Resource Management through Artificial Intelligence in Screening Programs-Key for the Successful Elimination of Hepatitis C

Abstract

Background: The elimination of the Hepatitis C virus (HCV) will only be possible if rapid and efficient actions are taken. Artificial neural networks (ANNs) are computing systems based on the topology of the biological brain, containing connected artificial neurons that can be tasked with solving medical problems.

Aim: We expanded the previously presented HCV micro-elimination project started in September 2020 that aimed to identify HCV infection through coordinated screening in asymptomatic populations and developed two ANN models able to identify at-risk subjects selected through a targeted questionnaire.

Material and method: Our study included 14,042 screened participants from a southwestern region of Oltenia, Romania. Each participant completed a 12-item questionnaire along with anti-HCV antibody rapid testing. Hepatitis-C-positive subjects were linked to care and ultimately could receive antiviral treatment if they had detectable viremia. We built two ANNs, trained and tested on the dataset derived from the questionnaires and then used to identify patients in a similar, already existing dataset.

Results: We found 114 HCV-positive patients (81 females), resulting in an overall prevalence of 0.81%. We identified sharing personal hygiene items, receiving blood transfusions, having dental work or surgery and re-using hypodermic needles as significant risk factors. When used on an existing dataset of 15,140 persons (119 HCV cases), the first ANN models correctly identified 97 (81.51%) HCV-positive subjects through 13,401 tests, while the second ANN model identified 81 (68.06%) patients through only 5192 tests.

Conclusions: The use of ANNs in selecting screening candidates may improve resource allocation and prioritize cases more prone to severe disease.

Keywords: artificial intelligence; hepatitis C virus; micro-elimination; prediction model; screening.

Figure 1

Proposed model for the ANN developed for the AI model.

Figure 2

Sigmoid activation function of the ANN model.

Figure 3

Dataset distribution after over-sampling.

Figure 4

Overview of the age distribution in the study lot.

Figure 5

Age distribution in persons with positive anti-HCV antibodies.

Figure 6

Distribution by gender and provenance in the entire lot.

Figure 7

Gender and residence of men and women in the subgroup of subjects with present anti-HCV antibodies.

Figure 8

Breakdown of the sub-lot containing subjects already aware of their HCV-positive status. Of the 61 persons, 16 chose to start an antiviral regimen, having detectable viremia and not receiving previous treatment.

Figure 9

Linkage-to-care analysis of the current lot, compared to the previously analyzed lot.

Figure 10

Receiver operating characteristic curve of both proposed models.

Figure 11

(A) Comparative overview of the number of correctly identified HCV-positive persons by each ANN model; we can observe the constantly high number of true positives identified by Model 1, however (B), at the cost of a significantly higher number of tests performed compared to Model 2. In contrast (C), Model 2 was more effective at identifying true not-at-risk persons; this feature, corroborated with the low HCV incidence of 0.79% in the chosen dataset of 15,140 subjects, resulted in an overall higher accuracy for Model 2.

Figure 11

(A) Comparative overview of the number of correctly identified HCV-positive persons by each ANN model; we can observe the constantly high number of true positives identified by Model 1, however (B), at the cost of a significantly higher number of tests performed compared to Model 2. In contrast (C), Model 2 was more effective at identifying true not-at-risk persons; this feature, corroborated with the low HCV incidence of 0.79% in the chosen dataset of 15,140 subjects, resulted in an overall higher accuracy for Model 2.