascular risk factors, treatment data, echocardiographic variables and the C congestive heart failure history H Hypertension history A Age D Diabetes S Sex S2 Stroke/TIA/Thromboembolism history Vasc Vascular Disease history (CHA2DS2-VASc) score. We studied 79 patients (age 77.9 ± 8.4 y). Mean tCBF was 65.5 ± 15.7 mL/100 g/min. Cox regression analysis found that CHA2 DS2-VASc score and ejection fraction were associated with tCBF. After a mean follow-up of 22 ± 8.5 mo, 7.6% of patients experienced a recurrent stroke, 12.7% experienced a vascular event and 21.5% of patients died. Clinical outcomes were not predicted by tCBF.

Transcranial diffuse optical monitoring of microvascular cerebral hemodynamics after thrombolysis in ischemic stroke.

The ultimate goal of therapeutic strategies for ischemic stroke is to reestablish the blood flow to the ischemic region of the brain. However, currently, the local cerebral hemodynamics (microvascular) is almost entirely inaccessible for stroke clinicians at the patient bed-side, and the recanalization of the major cerebral arteries (macrovascular) is the only available measure to evaluate the therapy, which does not always reflect the local conditions. Here we report the case of an ischemic stroke patient whose microvascular cerebral blood flow and oxygenation were monitored by a compact hybrid diffuse optical monitor during thrombolytic therapy. This monitor combined diffuse correlation spectroscopy and near-infrared spectroscopy. The reperfusion assessed by hybrid diffuse optics temporally correlated with the recanalization of the middle cerebral artery (assessed by transcranial-Doppler) and was in agreement with the patient outcome. This study suggests that upon further investigation, diffuse optics might have a potential for bed-side acute stroke monitoring and therapy guidance by providing hemodynamics information at the microvascular level.

Prognostic value of plasma ß-amyloid levels in patients with acute intracerebral hemorrhage.

BACKGROUND AND PURPOSE: It has been proposed that the deposition of the ß-amyloid peptide (Aß) in the brain parenchyma and brain blood vessels has deleterious effects. We tested the hypothesis that the levels of plasma Aß are related to the outcome in patients with intracerebral hemorrhage. METHODS: In a multicenter study, we prospectively included patients with spontaneous intracerebral hemorrhage within the first 24 hours after onset. At admission, we measured plasma Aß40 and Aß42 levels using ELISA techniques. Also, we recorded age, sex, vascular risk factors, National Institutes of Health Stroke Scale score, presence of intraventricular hemorrhage, localization, cause, and volume of the hematoma. We obtained the modified Rankin scale and defined a unfavorable outcome as modified Rankin scale >2 at 3 months. Bivariate and multivariate regression analyses were performed. RESULTS: We studied 160 patients (mean age, 73.8±11.3 years; 59.4% of them were men). A favorable outcome was observed in 64 (40%) of the patients. In the bivariate analyses, unfavorable outcome was associated with high age, female sex, diabetes mellitus, presence of intraventricular hemorrhage, high blood glucose, high National Institutes of Health Stroke Scale score, high volume, and high plasma levels of Aß42 and Aß40. The multivariate analysis showed that increased age (odds ratio, 1.07; 95% confidence interval, 1.035-1.21; P<0.0001), high admission National Institutes of Health Stroke Scale score (odds ratio, 1.29, 95% confidence interval, 1.17-1.42; P<0.0001), presence of diabetes mellitus (odds ratio, 4.15; 95% confidence interval, 1.21-14.1; P=0.02), and Aß42 levels >9.7 pg/mL (odds ratio, 4.11; 95% confidence interval, 1.65-10.1; P=0.02) were independently associated with an increased likelihood of an unfavorable outcome. CONCLUSIONS: High levels of plasma Aß42 in patients with acute intracerebral hemorrhage are associated with a poor functional prognosis.

Design and validation of a prehospital stroke scale to predict large arterial occlusion: the rapid arterial occlusion evaluation scale.

BACKGROUND AND PURPOSE: We aimed to develop and validate a simple prehospital stroke scale to predict the presence of large vessel occlusion (LVO) in patients with acute stroke. METHODS: The Rapid Arterial oCclusion Evaluation (RACE) scale was designed based on the National Institutes of Health Stroke Scale (NIHSS) items with a higher predictive value of LVO on a retrospective cohort of 654 patients with acute ischemic stroke: facial palsy (scored 0-2), arm motor function (0-2), leg motor function (0-2), gaze (0-1), and aphasia or agnosia (0-2). Thereafter, the RACE scale was validated prospectively in the field by trained medical emergency technicians in 357 consecutive patients transferred by Emergency Medical Services to our Comprehensive Stroke Center. Neurologists evaluated stroke severity at admission and LVO was diagnosed by transcranial duplex, computed tomography, or MR angiography. Receiver operating curve, sensitivity, specificity, and global accuracy of the RACE scale were analyzed to evaluate its predictive value for LVO. RESULTS: In the prospective cohort, the RACE scale showed a strong correlation with NIHSS (r=0.76; P<0.001). LVO was detected in 76 of 357 patients (21%). Receiver operating curves showed a similar capacity to predict LVO of the RACE scale compared with the NIHSS (area under the curve 0.82 and 0.85, respectively). A RACE scale≥5 had sensitivity 0.85, specificity 0.68, positive predictive value 0.42, and negative predictive value 0.94 for detecting LVO. CONCLUSIONS: The RACE scale is a simple tool that can accurately assess stroke severity and identify patients with acute stroke with large artery occlusion at prehospital setting by medical emergency technicians.

Microbleed burden and hematoma expansion in acute intracerebral hemorrhage.

BACKGROUND: Intracranial amyloid and hypertensive angiopathy have been related to impaired blood vessel function and the etiology of intracerebral hemorrhage (ICH). Microbleeds (MBs) are surrogate radiological markers that are associated with these underlying angiopathies. We assessed the hypothesis that MBs are associated with hematoma expansion (HE) in patients with hyperacute ICH. METHODS: We studied patients with spontaneous supratentorial ICH within the first 6 h after onset. HE was defined as an increase≥33% in the volume of hematoma on the follow-up CT in comparison with the admission CT. The volume was calculated using the ABC/2 formula. MBs were detected by specific magnetic resonance sequences (gradient-echo). The presence, number and distribution of MBs were analyzed. RESULTS: Our study included 44 patients. Their mean age was 68.9±11.1 years, and 70.5% of them were men. HE was observed in 14 of the patients (31.8%). HE was more prevalent in patients with more than 10 MBs compared with patients with 1-10 MBs (60 vs 12.5%; p=0.03). CONCLUSION: A high burden of MBs is associated with an increased risk of HE in patients with ICH. This is probably a marker of a more severe underlying angiopathy.