Dataset Information

Machine learning approach for differentiating cytomegalovirus esophagitis from herpes simplex virus esophagitis.

ABSTRACT: The endoscopic features between herpes simplex virus (HSV) and cytomegalovirus (CMV) esophagitis overlap significantly, and hence the differential diagnosis between HSV and CMV esophagitis is sometimes difficult. Therefore, we developed a machine-learning-based classifier to discriminate between CMV and HSV esophagitis. We analyzed 87 patients with HSV esophagitis and 63 patients with CMV esophagitis and developed a machine-learning-based artificial intelligence (AI) system using a total of 666 endoscopic images with HSV esophagitis and 416 endoscopic images with CMV esophagitis. In the five repeated five-fold cross-validations based on the hue-saturation-brightness color model, logistic regression with a least absolute shrinkage and selection operation showed the best performance (sensitivity, specificity, positive predictive value, negative predictive value, accuracy, and area under the receiver operating characteristic curve: 100%, 100%, 100%, 100%, 100%, and 1.0, respectively). Previous history of transplantation was included in classifiers as a clinical factor; the lower the performance of these classifiers, the greater the effect of including this clinical factor. Our machine-learning-based AI system for differential diagnosis between HSV and CMV esophagitis showed high accuracy, which could help clinicians with diagnoses.

SUBMITTER: Lee JS

PROVIDER: S-EPMC7878749 | biostudies-literature | 2021 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning approach for differentiating cytomegalovirus esophagitis from herpes simplex virus esophagitis.

Lee Jung Su JS Yun Jihye J Ham Sungwon S Park Hyunjung H Lee Hyunsu H Kim Jeongseok J Byeon Jeong-Sik JS Jung Hwoon-Yong HY Kim Namkug N Kim Do Hoon DH

Scientific reports 20210211 1

The endoscopic features between herpes simplex virus (HSV) and cytomegalovirus (CMV) esophagitis overlap significantly, and hence the differential diagnosis between HSV and CMV esophagitis is sometimes difficult. Therefore, we developed a machine-learning-based classifier to discriminate between CMV and HSV esophagitis. We analyzed 87 patients with HSV esophagitis and 63 patients with CMV esophagitis and developed a machine-learning-based artificial intelligence (AI) system using a total of 666 ...[more]

PMID: 33574361

Similar Datasets

Project description:BackgroundHerpesviridae reactivation among non-immunocompromised critically ill patients is associated with impaired prognosis, especially during acute respiratory distress syndrome (ARDS). However, little is known about herpes simplex virus (HSV) and Cytomegalovirus (CMV) reactivation occurring in patients with severe ARDS under veno-venous extracorporeal membrane oxygenation (ECMO). We tried to determine the frequency of Herpesviridae reactivation and its impact on patients' prognosis during ECMO for severe ARDS.ResultsDuring a 5-year period, 123 non-immunocompromised patients with a severe ARDS requiring a veno-venous ECMO were included. Sixty-seven patients (54%) experienced HSV and/or CMV reactivation during ECMO course (20 viral co-infection, 40 HSV alone, and 7 CMV alone). HSV reactivation occurred earlier than CMV after the beginning of MV [(6-15) vs. 19 (13-29) days; p < 0.01] and after ECMO implementation [(2-8) vs. 14 (10-20) days; p < 0.01]. In univariate analysis, HSV/CMV reactivation was associated with a longer duration of mechanical ventilation [(22-52.5) vs. 17.5 (9-28) days; p < 0.01], a longer duration of ECMO [15 (10-22.5) vs. 9 (5-14) days; p < 0.01], and a prolonged ICU [29 (19.5-47.5) vs. 16 (9-30) days; p < 0.01] and hospital stay [44 (29-63.5) vs. 24 (11-43) days; p < 0.01] as compared to non-reactivated patients. However, in multivariate analysis, viral reactivation remained associated with prolonged MV only. When considered separately, both HSV and CMV reactivation were associated with a longer duration of MV as compared to non-reactivation patients [29 (19.5-41) and 28 (20.5-37), respectively, vs. 17.5 (9-28) days; p < 0.05]. Co-reactivation patients had a longer duration of MV [58.5 (38-72.3); p < 0.05] and ICU stay [51.5 (32.5-69) vs. 27.5 (17.75-35.5) and 29 (20-30.5), respectively] as compared to patients with HSV or CMV reactivation alone. In multivariate analysis, HSV reactivation remained independently associated with a longer duration of MV and hospital length of stay.ConclusionsHerpesviridae reactivation is frequent among patients with severe ARDS under veno-venous ECMO and is associated with a longer duration of mechanical ventilation. The direct causative link between HSV and CMV reactivation and respiratory function worsening under ECMO remains to be confirmed.

Project description:Viral proteins pUL16 and pUL21 are required for efficient nuclear egress of herpes simplex virus 2 capsids. To better understand the role of these proteins in nuclear egress, we established whether nuclear egress complex (NEC) distribution and/or function was altered in the absence of either pUL16 or pUL21. NEC distribution in cells infected with pUL16-deficient viruses was indistinguishable from that observed in cells infected with wild-type viruses. In contrast, NEC distribution was aberrant in cells infected with pUL21-deficient virus and, instead, showed some similarity to the aberrant NEC distribution pattern observed in cells infected with pUs3-deficient virus. These results indicated that pUL16 plays a role in nuclear egress that is distinct from that of pUL21 and pUs3. Higher-resolution examination of nuclear envelope ultrastructure in cells infected with pUL21-deficient viruses by transmission electron microscopy showed different types of nuclear envelope perturbations, including some that were not observed in cells infected with pUs3 deficient virus. The formation of the nuclear envelope perturbations observed in pUL21-deficient virus infections was dependent on a functional NEC, revealing a novel role for pUL21 in regulating NEC activity. The results of comparisons of nuclear envelope ultrastructure in cells infected with viruses lacking pUs3, pUL16, or both pUs3 and pUL16 were consistent with a role for pUL16 in advance of primary capsid envelopment and shed new light on how pUs3 functions in nuclear egress.IMPORTANCE The membrane deformation activity of the herpesvirus nuclear egress complex (NEC) allows capsids to transit through both nuclear membranes into the cytoplasm. NEC activity must be precisely controlled during viral infection, and yet our knowledge of how NEC activity is controlled is incomplete. To determine how pUL16 and pUL21, two viral proteins required for nuclear egress of herpes simplex virus 2, function in nuclear egress, we examined how the lack of each protein impacted NEC distribution. These analyses revealed a function of pUL16 in nuclear egress distinct from that of pUL21, uncovered a novel role for pUL21 in regulating NEC activity, and shed new light on how a viral kinase, pUs3, regulates nuclear egress. Nuclear egress of capsids is required for all herpesviruses. A complete understanding of all aspects of nuclear egress, including how viral NEC activity is controlled, may yield strategies to disrupt this process and aid the development of herpes-specific antiviral therapies.

Dataset Information

Machine learning approach for differentiating cytomegalovirus esophagitis from herpes simplex virus esophagitis.

Publications

Machine learning approach for differentiating cytomegalovirus esophagitis from herpes simplex virus esophagitis.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets