ABSTRACT
Background and objectives: We aim to verify the use of ML algorithms to predict patient outcome using a relatively small dataset and to create a nomogram to assess in-hospital mortality of patients with COVID-19. Methods: A database of 200 COVID-19 patients admitted to the Clinical Hospital of State University of Campinas (UNICAMP) was used in this analysis. Patient features were divided into three categories: clinical, chest abnormalities, and body composition characteristics acquired by computerized tomography. These features were evaluated independently and combined to predict patient outcomes. To minimize performance fluctuations due to low sample number, reduce possible bias related to outliers, and evaluate the uncertainties generated by the small dataset, we developed a shuffling technique, a modified version of the Monte Carlo Cross Validation, creating several subgroups for training the algorithm and complementary testing subgroups. The following ML algorithms were tested: random forest, boosted decision trees, logistic regression, support vector machines, and neural networks. Performance was evaluated by analyzing Receiver operating characteristic (ROC) curves. The importance of each feature in the determination of the outcome predictability was also studied and a nomogram was created based on the most important features selected by the exclusion test. Results: Among the different sets of features, clinical variables age, lymphocyte number and weight were the most valuable features for prognosis prediction. However, we observed that skeletal muscle radiodensity and presence of pleural effusion were also important for outcome determination. Integrating these independent predictors was successfully developed to accurately predict mortality in COVID-19 in hospital patients. A nomogram based on these five features was created to predict COVID-19 mortality in hospitalized patients. The area under the ROC curve was 0.86 ± 0.04. Conclusion: ML algorithms can be reliable for the prediction of COVID-19-related in-hospital mortality, even when using a relatively small dataset. The success of ML techniques in smaller datasets broadens the applicability of these methods in several problems in the medical area. In addition, feature importance analysis allowed us to determine the most important variables for the prediction tasks resulting in a nomogram with good accuracy and clinical utility in predicting COVID-19 in-hospital mortality.
ABSTRACT
The prevalence of Alzheimer's disease (AD) has been rapidly increasing worldwide. We have developed a novel angiogenic therapy with low-intensity pulsed ultrasound (LIPUS), which is effective and safe in animal models of AD and vascular dementia. We performed two trials of LIPUS therapy for AD (mild cognitive impairment due to AD and mild AD); a roll-in open trial for safety, and a randomized, double-blind, placebo-controlled (RCT) trial for efficacy and safety. The LIPUS therapy was performed for whole brain through the bilateral temporal bones for one hour 3 times a week as one session under the special conditions (1.3 MPa, 32 cycles, 5% duty cycle) we identified. The LIPUS therapy was performed for one session in the roll-in trial, and 6 sessions in the RCT trial with 3-month intervals for 1.5 years. The primary endpoint was ADAS-J cog scores. The RCT trial was terminated prematurely due to the COVID-19 pandemic. In the roll-in trial (N = 5), no adverse effects were noted. In the RCT trial (N = 22), the worsening of ADAS-J cog scores tended to be suppressed in the LIPUS group compared with the placebo group at week 72 (P = 0.257). When responders were defined as those with no worsening of ADAS-J cog scores at week 72, the prevalence was 50% (5/10) and 0% (0/5) in the LIPUS and placebo groups, respectively (P = 0.053). No adverse effects were noted. These results suggest that the LIPUS therapy is safe and tends to suppress cognitive impairment although a next pivotal trial with a large number of subjects is warranted.
Subject(s)
Alzheimer Disease , COVID-19 , Animals , Humans , Alzheimer Disease/therapy , Alzheimer Disease/psychology , Pilot Projects , Pandemics , Brain/diagnostic imaging , Ultrasonic WavesABSTRACT
Due to the coronavirus disease 2019 (COVID-19) pandemic, the first state of emergency had been declared from April 7 to May 25, 2020, in Japan. This pandemic might affect the management for patients with acute myocardial infarction (AMI). Method and Results: To evaluate the critical care and outcomes of AMI patients during the COVID-19 outbreak, we examined the patients with AMI hospitalized in 2020 (n = 1186) and those in 2017-2019 (n = 4877) using a database of the Miyagi AMI Registry Study. The door-to-device time under the emergency declaration became longer as compared with that of the same period in 2017-2019 [83(65-111) vs 74(54-108) min, p = 0.04]. Importantly, the time delay was noted in only patients with Killip class I on arrival, but not in those with Killip class II-IV. Meanwhile, there were no significant changes in the duration from the symptom onset to hospital arrival, the use rate of ambulance and the performance rate of primary percutaneous coronary intervention before and after the COVID-19 outbreak. Eventually, in-hospital mortality had not deteriorated under the state of emergency (6.7 vs 7.8 %, P = 0.69). Conclusion: The emergence of the COVID-19 outbreak seemed to affect AMI management and highlight understanding the barriers to cardiovascular critical care.
ABSTRACT
[Figure: see text].