Omosivie Maduka
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Omosivie Maduka has a total of 216 co-authors. Here are the top co-authors by number of shared publications:
- Rosemary Ogu
University of Port Harcourt, Rivers State, Nigeria
5 shared publications - Sylvester Maleghemi
Vaccine Preventable Diseases Programme, World Health Organization Headquarters, Geneva, Switzerland
5 shared publications Vetty Agala
Rivers State Ministry of Health, Port Harcourt, Nigeria
5 shared publications- Datonye Alasia
Department of Internal Medicine, University of Port-Harcourt Teaching Hospital , Port-Harcourt, Rivers State , Nigeria
4 shared publications Charles Tobin-West
Rivers State Public Health Emergency Operations Centre, 50001, Port Harcourt, Nigeria
3 shared publications- Christie Akwaowo
Institute of Health Research and Development, University of Uyo Teaching Hospital, Uyo 520103, Nigeria
3 shared publications
Latest Publications
- Open Access
- 31 October 2026
- Published by GSC Online Press in World Journal of Advanced Research and Reviews
- Vol. 32 (1) , 1044-1056
DETERMINANTS OF DIAGNOSTIC ACCURACY OF TYPHOID FEVER TESTS IN GA WEST MUNICIPALITY, GHANA
Background: Typhoid fever remains a significant public health challenge in Ghana due to overlapping clinical presentations with malaria and other febrile illnesses. Accurate diagnosis is essential for effective patient management and rational antibiotic use. This study evaluated determinants of diagnostic accuracy of typhoid fever tests in Ga West Municipality, Ghana. Methods: A cross-sectional comparative diagnostic accuracy study was conducted in seven health facilities. A total of 284 participants with suspected typhoid fever were recruited using consecutive sampling. Data were collected through structured questionnaires, clinical evaluations, and laboratory investigations, including Widal test, RDT IgM, blood culture (reference standard), and PCR. Statistical analyses included descriptive statistics, chi-square tests, multivariable logistic regression, Cohen's kappa, and McNemar's test at a 0.05 significance level. Results: Previous hospitalization was the strongest predictor of diagnostic accuracy for the Widal test (aOR = 7.74; p < 0.001) and RDT IgM (aOR = 6.96; p < 0.001). Symptom duration of 4–7 days was significantly associated with accuracy for both tests (Widal: aOR = 2.03; p = 0.019; RDT IgM: aOR = 2.51; p = 0.002). Age ≥45 years predicted Widal accuracy (aOR = 3.52; p = 0.016), while prior antibiotic use influenced RDT IgM accuracy (aOR = 2.87; p = 0.016). Widal test showed poor agreement with blood culture (κ = 0.16), RDT IgM showed moderate agreement (κ = 0.58), and PCR showed excellent agreement (κ = 0.92). Combined testing strategies improved diagnostic accuracy. Conclusion: Diagnostic accuracy depends on patient-related and clinical factors. The Widal test performed poorly as a standalone tool, while PCR and blood culture provided superior accuracy. Integrated diagnostic algorithms combining rapid tests with confirmatory methods are recommended to improve typhoid fever diagnosis and patient outcomes.
- 31 October 2026
- Published by GSC Online Press in World Journal of Advanced Research and Reviews
- Vol. 32 (1) , 1044-1056
Open AccessDETERMINANTS OF DIAGNOSTIC ACCURACY OF TYPHOID FEVER TESTS IN GA WEST MUNICIPALITY, GHANA
Background: Typhoid fever remains a significant public health challenge in Ghana due to overlapping clinical presentations with malaria and other febrile illnesses. Accurate diagnosis is essential for effective patient management and rational antibiotic use. This study evaluated determinants of diagnostic accuracy of typhoid fever tests in Ga West Municipality, Ghana. Methods: A cross-sectional comparative diagnostic accuracy study was conducted in seven health facilities. A total of 284 participants with suspected typhoid fever were recruited using consecutive sampling. Data were collected through structured questionnaires, clinical evaluations, and laboratory investigations, including Widal test, RDT IgM, blood culture (reference standard), and PCR. Statistical analyses included descriptive statistics, chi-square tests, multivariable logistic regression, Cohen's kappa, and McNemar's test at a 0.05 significance level. Results: Previous hospitalization was the strongest predictor of diagnostic accuracy for the Widal test (aOR = 7.74; p < 0.001) and RDT IgM (aOR = 6.96; p < 0.001). Symptom duration of 4–7 days was significantly associated with accuracy for both tests (Widal: aOR = 2.03; p = 0.019; RDT IgM: aOR = 2.51; p = 0.002). Age ≥45 years predicted Widal accuracy (aOR = 3.52; p = 0.016), while prior antibiotic use influenced RDT IgM accuracy (aOR = 2.87; p = 0.016). Widal test showed poor agreement with blood culture (κ = 0.16), RDT IgM showed moderate agreement (κ = 0.58), and PCR showed excellent agreement (κ = 0.92). Combined testing strategies improved diagnostic accuracy. Conclusion: Diagnostic accuracy depends on patient-related and clinical factors. The Widal test performed poorly as a standalone tool, while PCR and blood culture provided superior accuracy. Integrated diagnostic algorithms combining rapid tests with confirmatory methods are recommended to improve typhoid fever diagnosis and patient outcomes.
- Open Access
- 22 October 2026
- Published by Sciencedomain International in Biotechnology Journal International
- Vol. 30 (6) , 68-78
Diagnostic Accuracy of Rapid Diagnostic Tests for Typhoid Fever among Patients in Ga West Municipality, Ghana
Background: Typhoid fever remains a major public health challenge in many low- and middle-income countries, particularly in sub-Saharan Africa, where diagnostic limitations complicate the management of febrile illness. Rapid diagnostic tests (RDTs) are widely used as point-of-care tools in resource-limited settings, but concerns remain about their diagnostic accuracy in endemic settings. Aim: This study assessed the diagnostic accuracy of RDTs for typhoid fever among febrile patients in Ga West Municipality, Ghana. Methods: A quantitative cross-sectional design was used to recruit 284 febrile patients from health facilities in Ga West Municipality. Each participant underwent both RDT and blood culture testing, with blood culture serving as the reference standard. We calculated diagnostic accuracy parameters, including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), and conducted hypothesis tests to assess diagnostic misclassification. Results: The RDT showed an observed sensitivity (100%) and a negative predictive value of 100% in the study sample. However, specificity was poor (52.5%), and the positive predictive value was extremely low (5.8%), reflecting a substantial number of false-positive results. The diagnostic misclassification rate was 46.1%, and the reported hypothesis tests rejected all three null hypotheses (P<0.05). Conclusion: Although the RDT showed high observed sensitivity in this sample, its poor specificity and low PPV make it unreliable for definitive typhoid diagnosis in this setting. The study recommends confirmatory testing for all positive RDT results and supports improvements in diagnostic algorithms to guide clinical decision-making and reduce inappropriate antibiotic use in febrile illness management.
- 22 October 2026
- Published by Sciencedomain International in Biotechnology Journal International
- Vol. 30 (6) , 68-78
Open AccessDiagnostic Accuracy of Rapid Diagnostic Tests for Typhoid Fever among Patients in Ga West Municipality, Ghana
Background: Typhoid fever remains a major public health challenge in many low- and middle-income countries, particularly in sub-Saharan Africa, where diagnostic limitations complicate the management of febrile illness. Rapid diagnostic tests (RDTs) are widely used as point-of-care tools in resource-limited settings, but concerns remain about their diagnostic accuracy in endemic settings. Aim: This study assessed the diagnostic accuracy of RDTs for typhoid fever among febrile patients in Ga West Municipality, Ghana. Methods: A quantitative cross-sectional design was used to recruit 284 febrile patients from health facilities in Ga West Municipality. Each participant underwent both RDT and blood culture testing, with blood culture serving as the reference standard. We calculated diagnostic accuracy parameters, including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), and conducted hypothesis tests to assess diagnostic misclassification. Results: The RDT showed an observed sensitivity (100%) and a negative predictive value of 100% in the study sample. However, specificity was poor (52.5%), and the positive predictive value was extremely low (5.8%), reflecting a substantial number of false-positive results. The diagnostic misclassification rate was 46.1%, and the reported hypothesis tests rejected all three null hypotheses (P<0.05). Conclusion: Although the RDT showed high observed sensitivity in this sample, its poor specificity and low PPV make it unreliable for definitive typhoid diagnosis in this setting. The study recommends confirmatory testing for all positive RDT results and supports improvements in diagnostic algorithms to guide clinical decision-making and reduce inappropriate antibiotic use in febrile illness management.
- Open Access
- 22 October 2026
- Published by Bluemark Publishers in International Journal of Medical and Health Research
- Vol. 4 (4) , 16-26
Prevalence of Typhoid Fever among Patients in Ga West Municipality, Ghana: A Blood Culture-Based Cross-Sectional Study
Typhoid fever remains a significant public health burden in sub-Saharan Africa, yet many Ghanaian municipalities lack laboratory-confirmed epidemiological data. This study determined the prevalence of blood culture-confirmed typhoid fever and identified associated demographic and clinical factors among febrile patients in Ga West Municipality, Ghana. We conducted a cross-sectional study from October 2025 to March 2026 among 284 febrile patients attending seven health facilities. We processed blood cultures using standard microbiological procedures to isolate Salmonella Typhi. Data on sociodemographic and clinical characteristics were collected using a structured questionnaire. We used descriptive statistics, chi-square tests, and binary logistic regression for analysis, with statistical significance set at p < 0.05. The prevalence of blood culture-confirmed typhoid fever was 2.8% (95% CI: 1.2%-5.4%), with 8 positive isolates among 284 participants. Fever (85.6%), headache (56.7%), loss of appetite (49.6%), and abdominal pain (48.6%) were the most common presenting symptoms. Prevalence varied widely by age group (0.0%–10.0%), being highest among participants aged ≥60 years (10.0%) and lowest among those aged 18–29 years (0.0%). Five positive cases were male, and three were female; sex was not significantly associated with blood culture positivity. Clinical characteristics were significantly associated with blood culture positivity (χ² = 13.527, df = 6, p = 0.036) and independently predicted laboratory-confirmed typhoid fever (model χ² = 15.214, df = 3, p = 0.018), whereas sociodemographic factors were not significantly associated. The low prevalence of culture-confirmed typhoid fever (2.8%) among febrile patients underscores the importance of laboratory confirmation to avoid overdiagnosis and inappropriate antibiotic use. Strengthening blood culture capacity and implementing evidence-based clinical algorithms are essential for improving typhoid diagnosis, antimicrobial stewardship, and public health interventions in Ga West Municipality and similar settings in Ghana.
- 22 October 2026
- Published by Bluemark Publishers in International Journal of Medical and Health Research
- Vol. 4 (4) , 16-26
Open AccessPrevalence of Typhoid Fever among Patients in Ga West Municipality, Ghana: A Blood Culture-Based Cross-Sectional Study
Typhoid fever remains a significant public health burden in sub-Saharan Africa, yet many Ghanaian municipalities lack laboratory-confirmed epidemiological data. This study determined the prevalence of blood culture-confirmed typhoid fever and identified associated demographic and clinical factors among febrile patients in Ga West Municipality, Ghana. We conducted a cross-sectional study from October 2025 to March 2026 among 284 febrile patients attending seven health facilities. We processed blood cultures using standard microbiological procedures to isolate Salmonella Typhi. Data on sociodemographic and clinical characteristics were collected using a structured questionnaire. We used descriptive statistics, chi-square tests, and binary logistic regression for analysis, with statistical significance set at p < 0.05. The prevalence of blood culture-confirmed typhoid fever was 2.8% (95% CI: 1.2%-5.4%), with 8 positive isolates among 284 participants. Fever (85.6%), headache (56.7%), loss of appetite (49.6%), and abdominal pain (48.6%) were the most common presenting symptoms. Prevalence varied widely by age group (0.0%–10.0%), being highest among participants aged ≥60 years (10.0%) and lowest among those aged 18–29 years (0.0%). Five positive cases were male, and three were female; sex was not significantly associated with blood culture positivity. Clinical characteristics were significantly associated with blood culture positivity (χ² = 13.527, df = 6, p = 0.036) and independently predicted laboratory-confirmed typhoid fever (model χ² = 15.214, df = 3, p = 0.018), whereas sociodemographic factors were not significantly associated. The low prevalence of culture-confirmed typhoid fever (2.8%) among febrile patients underscores the importance of laboratory confirmation to avoid overdiagnosis and inappropriate antibiotic use. Strengthening blood culture capacity and implementing evidence-based clinical algorithms are essential for improving typhoid diagnosis, antimicrobial stewardship, and public health interventions in Ga West Municipality and similar settings in Ghana.
- Open Access
- 1 January 2026
- Published by AOSIS in African Vision and Eye Health
- Vol. 85 (1)
Intraocular pressure changes in bronchial asthma patients on treatment with inhalational steroids at a tertiary hospital in Southern Nigeria
Background: Elevated intraocular pressure (IOP) has been linked with the use of steroids. Aim: This study aimed to evaluate the association between the use of inhalational steroids and IOP among patients receiving treatment for bronchial asthma. Setting: It was a hospital-based cohort study at a tertiary hospital. Methods: Newly diagnosed asthma patients yet to commence inhalational steroids as cohort 1 gotten from the Respiratory clinic and Family Medicine Department. Cohort 2 was the control group who were age and sex matched with cohort 1 gotten from patients visiting the Ophthalmology Clinic with mild ocular disorders (without any respiratory symptoms). Their baseline IOP and central corneal thickness were used to adjust for corrected IOP. The patients in cohort 1 were placed on inhalational fluticasone 250 mcg twice daily. The IOP was thereafter measured monthly over a period of 5 months for both cohorts. Results: A total of 210 participants who were enrolled completed the study with 105 participants (mean age ± standard deviation 41.08 ± 11.10) in cohort 1 and another 105 participants (mean age 41.34 ± 10.55 years) in cohort 2. The mean IOP increase was 4.51 ± 2.15 right eye and 4.56 ± 2.03 left eye for cohort 1, while cohort had a mean IOP increase of 4.37 ± 2.40 and 3.83 ± 2.80 for the right and left eyes, respectively. There was no statistically significant difference in IOP increase from the first to third month of study, but a statistically significant association between inhaled steroid and IOP increase was observed in the fourth and fifth months of study (P < 0.005). The predictors of elevated IOP among study participants were primary or secondary level education and the presence of allergies. Conclusion: The study showed a significant increase in IOP after 3 months of use of inhalational steroids. Contribution: There is a need for physicians to liaise with ophthalmologists to monitor the IOP and associated risk of glaucoma of their asthmatic patients on inhalational steroids.
- 1 January 2026
- Published by AOSIS in African Vision and Eye Health
- Vol. 85 (1)
Open AccessIntraocular pressure changes in bronchial asthma patients on treatment with inhalational steroids at a tertiary hospital in Southern Nigeria
Background: Elevated intraocular pressure (IOP) has been linked with the use of steroids. Aim: This study aimed to evaluate the association between the use of inhalational steroids and IOP among patients receiving treatment for bronchial asthma. Setting: It was a hospital-based cohort study at a tertiary hospital. Methods: Newly diagnosed asthma patients yet to commence inhalational steroids as cohort 1 gotten from the Respiratory clinic and Family Medicine Department. Cohort 2 was the control group who were age and sex matched with cohort 1 gotten from patients visiting the Ophthalmology Clinic with mild ocular disorders (without any respiratory symptoms). Their baseline IOP and central corneal thickness were used to adjust for corrected IOP. The patients in cohort 1 were placed on inhalational fluticasone 250 mcg twice daily. The IOP was thereafter measured monthly over a period of 5 months for both cohorts. Results: A total of 210 participants who were enrolled completed the study with 105 participants (mean age ± standard deviation 41.08 ± 11.10) in cohort 1 and another 105 participants (mean age 41.34 ± 10.55 years) in cohort 2. The mean IOP increase was 4.51 ± 2.15 right eye and 4.56 ± 2.03 left eye for cohort 1, while cohort had a mean IOP increase of 4.37 ± 2.40 and 3.83 ± 2.80 for the right and left eyes, respectively. There was no statistically significant difference in IOP increase from the first to third month of study, but a statistically significant association between inhaled steroid and IOP increase was observed in the fourth and fifth months of study (P < 0.005). The predictors of elevated IOP among study participants were primary or secondary level education and the presence of allergies. Conclusion: The study showed a significant increase in IOP after 3 months of use of inhalational steroids. Contribution: There is a need for physicians to liaise with ophthalmologists to monitor the IOP and associated risk of glaucoma of their asthmatic patients on inhalational steroids.
- Open Access
- 30 April 2025
- Published by The Korean Society of Medical Informatics in Healthcare Informatics Research
- Vol. 31 (2) , 125-135
Developing an Explainable Artificial Intelligence System for the Mobile-Based Diagnosis of Febrile Diseases Using Random Forest, LIME, and GPT
Objectives: This study proposes a mobile-based explainable artificial intelligence (XAI) platform designed for diagnosing febrile illnesses.Methods: We integrated the interpretability offered by local interpretable model-agnostic explanations (LIME) and the explainability provided by generative pre-trained transformers (GPT) to bridge the gap in understanding and trust often created by machine learning models in critical healthcare decision-making. The developed system employed random forest for disease diagnosis, LIME for interpretation of the results, and GPT-3.5 for generating explanations in easy-to-understand language.Results: Our model demonstrated robust performance in detecting malaria, achieving precision, recall, and F1-scores of 85%, 91%, and 88%, respectively. It performed moderately well in detecting urinary tract and respiratory tract infections, with precision, recall, and F1-scores of 80%, 65%, and 72%, and 77%, 68%, and 72%, respectively, maintaining an effective balance between sensitivity and specificity. However, the model exhibited limitations in detecting typhoid fever and human immunodeficiency virus/acquired immune deficiency syndrome, achieving lower precision, recall, and F1-scores of 69%, 53%, and 60%, and 75%, 39%, and 51%, respectively. These results indicate missed true-positive cases, necessitating further model fine-tuning. LIME and GPT-3.5 were integrated to enhance transparency and provide natural language explanations, thereby aiding decision-making and improving user comprehension of the diagnoses.Conclusions: The LIME plots revealed key symptoms influencing the diagnoses, with bitter taste in the mouth and fever showing the highest negative influence on predictions, and GPT-3.5 provided natural language explanations that increased the reliability and trustworthiness of the system, promoting improved patient outcomes and reducing the healthcare burden.
- 30 April 2025
- Published by The Korean Society of Medical Informatics in Healthcare Informatics Research
- Vol. 31 (2) , 125-135
Open AccessDeveloping an Explainable Artificial Intelligence System for the Mobile-Based Diagnosis of Febrile Diseases Using Random Forest, LIME, and GPT
Objectives: This study proposes a mobile-based explainable artificial intelligence (XAI) platform designed for diagnosing febrile illnesses.Methods: We integrated the interpretability offered by local interpretable model-agnostic explanations (LIME) and the explainability provided by generative pre-trained transformers (GPT) to bridge the gap in understanding and trust often created by machine learning models in critical healthcare decision-making. The developed system employed random forest for disease diagnosis, LIME for interpretation of the results, and GPT-3.5 for generating explanations in easy-to-understand language.Results: Our model demonstrated robust performance in detecting malaria, achieving precision, recall, and F1-scores of 85%, 91%, and 88%, respectively. It performed moderately well in detecting urinary tract and respiratory tract infections, with precision, recall, and F1-scores of 80%, 65%, and 72%, and 77%, 68%, and 72%, respectively, maintaining an effective balance between sensitivity and specificity. However, the model exhibited limitations in detecting typhoid fever and human immunodeficiency virus/acquired immune deficiency syndrome, achieving lower precision, recall, and F1-scores of 69%, 53%, and 60%, and 75%, 39%, and 51%, respectively. These results indicate missed true-positive cases, necessitating further model fine-tuning. LIME and GPT-3.5 were integrated to enhance transparency and provide natural language explanations, thereby aiding decision-making and improving user comprehension of the diagnoses.Conclusions: The LIME plots revealed key symptoms influencing the diagnoses, with bitter taste in the mouth and fever showing the highest negative influence on predictions, and GPT-3.5 provided natural language explanations that increased the reliability and trustworthiness of the system, promoting improved patient outcomes and reducing the healthcare burden.
- Open Access
- 16 September 2024
- Published by MDPI AG in Tropical Medicine and Infectious Disease
- Vol. 9 (9) , 216
Enhancing the Interpretability of Malaria and Typhoid Diagnosis with Explainable AI and Large Language Models
Malaria and Typhoid fever are prevalent diseases in tropical regions, and both are exacerbated by unclear protocols, drug resistance, and environmental factors. Prompt and accurate diagnosis is crucial to improve accessibility and reduce mortality rates. Traditional diagnosis methods cannot effectively capture the complexities of these diseases due to the presence of similar symptoms. Although machine learning (ML) models offer accurate predictions, they operate as “black boxes” with non-interpretable decision-making processes, making it challenging for healthcare providers to comprehend how the conclusions are reached. This study employs explainable AI (XAI) models such as Local Interpretable Model-agnostic Explanations (LIME), and Large Language Models (LLMs) like GPT to clarify diagnostic results for healthcare workers, building trust and transparency in medical diagnostics by describing which symptoms had the greatest impact on the model’s decisions and providing clear, understandable explanations. The models were implemented on Google Colab and Visual Studio Code because of their rich libraries and extensions. Results showed that the Random Forest model outperformed the other tested models; in addition, important features were identified with the LIME plots while ChatGPT 3.5 had a comparative advantage over other LLMs. The study integrates RF, LIME, and GPT in building a mobile app to enhance the interpretability and transparency in malaria and typhoid diagnosis system. Despite its promising results, the system’s performance is constrained by the quality of the dataset. Additionally, while LIME and GPT improve transparency, they may introduce complexities in real-time deployment due to computational demands and the need for internet service to maintain relevance and accuracy. The findings suggest that AI-driven diagnostic systems can significantly enhance healthcare delivery in environments with limited resources, and future works can explore the applicability of this framework to other medical conditions and datasets.
- 16 September 2024
- Published by MDPI AG in Tropical Medicine and Infectious Disease
- Vol. 9 (9) , 216
Open AccessEnhancing the Interpretability of Malaria and Typhoid Diagnosis with Explainable AI and Large Language Models
Malaria and Typhoid fever are prevalent diseases in tropical regions, and both are exacerbated by unclear protocols, drug resistance, and environmental factors. Prompt and accurate diagnosis is crucial to improve accessibility and reduce mortality rates. Traditional diagnosis methods cannot effectively capture the complexities of these diseases due to the presence of similar symptoms. Although machine learning (ML) models offer accurate predictions, they operate as “black boxes” with non-interpretable decision-making processes, making it challenging for healthcare providers to comprehend how the conclusions are reached. This study employs explainable AI (XAI) models such as Local Interpretable Model-agnostic Explanations (LIME), and Large Language Models (LLMs) like GPT to clarify diagnostic results for healthcare workers, building trust and transparency in medical diagnostics by describing which symptoms had the greatest impact on the model’s decisions and providing clear, understandable explanations. The models were implemented on Google Colab and Visual Studio Code because of their rich libraries and extensions. Results showed that the Random Forest model outperformed the other tested models; in addition, important features were identified with the LIME plots while ChatGPT 3.5 had a comparative advantage over other LLMs. The study integrates RF, LIME, and GPT in building a mobile app to enhance the interpretability and transparency in malaria and typhoid diagnosis system. Despite its promising results, the system’s performance is constrained by the quality of the dataset. Additionally, while LIME and GPT improve transparency, they may introduce complexities in real-time deployment due to computational demands and the need for internet service to maintain relevance and accuracy. The findings suggest that AI-driven diagnostic systems can significantly enhance healthcare delivery in environments with limited resources, and future works can explore the applicability of this framework to other medical conditions and datasets.
- Open Access
- 14 August 2024
- Published by MDPI AG in Preprints.org
Enhancing the Interpretability of Malaria and Typhoid Diagnosis with Explainable AI and Large Language Models
Malaria and Typhoid fever are prevalent diseases in tropical regions, exacerbated by unclear protocols, drug resistance, and environmental factors. Prompt and accurate diagnosis is crucial to improve accessibility and reduce mortality rates. Traditional diagnosis methods cannot effectively capture the complexities of these diseases due to the presence of confusable symptoms. Although machine learning (ML) models offer accurate predictions, they operate as "black boxes" with non-interpretable decision-making processes, making it challenging for healthcare providers to comprehend how the conclusions are reached. This study employs explainable AI (XAI) models such as Local Interpretable Model-agnostic Explanations (LIME), and Large Language Models (LLMs) like ChatGPT, Gemini, and Perplexity, to clarify diagnostic results for healthcare workers, building trust and transparency in medical diagnostics. The models were implemented on Google Colab and Visual Studio Code because of their rich libraries and extensions. Results showed that the Random Forest model outperformed Extreme Gradient Boost and Support Vector Machines and important features for predicting Malaria and Typhoid were identified with the LIME plots. Among LLMs, ChatGPT 3.5 demonstrates a comparative advantage over Gemini and Perplexity, hence highlighting how crucial it is to use a hybrid strategy that combines ML, XAI, and LLMs to improve diagnostic performance and reliability in healthcare applications. The study recommends implementing the integrated ML, LIME, and LLM processes because it streamlines the development and maintenance process's workflow overall, makes use of resources more effectively, and improves explainability by allowing the LLM to consider the complete context that LIME provides.
- 14 August 2024
- Published by MDPI AG in Preprints.org
Open AccessEnhancing the Interpretability of Malaria and Typhoid Diagnosis with Explainable AI and Large Language Models
Malaria and Typhoid fever are prevalent diseases in tropical regions, exacerbated by unclear protocols, drug resistance, and environmental factors. Prompt and accurate diagnosis is crucial to improve accessibility and reduce mortality rates. Traditional diagnosis methods cannot effectively capture the complexities of these diseases due to the presence of confusable symptoms. Although machine learning (ML) models offer accurate predictions, they operate as "black boxes" with non-interpretable decision-making processes, making it challenging for healthcare providers to comprehend how the conclusions are reached. This study employs explainable AI (XAI) models such as Local Interpretable Model-agnostic Explanations (LIME), and Large Language Models (LLMs) like ChatGPT, Gemini, and Perplexity, to clarify diagnostic results for healthcare workers, building trust and transparency in medical diagnostics. The models were implemented on Google Colab and Visual Studio Code because of their rich libraries and extensions. Results showed that the Random Forest model outperformed Extreme Gradient Boost and Support Vector Machines and important features for predicting Malaria and Typhoid were identified with the LIME plots. Among LLMs, ChatGPT 3.5 demonstrates a comparative advantage over Gemini and Perplexity, hence highlighting how crucial it is to use a hybrid strategy that combines ML, XAI, and LLMs to improve diagnostic performance and reliability in healthcare applications. The study recommends implementing the integrated ML, LIME, and LLM processes because it streamlines the development and maintenance process's workflow overall, makes use of resources more effectively, and improves explainability by allowing the LLM to consider the complete context that LIME provides.
- Open Access
- 25 August 2022
- Published by Springer Nature in Diabetes Therapy
- Vol. 13 (10) , 1769-1778
The Case for Early and Universal Screening for Gestational Diabetes Mellitus: Findings from 9314 Pregnant Women in a Major City in Nigeria
Introduction Risk-based screening has been replaced by universal screening as the recommended course of care for gestational diabetes mellitus (GDM). As of 2016, no state in Nigeria had implemented a policy of universal screening for GDM. This research aimed to assess findings from a universal screening programme and its implication for scaling up universal and early screening for GDM. Methods This was a descriptive cross-sectional study conducted in Rivers State Nigeria between February 2017 and January 2020. Multistage sampling was used to recruit 9314 pregnant women from 30 primary, secondary, and tertiary health facilities in the state. An interviewer-administered structured questionnaire was used by trained healthcare workers to collect socio-demographic, obstetric and medical information. All study participants had a plasma glucose test on their first hospital visit and a diagnosis made using the World Health Organization (WHO) criteria. Data obtained was analysed using the IBM Statistical Package for Social Sciences (SPSS) version 23. Results Most women [5683 (61.0%)] were aged 25–34 (mean 29.60 ± 5.64) years. The prevalence of GDM in this study was 5.2% with a prevalence of GDM in the first, second and third trimesters of 4.9%, 4.2% and 6.7%, respectively. The prevalence of GDM among persons with a family history of diabetes was 13.2% (97 persons) while 4.6% (391 persons) without family history were diagnosed with GDM. Gestational age, family history of diabetes and age group were found to be significant predictors of GDM among the study participants after adjusting for confounding variables. Conclusion The practice of universal screening was useful in identifying GDM in 1 out of 20 pregnant women in the study sample. Screening at all trimesters was useful in identifying GDM. There is an urgent need to scale up early and universal screening for GDM across sub-Saharan Africa.
- 25 August 2022
- Published by Springer Nature in Diabetes Therapy
- Vol. 13 (10) , 1769-1778
Open AccessThe Case for Early and Universal Screening for Gestational Diabetes Mellitus: Findings from 9314 Pregnant Women in a Major City in Nigeria
Introduction Risk-based screening has been replaced by universal screening as the recommended course of care for gestational diabetes mellitus (GDM). As of 2016, no state in Nigeria had implemented a policy of universal screening for GDM. This research aimed to assess findings from a universal screening programme and its implication for scaling up universal and early screening for GDM. Methods This was a descriptive cross-sectional study conducted in Rivers State Nigeria between February 2017 and January 2020. Multistage sampling was used to recruit 9314 pregnant women from 30 primary, secondary, and tertiary health facilities in the state. An interviewer-administered structured questionnaire was used by trained healthcare workers to collect socio-demographic, obstetric and medical information. All study participants had a plasma glucose test on their first hospital visit and a diagnosis made using the World Health Organization (WHO) criteria. Data obtained was analysed using the IBM Statistical Package for Social Sciences (SPSS) version 23. Results Most women [5683 (61.0%)] were aged 25–34 (mean 29.60 ± 5.64) years. The prevalence of GDM in this study was 5.2% with a prevalence of GDM in the first, second and third trimesters of 4.9%, 4.2% and 6.7%, respectively. The prevalence of GDM among persons with a family history of diabetes was 13.2% (97 persons) while 4.6% (391 persons) without family history were diagnosed with GDM. Gestational age, family history of diabetes and age group were found to be significant predictors of GDM among the study participants after adjusting for confounding variables. Conclusion The practice of universal screening was useful in identifying GDM in 1 out of 20 pregnant women in the study sample. Screening at all trimesters was useful in identifying GDM. There is an urgent need to scale up early and universal screening for GDM across sub-Saharan Africa.