Machine learning prediction of pathological complete response and overall survival of breast cancer patients in an underserved inner-city population.

BACKGROUND: Generalizability of predictive models for pathological complete response (pCR) and overall survival (OS) in breast cancer patients requires diverse datasets. This study employed four machine learning models to predict pCR and OS up to 7.5 years using data from a diverse and underserved inner-city population. METHODS: Demographics, staging, tumor subtypes, income, insurance status, and data from radiology reports were obtained from 475 breast cancer patients on neoadjuvant chemotherapy in an inner-city health system (01/01/2012 to 12/31/2021). Logistic regression, Neural Network, Random Forest, and Gradient Boosted Regression models were used to predict outcomes (pCR and OS) with fivefold cross validation. RESULTS: pCR was not associated with age, race, ethnicity, tumor staging, Nottingham grade, income, and insurance status (p > 0.05). ER-/HER2+ showed the highest pCR rate, followed by triple negative, ER+/HER2+, and ER+/HER2- (all p < 0.05), tumor size (p < 0.003) and background parenchymal enhancement (BPE) (p < 0.01). Machine learning models ranked ER+/HER2-, ER-/HER2+, tumor size, and BPE as top predictors of pCR (AUC = 0.74-0.76). OS was associated with race, pCR status, tumor subtype, and insurance status (p < 0.05), but not ethnicity and incomes (p > 0.05). Machine learning models ranked tumor stage, pCR, nodal stage, and triple-negative subtype as top predictors of OS (AUC = 0.83-0.85). When grouping race and ethnicity by tumor subtypes, neither OS nor pCR were different due to race and ethnicity for each tumor subtype (p > 0.05). CONCLUSION: Tumor subtypes and imaging characteristics were top predictors of pCR in our inner-city population. Insurance status, race, tumor subtypes and pCR were associated with OS. Machine learning models accurately predicted pCR and OS.

Long-term outcomes of hospitalized patients with SARS-CoV-2/COVID-19 with and without neurological involvement: 3-year follow-up assessment.

BACKGROUND: Acute neurological manifestation is a common complication of acute Coronavirus Disease 2019 (COVID-19) disease. This retrospective cohort study investigated the 3-year outcomes of patients with and without significant neurological manifestations during initial COVID-19 hospitalization. METHODS AND FINDINGS: Patients hospitalized for Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection between 03/01/2020 and 4/16/2020 in the Montefiore Health System in the Bronx, an epicenter of the early pandemic, were included. Follow-up data was captured up to 01/23/2023 (3 years post-COVID-19). This cohort consisted of 414 patients with COVID-19 with significant neurological manifestations and 1,199 propensity-matched patients (for age and COVID-19 severity score) with COVID-19 without neurological manifestations. Neurological involvement during the acute phase included acute stroke, new or recrudescent seizures, anatomic brain lesions, presence of altered mentation with evidence for impaired cognition or arousal, and neuro-COVID-19 complex (headache, anosmia, ageusia, chemesthesis, vertigo, presyncope, paresthesias, cranial nerve abnormalities, ataxia, dysautonomia, and skeletal muscle injury with normal orientation and arousal signs). There were no significant group differences in female sex composition (44.93% versus 48.21%, p = 0.249), ICU and IMV status, white, not Hispanic (6.52% versus 7.84%, p = 0.380), and Hispanic (33.57% versus 38.20%, p = 0.093), except black non-Hispanic (42.51% versus 36.03%, p = 0.019). Primary outcomes were mortality, stroke, heart attack, major adverse cardiovascular events (MACE), reinfection, and hospital readmission post-discharge. Secondary outcomes were neuroimaging findings (hemorrhage, active and prior stroke, mass effect, microhemorrhages, white matter changes, microvascular disease (MVD), and volume loss). More patients in the neurological cohort were discharged to acute rehabilitation (10.39% versus 3.34%, p < 0.001) or skilled nursing facilities (35.75% versus 25.35%, p < 0.001) and fewer to home (50.24% versus 66.64%, p < 0.001) than matched controls. Incidence of readmission for any reason (65.70% versus 60.72%, p = 0.036), stroke (6.28% versus 2.34%, p < 0.001), and MACE (20.53% versus 16.51%, p = 0.032) was higher in the neurological cohort post-discharge. Per Kaplan-Meier univariate survival curve analysis, such patients in the neurological cohort were more likely to die post-discharge compared to controls (hazard ratio: 2.346, (95% confidence interval (CI) [1.586, 3.470]; p < 0.001)). Across both cohorts, the major causes of death post-discharge were heart disease (13.79% neurological, 15.38% control), sepsis (8.63%, 17.58%), influenza and pneumonia (13.79%, 9.89%), COVID-19 (10.34%, 7.69%), and acute respiratory distress syndrome (ARDS) (10.34%, 6.59%). Factors associated with mortality after leaving the hospital involved the neurological cohort (odds ratio (OR): 1.802 (95% CI [1.237, 2.608]; p = 0.002)), discharge disposition (OR: 1.508 (95% CI [1.276, 1.775]; p < 0.001)), congestive heart failure (OR: 2.281 (95% CI [1.429, 3.593]; p < 0.001)), higher COVID-19 severity score (OR: 1.177 (95% CI [1.062, 1.304]; p = 0.002)), and older age (OR: 1.027 (95% CI [1.010, 1.044]; p = 0.002)). There were no group differences in radiological findings, except that the neurological cohort showed significantly more age-adjusted brain volume loss (p = 0.045) than controls. The study's patient cohort was limited to patients infected with COVID-19 during the first wave of the pandemic, when hospitals were overburdened, vaccines were not yet available, and treatments were limited. Patient profiles might differ when interrogating subsequent waves. CONCLUSIONS: Patients with COVID-19 with neurological manifestations had worse long-term outcomes compared to matched controls. These findings raise awareness and the need for closer monitoring and timely interventions for patients with COVID-19 with neurological manifestations, as their disease course involving initial neurological manifestations is associated with enhanced morbidity and mortality.

Time-dependent density functional theory with the orthogonal projector augmented wave method.

The projector augmented wave (PAW) method of Blöchl linearly maps smooth pseudo wavefunctions to the highly oscillatory all-electron DFT orbitals. Compared to norm-conserving pseudopotentials (NCPP), PAW has the advantage of lower kinetic energy cutoffs and larger grid spacing at the cost of having to solve for non-orthogonal wavefunctions. We earlier developed orthogonal PAW (OPAW) to allow the use of PAW when orthogonal wavefunctions are required. In OPAW, the pseudo wavefunctions are transformed through the efficient application of powers of the PAW overlap operator with essentially no extra cost compared to NCPP methods. Previously, we applied OPAW to DFT. Here, we take the first step to make OPAW viable for post-DFT methods by implementing it in real-time time-dependent (TD) DFT. Using fourth-order Runge-Kutta for the time-propagation, we compare calculations of absorption spectra for various organic and biological molecules and show that very large grid spacings are sufficient, 0.6-0.7 bohr in OPAW-TDDFT rather than the 0.4-0.5 bohr used in traditional NCPP-TDDFT calculations. This reduces the memory and propagation costs by around a factor of 3. Our method would be directly applicable to any post-DFT methods that require time-dependent propagations such as the GW approximation and the Bethe-Salpeter equation.

Deep learning applications to breast cancer detection by magnetic resonance imaging: a literature review.

Deep learning analysis of radiological images has the potential to improve diagnostic accuracy of breast cancer, ultimately leading to better patient outcomes. This paper systematically reviewed the current literature on deep learning detection of breast cancer based on magnetic resonance imaging (MRI). The literature search was performed from 2015 to Dec 31, 2022, using Pubmed. Other database included Semantic Scholar, ACM Digital Library, Google search, Google Scholar, and pre-print depositories (such as Research Square). Articles that were not deep learning (such as texture analysis) were excluded. PRISMA guidelines for reporting were used. We analyzed different deep learning algorithms, methods of analysis, experimental design, MRI image types, types of ground truths, sample sizes, numbers of benign and malignant lesions, and performance in the literature. We discussed lessons learned, challenges to broad deployment in clinical practice and suggested future research directions.

Deep learning combining FDG-PET and neurocognitive data accurately predicts MCI conversion to Alzheimer's dementia 3-year post MCI diagnosis.

INTRODUCTION: This study reports a novel deep learning approach to predict mild cognitive impairment (MCI) conversion to Alzheimer's dementia (AD) within three years using whole-brain fluorodeoxyglucose (FDG) positron emission tomography (PET) and cognitive scores (CS). METHODS: This analysis consisted of 150 normal controls (CN), 257 MCI, and 205  AD subjects from ADNI. FDG-PET and CS were obtained at MCI diagnosis to predict AD conversion within three years of MCI diagnosis using convolutional neural networks. RESULTS: Neurocognitive scores predicted better than FDG-PET per se, but the best model was a combination of FDG-PET, age, and neurocognitive data, yielding an AUC of 0.785 ± 0.096 and a balanced accuracy of 0.733 ± 0.098. Saliency maps highlighted putamen, thalamus, inferior frontal gyrus, parietal operculum, precuneus cortices, calcarine cortices, temporal gyrus, and planum temporale to be important for prediction. DISCUSSION: Deep learning accurately predicts MCI conversion to AD and provides neural correlates of brain regions associated with AD conversion.

Long-term outcomes of COVID-19 survivors with hospital AKI: association with time to recovery from AKI.

BACKGROUND: Although coronavirus disease 2019 (COVID-19) patients who develop in-hospital acute kidney injury (AKI) have worse short-term outcomes, their long-term outcomes have not been fully characterized. We investigated 90-day and 1-year outcomes after hospital AKI grouped by time to recovery from AKI. METHODS: This study consisted of 3296 COVID-19 patients with hospital AKI stratified by early recovery (<48 hours), delayed recovery (2-7 days) and prolonged recovery (>7-90 days). Demographics, comorbidities and laboratory values were obtained at admission and up to the 1-year follow-up. The incidence of major adverse cardiovascular events (MACE) and major adverse kidney events (MAKE), rehospitalization, recurrent AKI and new-onset chronic kidney disease (CKD) were obtained 90-days after COVID-19 discharge. RESULTS: The incidence of hospital AKI was 28.6%. Of the COVID-19 patients with AKI, 58.0% experienced early recovery, 14.8% delayed recovery and 27.1% prolonged recovery. Patients with a longer AKI recovery time had a higher prevalence of CKD (P < .05) and were more likely to need invasive mechanical ventilation (P < .001) and to die (P < .001). Many COVID-19 patients developed MAKE, recurrent AKI and new-onset CKD within 90 days, and these incidences were higher in the prolonged recovery group (P < .05). The incidence of MACE peaked 20-40 days postdischarge, whereas MAKE peaked 80-90 days postdischarge. Logistic regression models predicted 90-day MACE and MAKE with 82.4 ± 1.6% and 79.6 ± 2.3% accuracy, respectively. CONCLUSION: COVID-19 survivors who developed hospital AKI are at high risk for adverse cardiovascular and kidney outcomes, especially those with longer AKI recovery times and those with a history of CKD. These patients may require long-term follow-up for cardiac and kidney complications.

Long-term clinical outcomes and healthcare utilization of sickle cell disease patients with COVID-19: A 2.5-year follow-up study.

OBJECTIVES: This study investigated whether patients with sickle cell disease (SCD) had elevated risk of worse long-term clinical outcomes and healthcare utilization 2.5 years post-SARS-CoV-2 infection. METHODS: This study consisted of 178 patients with SCD who tested positive for COVID-19 between February 1, 2020 and January 30, 2022 in a major academic health system in New York City. The control cohort consisted of two-to-one matches of 356 SCD patients without a COVID-19 positive test. The last follow-up was July 18, 2022. The primary outcome was mortality. Secondary outcomes were annualized emergency department visits due to pain, pain hospital admission, length of stay due to pain, acute chest syndrome, episodic transfusion, and episodic exchange transfusion. RESULTS: There was no significant difference in mortality between SCD patients with and without COVID-19 (p > .05). There were no significant differences in secondary outcomes between pre- and postpandemic (p > .05). There were also no significant differences in these outcomes between SCD patients with and without COVID-19 (p > .05). SCD care utilization was not significantly associated with COVID-19 hospitalization status (p > .05). CONCLUSIONS: SCD patients with SARS-CoV-2 infection incurred no additional risk of worse long-term outcomes compared to matched controls of SCD patients not infected by SARS-CoV-2.

COVID-19 associated ketosis and diabetic ketoacidosis: A rapid review.

SARS-CoV-2 infection could disrupt the endocrine system directly or indirectly, which could result in endocrine dysfunction and glycaemic dysregulation, triggering transient or persistent diabetes mellitus. The literature on the complex relationship between COVID-19 and endocrine dysfunctions is still evolving and remains incompletely understood. Thus, we conducted a review on all literature to date involving COVID-19 associated ketosis or diabetic ketoacidosis (DKA). In total, 27 publications were included and analysed quantitatively and qualitatively. Studies included patients with DKA with existing or new onset diabetes. While the number of case and cohort studies was limited, DKA in the setting of COVID-19 seemed to increase risk of death, particularly in patients with new onset diabetes. Future studies with more specific variables and larger sample sizes are needed to draw better conclusions.

Incidence, characteristics, risk factors and outcomes of diabetic ketoacidosis in COVID-19 patients: Comparison with influenza and pre-pandemic data.

AIMS: This study characterized incidence, patient profiles, risk factors and outcomes of in-hospital diabetic ketoacidosis (DKA) in patients with COVID-19 compared with influenza and pre-pandemic data. METHODS: This study consisted of 13 383 hospitalized patients with COVID-19 (March 2020-July 2022), 19 165 hospitalized patients with influenza (January 2018-July 2022) and 35 000 randomly sampled hospitalized pre-pandemic patients (January 2017-December 2019) in Montefiore Health System, Bronx, NY, USA. Primary outcomes were incidence of in-hospital DKA, in-hospital mortality, and insulin use at 3 and 6 months post-infection. Risk factors for developing DKA were identified. RESULTS: The overall incidence of DKA in patients with COVID-19 and influenza, and pre-pandemic were 2.1%, 1.4% and 0.5%, respectively (p < .05 pairwise). Patients with COVID-19 with DKA had worse acute outcomes (p < .05) and higher incidence of new insulin treatment 3 and 6 months post-infection compared with patients with influenza with DKA (p < .05). The incidence of DKA in patients with COVID-19 was highest among patients with type 1 diabetes (12.8%), followed by patients with insulin-dependent type 2 diabetes (T2D; 5.2%), non-insulin dependent T2D (2.3%) and, lastly, patients without T2D (1.3%). Patients with COVID-19 with DKA had worse disease severity and higher mortality [odds ratio = 6.178 (4.428-8.590), p < .0001] compared with those without DKA. Type 1 diabetes, steroid therapy for COVID-19, COVID-19 status, black race and male gender were associated with increased risk of DKA. CONCLUSIONS: The incidence of DKA was higher in COVID-19 cohort compared to the influenza and pre-pandemic cohort. Patients with COVID-19 with DKA had worse outcomes compared with those without. Many COVID-19 survivors who developed DKA during hospitalization became insulin dependent. Identification of risk factors for DKA and new insulin-dependency could enable careful monitoring and timely intervention.

Longitudinal prediction of hospital-acquired acute kidney injury in COVID-19: a two-center study.

BACKGROUND: To investigate the temporal characteristics of clinical variables of hospital-acquired acute kidney injury (AKI) in COVID-19 patients and to longitudinally predict AKI onset. METHODS: There were 308 hospital-acquired AKI and 721 non-AKI (NAKI) COVID-19 patients from Stony Brook Hospital (New York, USA) data, and 72 hospital-acquired AKI and 303 NAKI COVID-19 patients from Tongji Hospital (Wuhan, China). Demographic, comorbidities, and longitudinal (3 days before and 3 days after AKI onset) clinical variables were used to compute odds ratios for and longitudinally predict hospital-acquired AKI onset. RESULTS: COVID-19 patients with AKI were more likely to die than NAKI patients (31.5% vs 6.9%, adjusted p < 0.001, OR = 4.67 [95% CI 3.1, 7.0], Stony Brook data). AKI developed on average 3.3 days after hospitalization. Procalcitonin was elevated prior to AKI onset (p < 0.05), peaked, and remained elevated (p < 0.05). Alanine aminotransferase, aspartate transaminase, ferritin, and lactate dehydrogenase peaked the same time as creatinine, whereas D-dimer and brain natriuretic peptide peaked a day later. C-reactive protein, white blood cell and lymphocyte showed group differences - 2 days prior (p < 0.05). Top predictors were creatinine, procalcitonin, white blood cells, lactate dehydrogenase, and lymphocytes. They predicted AKI onset with areas under curves (AUCs) of 0.78, 0.66, and 0.56 at 0, - 1, and - 2 days prior, respectively. When tested on the Tongji Hospital data, the AUCs were 0.80, 0.79, and 0.77, respectively. CONCLUSIONS: Time-locked longitudinal data provide insight into AKI progression. Commonly clinical variables reasonably predict AKI onset a few days prior. This work may lead to earlier recognition of AKI and treatment to improve clinical outcomes.

Brain Reactions to Opening and Closing the Eyes: Salivary Cortisol and Functional Connectivity.

This study empirically assessed the strength and duration of short-term effects induced by brain reactions to closing/opening the eyes on a few well-known resting-state networks. We also examined the association between these reactions and subjects' cortisol levels. A total of 55 young adults underwent 8-min resting-state fMRI (rs-fMRI) scans under 4-min eyes-closed and 4-min eyes-open conditions. Saliva samples were collected from 25 of the 55 subjects before and after the fMRI sessions and assayed for cortisol levels. Our empirical results indicate that when the subjects were relaxed with their eyes closed, the effect of opening the eyes on conventional resting-state networks (e.g., default-mode, frontal-parietal, and saliency networks) lasted for roughly 60-s, during which we observed a short-term increase in activity in rs-fMRI time courses. Moreover, brain reactions to opening the eyes had a pronounced effect on time courses in the temporo-parietal lobes and limbic structures, both of which presented a prolonged decrease in activity. After controlling for demographic factors, we observed a significantly positive correlation between pre-scan cortisol levels and connectivity in the limbic structures under both conditions. Under the eyes-closed condition, the temporo-parietal lobes presented significant connectivity to limbic structures and a significantly positive correlation with pre-scan cortisol levels. Future research on rs-fMRI could consider the eyes-closed condition when probing resting-state connectivity and its neuroendocrine correlates, such as cortisol levels. It also appears that abrupt instructions to open the eyes while the subject is resting quietly with eyes closed could be used to probe brain reactivity to aversive stimuli in the ventral hippocampus and other limbic structures.

Deep learning of longitudinal chest X-ray and clinical variables predicts duration on ventilator and mortality in COVID-19 patients.

OBJECTIVES: To use deep learning of serial portable chest X-ray (pCXR) and clinical variables to predict mortality and duration on invasive mechanical ventilation (IMV) for Coronavirus disease 2019 (COVID-19) patients. METHODS: This is a retrospective study. Serial pCXR and serial clinical variables were analyzed for data from day 1, day 5, day 1-3, day 3-5, or day 1-5 on IMV (110 IMV survivors and 76 IMV non-survivors). The outcome variables were duration on IMV and mortality. With fivefold cross-validation, the performance of the proposed deep learning system was evaluated by receiver operating characteristic (ROC) analysis and correlation analysis. RESULTS: Predictive models using 5-consecutive-day data outperformed those using 3-consecutive-day and 1-day data. Prediction using data closer to the outcome was generally better (i.e., day 5 data performed better than day 1 data, and day 3-5 data performed better than day 1-3 data). Prediction performance was generally better for the combined pCXR and non-imaging clinical data than either alone. The combined pCXR and non-imaging data of 5 consecutive days predicted mortality with an accuracy of 85 ± 3.5% (95% confidence interval (CI)) and an area under the curve (AUC) of 0.87 ± 0.05 (95% CI) and predicted the duration needed to be on IMV to within 2.56 ± 0.21 (95% CI) days on the validation dataset. CONCLUSIONS: Deep learning of longitudinal pCXR and clinical data have the potential to accurately predict mortality and duration on IMV in COVID-19 patients. Longitudinal pCXR could have prognostic value if these findings can be validated in a large, multi-institutional cohort.

Machine learning classification of texture features of MRI breast tumor and peri-tumor of combined pre- and early treatment predicts pathologic complete response.

PURPOSE: This study used machine learning classification of texture features from MRI of breast tumor and peri-tumor at multiple treatment time points in conjunction with molecular subtypes to predict eventual pathological complete response (PCR) to neoadjuvant chemotherapy. MATERIALS AND METHOD: This study employed a subset of patients (N = 166) with PCR data from the I-SPY-1 TRIAL (2002-2006). This cohort consisted of patients with stage 2 or 3 breast cancer that underwent anthracycline-cyclophosphamide and taxane treatment. Magnetic resonance imaging (MRI) was acquired pre-neoadjuvant chemotherapy, early, and mid-treatment. Texture features were extracted from post-contrast-enhanced MRI, pre- and post-contrast subtraction images, and with morphological dilation to include peri-tumoral tissue. Molecular subtypes and Ki67 were also included in the prediction model. Performance of classification models used the receiver operating characteristics curve analysis including area under the curve (AUC). Statistical analysis was done using unpaired two-tailed t-tests. RESULTS: Molecular subtypes alone yielded moderate prediction performance of PCR (AUC = 0.82, p = 0.07). Pre-, early, and mid-treatment data alone yielded moderate performance (AUC = 0.88, 0.72, and 0.78, p = 0.03, 0.13, 0.44, respectively). The combined pre- and early treatment data markedly improved performance (AUC = 0.96, p = 0.0003). Addition of molecular subtypes improved performance slightly for individual time points but substantially for the combined pre- and early treatment (AUC = 0.98, p = 0.0003). The optimal morphological dilation was 3-5 pixels. Subtraction of post- and pre-contrast MRI further improved performance (AUC = 0.98, p = 0.00003). Finally, among the machine-learning algorithms evaluated, the RUSBoosted Tree machine-learning method yielded the highest performance. CONCLUSION: AI-classification of texture features from MRI of breast tumor at multiple treatment time points accurately predicts eventual PCR. Longitudinal changes in texture features and peri-tumoral features further improve PCR prediction performance. Accurate assessment of treatment efficacy early on could minimize unnecessary toxic chemotherapy and enable mid-treatment modification for patients to achieve better clinical outcomes.

Survival of COVID-19 Patients With Respiratory Failure is Related to Temporal Changes in Gas Exchange and Mechanical Ventilation.

Background: Respiratory failure due to coronavirus disease of 2019 (COVID-19) often presents with worsening gas exchange over a period of days. Once patients require mechanical ventilation (MV), the temporal change in gas exchange and its relation to clinical outcome is poorly described. We investigated whether gas exchange over the first 5 days of MV is associated with mortality and ventilator-free days at 28 days in COVID-19. Methods: In a cohort of 294 COVID-19 patients, we used data during the first 5 days of MV to calculate 4 daily respiratory scores: PaO2/FiO2 (P/F), oxygenation index (OI), ventilatory ratio (VR), and Murray lung injury score. The association between these scores at early (days 1-3) and late (days 4-5) time points with mortality was evaluated using logistic regression, adjusted for demographics. Correlation with ventilator-free days was assessed (Spearman rank-order coefficients). Results: Overall mortality was 47.6%. Nonsurvivors were older (P < .0001), more male (P = .029), with more preexisting cardiopulmonary disease compared to survivors. Mean PaO2 and PaCO2 were similar during this timeframe. However, by days 4 to 5 values for all airway pressures and FiO2 had diverged, trending lower in survivors and higher in nonsurvivors. The most substantial between-group difference was the temporal change in OI, improving 15% in survivors and worsening 11% in nonsurvivors (P < .05). The adjusted mortality OR was significant for age (1.819, P = .001), OI at days 4 to 5 (2.26, P = .002), and OI percent change (1.90, P = .02). The number of ventilator-free days correlated significantly with late VR (-0.166, P < .05), early and late OI (-0.216, P < .01; -0.278, P < .01, respectively) and early and late P/F (0.158, P < .05; 0.283, P < .01, respectively). Conclusion: Nonsurvivors of COVID-19 needed increasing intensity of MV to sustain gas exchange over the first 5 days, unlike survivors. Temporal change OI, reflecting both PaO2 and the intensity of MV, is a potential marker of outcome in respiratory failure due to COVID-19.

Machining learning predicts the need for escalated care and mortality in COVID-19 patients from clinical variables.

Objective: This study aimed to develop a machine learning algorithm to identify key clinical measures to triage patients more effectively to general admission versus intensive care unit (ICU) admission and to predict mortality in COVID-19 pandemic. Materials and methods: This retrospective study consisted of 1874 persons-under-investigation for COVID-19 between February 7, 2020, and May 27, 2020 at Stony Brook University Hospital, New York. Two primary outcomes were ICU admission and mortality compared to COVID-19 positive patients in general hospital admission. Demographic, vitals, symptoms, imaging findings, comorbidities, and laboratory tests at presentation were collected. Predictions of mortality and ICU admission were made using machine learning with 80% training and 20% testing. Performance was evaluated using receiver operating characteristic (ROC) area under the curve (AUC). Results: A total of 635 patients were included in the analysis (age 60±11, 40.2% female). The top 6 mortality predictors were age, procalcitonin, C-creative protein, lactate dehydrogenase, D-dimer and lymphocytes. The top 6 ICU admission predictors are procalcitonin, lactate dehydrogenase, C-creative protein, pulse oxygen saturation, temperature and ferritin. The best machine learning algorithms predicted mortality with 89% AUC and ICU admission with 79% AUC. Conclusion: This study identifies key independent clinical parameters that predict ICU admission and mortality associated with COVID-19 infection. The predictive model is practical, readily enhanced and retrained using additional data. This approach has immediate translation and may prove useful for frontline physicians in clinical decision making under time-sensitive and resource-constrained environment.

Continued In-Hospital Angiotensin-Converting Enzyme Inhibitor and Angiotensin II Receptor Blocker Use in Hypertensive COVID-19 Patients Is Associated With Positive Clinical Outcome.

BACKGROUND: This study investigated continued and discontinued use of angiotensin-converting enzyme inhibitors (ACEi) or angiotensin II receptor blockers (ARB) during hospitalization of 614 hypertensive laboratory-confirmed COVID-19 patients. METHODS: Demographics, comorbidities, vital signs, laboratory data, and ACEi/ARB usage were analyzed. To account for confounders, patients were substratified by whether they developed hypotension and acute kidney injury (AKI) during the index hospitalization. RESULTS: Mortality (22% vs 17%, P > .05) and intensive care unit (ICU) admission (26% vs 12%, P > .05) rates were not significantly different between non-ACEi/ARB and ACEi/ARB groups. However, patients who continued ACEi/ARBs in the hospital had a markedly lower ICU admission rate (12% vs 26%; P = .001; odds ratio [OR] = 0.347; 95% confidence interval [CI], .187-.643) and mortality rate (6% vs 28%; P = .001; OR = 0.215; 95% CI, .101-.455) compared to patients who discontinued ACEi/ARB. The odds ratio for mortality remained significantly lower after accounting for development of hypotension or AKI. CONCLUSIONS: These findings suggest that continued ACEi/ARB use in hypertensive COVID-19 patients yields better clinical outcomes.

Machine-learning classification of texture features of portable chest X-ray accurately classifies COVID-19 lung infection.

BACKGROUND: The large volume and suboptimal image quality of portable chest X-rays (CXRs) as a result of the COVID-19 pandemic could post significant challenges for radiologists and frontline physicians. Deep-learning artificial intelligent (AI) methods have the potential to help improve diagnostic efficiency and accuracy for reading portable CXRs. PURPOSE: The study aimed at developing an AI imaging analysis tool to classify COVID-19 lung infection based on portable CXRs. MATERIALS AND METHODS: Public datasets of COVID-19 (N = 130), bacterial pneumonia (N = 145), non-COVID-19 viral pneumonia (N = 145), and normal (N = 138) CXRs were analyzed. Texture and morphological features were extracted. Five supervised machine-learning AI algorithms were used to classify COVID-19 from other conditions. Two-class and multi-class classification were performed. Statistical analysis was done using unpaired two-tailed t tests with unequal variance between groups. Performance of classification models used the receiver-operating characteristic (ROC) curve analysis. RESULTS: For the two-class classification, the accuracy, sensitivity and specificity were, respectively, 100%, 100%, and 100% for COVID-19 vs normal; 96.34%, 95.35% and 97.44% for COVID-19 vs bacterial pneumonia; and 97.56%, 97.44% and 97.67% for COVID-19 vs non-COVID-19 viral pneumonia. For the multi-class classification, the combined accuracy and AUC were 79.52% and 0.87, respectively. CONCLUSION: AI classification of texture and morphological features of portable CXRs accurately distinguishes COVID-19 lung infection in patients in multi-class datasets. Deep-learning methods have the potential to improve diagnostic efficiency and accuracy for portable CXRs.

3D MRI of whole-brain water permeability with intrinsic diffusivity encoding of arterial labeled spin (IDEALS).

This work proposes a novel MRI method - Intrinsic Diffusivity Encoding of Arterial Labeled Spin (IDEALS) - for the whole-brain mapping of water permeability in the human brain without an exogenous contrast agent. Quantitative separation of the intravascular and extravascular labeled water MRI signal was achieved in arterial spin labeling experiments with segmented 3D-GRASE acquisition by modulating the relative sensitivity between relaxation, true diffusion, and pseudodiffusion. The intrinsic diffusivity encoding in k-space created different broadening of the image-domain point spread functions for intravascular and extravascular labeled spins, from which blood-brain barrier (BBB) water extraction fraction (Ew) and water permeability surface area product (PSw) were estimated. The feasibility and sensitivity of this method was evaluated in healthy subjects at baseline and after caffeine challenge. The estimated baseline Ew and PSw maps showed contrast among gray matter (GM) and white matter (WM). GM Ew was significantly lower than that of WM (78.8% ±â€¯3.3% in GM vs. 83.9% ±â€¯4.6% in WM; p < 0.05) and GM PSw was significantly higher than that of WM (131.7 ±â€¯29.5 mL/100  g/min in GM vs. 76.2 ±â€¯18.4 mL/100  g/min in WM; p < 0.05). BBB Ew was significantly lower for females than males (74.9% ±â€¯3.7% for females vs. 81.3% ±â€¯3.3% for males in GM; 80.5% ±â€¯4.7% for females vs. 86.1 ±â€¯3.0 for males in WM; p < 0.05 for both), while significant PSw differences were only observed in WM (143.8 ±â€¯34.4 mL/100  g/min for females vs. 123.6 ±â€¯24.4 mL/100  g/min for males in GM; 91.6 ±â€¯15.0 mL/100  g/min for females vs. 65.9 ±â€¯12.5 mL/100  g/min for males in WM; p = 0.20 and p < 0.05 for GM and WM respectively). Significant correlations between Ew and CBF (r = -0.32, p < 0.05) and between PSw and CBF (r = 0.89, p < 0.05) were observed, consistent with 15O-H2O PET findings. After caffeine challenge, reduced CBF, Ew and PSw were observed, demonstrating the sensitivity of IDEALS approach.

Predicting Four-Year's Alzheimer's Disease Onset Using Longitudinal Neurocognitive Tests and MRI Data Using Explainable Deep Convolutional Neural Networks.

BACKGROUND: Prognosis of future risk of dementia from neuroimaging and cognitive data is important for optimizing clinical management for patients at early stage of Alzheimer's disease (AD). However, existing studies lack an efficient way to integrate longitudinal information from both modalities to improve prognosis performance. OBJECTIVE: In this study, we aim to develop and evaluate an explainable deep learning-based framework to predict mild cognitive impairment (MCI) to AD conversion within four years using longitudinal whole-brain 3D MRI and neurocognitive tests. METHODS: We proposed a two-stage framework that first uses a 3D convolutional neural network to extract single-timepoint MRI-based AD-related latent features, followed by multi-modal longitudinal feature concatenation and a 1D convolutional neural network to predict the risk of future dementia onset in four years. RESULTS: The proposed deep learning framework showed promising to predict MCI to AD conversion within 4 years using longitudinal whole-brain 3D MRI and cognitive data without extracting regional brain volumes or cortical thickness, reaching a balanced accuracy of 0.834, significantly improved from models trained from single timepoint or single modality. The post hoc model explainability revealed heatmap indicating regions that are important for predicting future risk of AD. CONCLUSIONS: The proposed framework sets the stage for future studies for using multi-modal longitudinal data to achieve optimal prediction for prognosis of AD onset, leading to better management of the diseases, thereby improving the quality of life.