Manav Doshi
Research Subjects
Manav Doshi has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Chris A. Rees
Department of Pediatrics Emory University School of Medicine Atlanta Georgia USA
1 shared publication - Elias Kumbakumba
Department of Pediatrics and Child Health, Faculty of Medicine Mbarara University of Science and Technology Mbarara Uganda
1 shared publication - Jerome Kabakyenga
Mbarara University of Science and Technology
1 shared publication - John Mark Ansermino
University of British Columbia
1 shared publication Martina Knappett
1 shared publication- Matthew O. Wiens
University of British Columbia
1 shared publication
Latest Publications
- Open Access
- 29 July 2026
- Published by Elsevier in SSRN Electronic Journal
Soft-Label Machine Learning for Verbal Autopsy: Incorporating Diagnostic Uncertainty in Cause-of-Death Estimation
PurposeVerbal autopsy (VA) is widely used to infer causes of death (COD) in settings where medical certification and autopsy are unavailable. Physician-certified COD determination after review of verbal autopsy is commonly used, but it is time- and effort-intensive, and often cloaked in diagnostic uncertainty. We introduce a framework for soft-label (probabilistic) supervised VA modelling that accommodates physician diagnostic uncertainty in settings where gold-standard labels are unavailable, and compare this to a traditional hard-label (deterministic) approach.MethodsUsing pediatric post-discharge mortality VA data (N = 356) from a multisite cohort study in Uganda, we constructed soft labels (probability vectors) from primary and alternative physician-assigned causes, along with associated confidence levels. XGBoost was used to create and compare models.ResultsThe soft-label model had population-level performance, as assessed by the cause-specific mortality fraction (CSMF) accuracy, of 0.88 ± 0.04, compared with 0.78 ± 0.03 for hard labels. Individual-level performance was slightly lower with a chance-corrected concordance of 0.38 ± 0.05 vs 0.40 ± 0.06.ConclusionsBy explicitly modelling physician uncertainty, soft-label learning substantially improves population-level cause-specific mortality estimation derived from VA data. The proposed framework is well-suited for mortality surveillance and policy-relevant applications in low-resource settings.
- 29 July 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessSoft-Label Machine Learning for Verbal Autopsy: Incorporating Diagnostic Uncertainty in Cause-of-Death Estimation
PurposeVerbal autopsy (VA) is widely used to infer causes of death (COD) in settings where medical certification and autopsy are unavailable. Physician-certified COD determination after review of verbal autopsy is commonly used, but it is time- and effort-intensive, and often cloaked in diagnostic uncertainty. We introduce a framework for soft-label (probabilistic) supervised VA modelling that accommodates physician diagnostic uncertainty in settings where gold-standard labels are unavailable, and compare this to a traditional hard-label (deterministic) approach.MethodsUsing pediatric post-discharge mortality VA data (N = 356) from a multisite cohort study in Uganda, we constructed soft labels (probability vectors) from primary and alternative physician-assigned causes, along with associated confidence levels. XGBoost was used to create and compare models.ResultsThe soft-label model had population-level performance, as assessed by the cause-specific mortality fraction (CSMF) accuracy, of 0.88 ± 0.04, compared with 0.78 ± 0.03 for hard labels. Individual-level performance was slightly lower with a chance-corrected concordance of 0.38 ± 0.05 vs 0.40 ± 0.06.ConclusionsBy explicitly modelling physician uncertainty, soft-label learning substantially improves population-level cause-specific mortality estimation derived from VA data. The proposed framework is well-suited for mortality surveillance and policy-relevant applications in low-resource settings.