Laboratory testing is routinely used to assay blood biomarkers to provide information on physiologic state beyond what clinicians can evaluate from interpreting medical imaging. We hypothesized that ...deep learning interpretation of echocardiogram videos can provide additional value in understanding disease states and can evaluate common biomarkers results.
We developed EchoNet-Labs, a video-based deep learning algorithm to detect evidence of anemia, elevated B-type natriuretic peptide (BNP), troponin I, and blood urea nitrogen (BUN), as well as values of ten additional lab tests directly from echocardiograms. We included patients (n = 39,460) aged 18 years or older with one or more apical-4-chamber echocardiogram videos (n = 70,066) from Stanford Healthcare for training and internal testing of EchoNet-Lab's performance in estimating the most proximal biomarker result. Without fine-tuning, the performance of EchoNet-Labs was further evaluated on an additional external test dataset (n = 1,301) from Cedars-Sinai Medical Center. We calculated the area under the curve (AUC) of the receiver operating characteristic curve for the internal and external test datasets.
On the held-out test set of Stanford patients not previously seen during model training, EchoNet-Labs achieved an AUC of 0.80 (0.79-0.81) in detecting anemia (low hemoglobin), 0.86 (0.85-0.88) in detecting elevated BNP, 0.75 (0.73-0.78) in detecting elevated troponin I, and 0.74 (0.72-0.76) in detecting elevated BUN. On the external test dataset from Cedars-Sinai, EchoNet-Labs achieved an AUC of 0.80 (0.77-0.82) in detecting anemia, of 0.82 (0.79-0.84) in detecting elevated BNP, of 0.75 (0.72-0.78) in detecting elevated troponin I, and of 0.69 (0.66-0.71) in detecting elevated BUN. We further demonstrate the utility of the model in detecting abnormalities in 10 additional lab tests. We investigate the features necessary for EchoNet-Labs to make successful detection and identify potential mechanisms for each biomarker using well-known and novel explainability techniques.
These results show that deep learning applied to diagnostic imaging can provide additional clinical value and identify phenotypic information beyond current imaging interpretation methods.
J.W.H. and B.H. are supported by the NSF Graduate Research Fellowship. D.O. is supported by NIH K99 HL157421-01. J.Y.Z. is supported by NSF CAREER 1942926, NIH R21 MD012867-01, NIH P30AG059307 and by a Chan-Zuckerberg Biohub Fellowship.
Over the last two decades, there have been many reported advances in the clinical management of pancreatic ductal adenocarcinoma (PDAC). We sought to evaluate changes in survival for patients ...diagnosed with PDAC between 2004 and 2017. The National Cancer Database was queried for patients diagnosed with PDAC between 2004 and 2017. There were 55,401 patients who underwent surgery and 109,477 patients who underwent non-surgical treatment for PDAC between 2004 and 2017. Patients were categorized into four groups by year of diagnosis. Median survival improved from 15.5 months to 25.3 months for patients treated with surgery between the years 2016 and 2017 compared with between 2004 and 2007 (p < 0.001). Median survival improved from 7.2 months to 10.1 months for patients treated without surgery during the same years (p < 0.001). On multivariable analysis, the hazard ratio for death was estimated to multiply by 0.975 per year for patients treated with surgery and 0.959 per year for patients treated without surgery (p < 0.001). This increase in survival in the setting of evolving care validates continued efforts aimed at improving survival for patients with this devastating disease.
Cyathostomins are the most common and highly prevalent parasites of horses worldwide. Historically, the control of cyathostomins has mainly relied on the routine use of anthelmintic products. ...Increasing reports on anthelmintic resistance (AR) in cyathostomins are concerning. A potential method proposed for detecting emerging AR in cyathostomins has been estimating the egg reappearance period (ERP). This paper reviews the data available for the ERP of cyathostomins against the three major classes of anthelmintics, macrocyclic lactones, tetrahydropyrimidines, and benzimidazoles. Published peer-reviewed original research articles were obtained from three databases (PubMed, CAB Direct and Web of Science) and were evaluated for their inclusion in a systematic review. Subsets of articles were then subjected to a review of ERP data. A total of 54 (of 134) studies published between 1972 and 2022 met the criteria for inclusion in the systematic review. Until the beginning of 2022, there was no agreed definition of the ERP; eight definitions of ERP were identified in the literature, complicating the comparison between studies. Additionally, potential risk factors for the shortening of the ERP, including previous anthelmintic use and climate, were frequently not described. Reports of shortened ERP for moxidectin and ivermectin are frequent: 20 studies that used comparable ERP definitions reported shortened moxidectin and ivermectin ERPs of 35 and 28 days, respectively. It is unclear whether the ERPs of these anthelmintics reduced to such levels are due to the development of AR or some biological factors related to horses, cyathostomin species, and/or the environment. The ERPs for other anthelmintics, such as fenbendazole and pyrantel, were frequently not reported due to established resistance against these drugs. Future research in horses is required to understand the mechanism(s) behind the shortening of ERP for cyathostomins. Based on this systematic review, we propose recommendations for future ERP studies.
Display omitted
•Is the egg reappearance period (ERP) an early indicator of anthelmintic resistance?•A summary of cyathostomins ERP for equine registered anthelmintics from 1972 to 2022.•Recent developments on the definition of ERP.•Proposed recommendations for future ERP studies.
Duchenne muscular dystrophy (DMD) is a genetic disorder caused by loss of the protein dystrophin. In humans, DMD has early onset, causes developmental delays, muscle necrosis, loss of ambulation, and ...death. Current animal models have been challenged by their inability to model the early onset and severity of the disease. It remains unresolved whether increased sarcoplasmic calcium observed in dystrophic muscles follows or leads the mechanical insults caused by the muscle’s disrupted contractile machinery. This knowledge has important implications for patients, as potential physiotherapeutic treatments may either help or exacerbate symptoms, depending on how dystrophic muscles differ from healthy ones. Recently we showed how burrowing dystrophic (dys-1) C. elegans recapitulate many salient phenotypes of DMD, including loss of mobility and muscle necrosis. Here, we report that dys-1 worms display early pathogenesis, including dysregulated sarcoplasmic calcium and increased lethality. Sarcoplasmic calcium dysregulation in dys-1 worms precedes overt structural phenotypes (e.g., mitochondrial, and contractile machinery damage) and can be mitigated by reducing calmodulin expression. To learn how dystrophic musculature responds to altered physical activity, we cultivated dys-1 animals in environments requiring high intensity or high frequency of muscle exertion during locomotion. We find that several muscular parameters (e.g., size) improve with increased activity. However, longevity in dystrophic animals was negatively associated with muscular exertion, regardless of effort duration. The high degree of phenotypic conservation between dystrophic worms and humans provides a unique opportunity to gain insight into the pathology of the disease as well as the initial assessment of potential treatment strategies.
Display omitted
Despite the enormous advances in the field of clinical pancreatic islet transplantation over the past two decades, the human islet isolation procedure remains suboptimal. Islets are ...extracted (isolated) from the exocrine tissue of donor pancreases using neutral protease (NP) and collagenase-based enzymes, which digest the extracellular matrix (ECM) scaffold surrounding human islets. This process remains highly variable and current isolation enzyme blends are ineffective at digesting pancreases from younger donors with low body mass indexes (BMI). However, age-related differences in pancreatic matrix digestion have not been studied in detail at the molecular level. To address this, we investigated ECM digestion in purified ECM proteins and in pancreatic tissue sections from younger (≤30 years; n = 5) and older (>55 years; n = 5) BMI matched donors, using Raman microspectroscopy (RMS). The Raman spectral profiles for purified collagens I, IV, VI and laminins were significantly altered following controlled enzyme treatment. Pancreatic cryosections were treated with Serva collagenase, NP, or the two enzymes combined, at clinically relevant concentrations. RMS demonstrated that the ECM at the islet-exocrine interface was differentially digested with respect to donor age. The action of collagenase was affected to a greater extent than NP.
RMS is a powerful, marker-independent technology for characterising the human pancreatic ECM and demonstrating differences between donor types. Ongoing detailed studies using RMS will assist the development of donor-specific enzyme blends, increasing the overall success of human islet isolation and benefiting many people with type 1 diabetes worldwide.
Pancreatic islet transplantation is a minimally invasive treatment, which can reverse Type 1 Diabetes Mellitus (T1DM) in selected patients. Islets of Langerhans are extracted (isolated) from the exocrine tissue of human donor pancreases using neutral protease (NP) and collagenase-based enzymes, which digest the extracellular matrix (ECM) scaffold surrounding human islets. This process remains highly variable and current enzymes are ineffective at digesting pancreases from younger donors. Using Raman microspectroscopy we demonstrate that donor age affects the enzymatic digestion of the pancreatic ECM at the molecular level. Collagenase activity is affected to a greater extent than NP. These findings will assist the development of donor-specific enzymes, thereby increasing the overall success of islet isolation and benefiting many people with T1DM worldwide.
Left ventricular assist devices (LVADs) are continuous flow pumps that are employed in patients with severe heart failure. Although their emergence has significantly improved therapeutic options for ...patients with heart failure, detailed studies of the impact of LVADs on hemodynamics are notably lacking. To this end we initiate a computational study of the Jarvik 2000 LVAD model employing isogeometric fluid–structure interaction analysis. We focus on a patient-specific configuration in which the LVAD is implanted in the descending thoracic aorta. We perform computations for three pump settings and report our observations for several quantities of hemodynamic interest. It should be noted that this paper presents the first three-dimensional, patient-specific fluid–structure interaction simulation of LVADs.
The genome sequence of Helicobacter pylori suggests that this bacterium possesses several Fe acquisition systems, including both Fe2+‐ and Fe3+‐citrate transporters. The role of these transporters ...was investigated by generating insertion mutants in feoB, tonB, fecA1 and fecDE. Fe transport in the feoB mutant was ≈ 10‐fold lower than in the wild type (with 0.5 μM Fe), irrespective of whether Fe was supplied in the Fe2+ or Fe3+ form. In contrast, transport rates were unaffected by the other mutations. Complementation of the feoB mutation fully restored both Fe2+ and Fe3+ transport. The growth inhibition exhibited by the feoB mutant in Fe‐deficient media was relieved by human holo‐transferrin, holo‐lactoferrin and Fe3+‐dicitrate, but not by FeSO4. The feoB mutant had less cellular Fe and was more sensitive to growth inhibition by transition metals in comparison with the wild type. Biphasic kinetics of Fe2+ transport in the wild type suggested the presence of high‐ and low‐affinity uptake systems. The high‐affinity system (apparent Ks = 0.54 μM) is absent in the feoB mutant. Transport via FeoB is highly specific for Fe2+ and was inhibited by FCCP, DCCD and vanadate, indicating an active process energized by ATP. Ferrozine inhibition of Fe2+ and Fe3+ uptake implied the concerted involvement of both an Fe3+ reductase and FeoB in the uptake of Fe supplied as Fe3+. Taken together, the results are consistent with FeoB‐mediated Fe2+ uptake being a major pathway for H. pylori Fe acquisition. feoB mutants were unable to colonize the gastric mucosa of mice, indicating that FeoB makes an important contribution to Fe acquisition by H. pylori in the low‐pH, low‐O2 environment of the stomach.
Diabetic Macular Edema (DME) is the most common sight-threatening complication of type 2 diabetes. Optical Coherence Tomography (OCT) is the most useful imaging technique to diagnose, follow up, and ...evaluate treatments for DME. However, OCT exam and devices are expensive and unavailable in all clinics in low- and middle-income countries. Our primary goal was therefore to develop an alternative method to OCT for DME diagnosis by introducing spectral information derived from spontaneous electroretinogram (ERG) signals as a single input or combined with fundus that is much more widespread. Baseline ERGs were recorded in 233 patients and transformed into scalograms and spectrograms via Wavelet and Fourier transforms, respectively. Using transfer learning, distinct Convolutional Neural Networks (CNN) were trained as classifiers for DME using OCT, scalogram, spectrogram, and eye fundus images. Input data were randomly split into training and test sets with a proportion of 80 %–20 %, respectively. The top performers for each input type were selected, OpticNet-71 for OCT, DenseNet-201 for eye fundus, and non-evoked ERG-derived scalograms, to generate a combined model by assigning different weights for each of the selected models. Model validation was performed using a dataset alien to the training phase of the models. None of the models powered by mock ERG-derived input performed well. In contrast, hybrid models showed better results, in particular, the model powered by eye fundus combined with mock ERG-derived information with a 91 % AUC and 86 % F1-score, and the model powered by OCT and mock ERG-derived scalogram images with a 93 % AUC and 89 % F1-score. These data show that the spontaneous ERG-derived input adds predictive value to the fundus- and OCT-based models to diagnose DME, except for the sensitivity of the OCT model which remains the same. The inclusion of mock ERG signals, which have recently been shown to take only 5 min to record in daylight conditions, therefore represents a potential improvement over existing OCT-based models, as well as a reliable and cost-effective alternative when combined with the fundus, especially in underserved areas, to predict DME.
•Images of the oscillatory components of the non-evoked, baseline electroretinogram signals can be combined to fundus input to train models for diabetic macular edema diagnosis with better predictive performance than when trained only with fundus.•Alone, images of the oscillatory components of the non-evoked, baseline electroretinogram signals poorly predict diabetic macular edema.•A hybrid model powered by fundus and mock electroretinogram-derived wavelet scalogram images performs as well as the optical coherence tomography one.•Images of the oscillatory components of the non-evoked, baseline electroretinogram signals can be combined to optical coherence tomography input to train models for diabetic macular edema diagnosis with better predictive performance than when trained only with optical coherence tomography images.
Background: The presence of Lewy bodies (LB) in the neocortex and limbic system in patients with Parkinson’s disease (PD) is commonly thought to be linked with cognitive impairment. The authors ...present here a series of patients with diagnosis of PD in life and no significant cognitive impairment who, at necropsy, satisfied the current neuropathological criteria for dementia with Lewy bodies (DLB). Methods: Two hundred and seventy six brains with PD pathology were examined at the Queen Square Brain Bank in London between 1993 and 1999. The neuropathological diagnosis was PD, but 117 patients also had sufficient LB involvement above the brain stem to satisfy the current neuropathological criteria for DLB (50 patients had a neuropathological picture consistent with the limbic category of DLB and 67 with neocortical DLB). Forty eight cases were excluded who developed early cognitive impairment together with motor features of parkinsonism, 12 cases for lack of detailed clinical history, and 19 cases with coexistent features of advanced Alzheimer’s disease changes. Thirty eight patients (13.8% of the total with PD pathology and 32.5 % of the total with DLB pathology) were found where there was no or very late cognitive impairment reported in the clinical records. Results: Selected cases were 24 men and 14 women, with a mean (SD) age at onset of parkinsonian symptoms of 60.1 (10.1) years and a mean disease duration of 15.3 (5.5) years. At some time during the evolution of the disease 21 patients developed different degrees of cognitive impairment (after a mean disease duration of 12.2 (4.8) years). Clinical diagnosis at death was PD in 10 cases and PD with dementia in 11. In the remaining 17 patients no history of cognitive impairment was ever recorded in life and all of them had a clinical diagnosis of PD at death; in this subgroup, nine patients later revealed a neuropathological picture consistent with limbic (or transitional) category of DLB and eight with neocortical DLB. Interestingly, in all these patients the parkinsonian features including the response to dopaminergic drugs were indistinguishable from classic brain stem PD. Conclusions: The authors demonstrate that the classic pathology of DLB can commonly be seen outside the generally accepted clinical spectrum for DLB and that important factors other than the absolute number of LB in the neocortex and limbic system influence the development of cognitive impairment in PD. Furthermore, the pathology of PD may be indistinguishable from that reported in DLB, suggesting that the two clinicopathological syndromes may be attributable to the same biological abnormality.