Investigating White Blood Cells as a Source of False-Positive Malaria Parasite Detection in African Blood-Smear Images

Samuel A. Adeniji; Goodness C. Obasi; Chris-Victor Ntwali; Aondana M. Iorumbur; Confidence Raymond; Lowami Uwimana; Ahmed Tahiru Issah

cs.CVAug 30, 2026

Investigating White Blood Cells as a Source of False-Positive Malaria Parasite Detection in African Blood-Smear Images

Authors: Samuel A. Adeniji, Goodness C. Obasi, Chris-Victor Ntwali, Aondana M. Iorumbur, Confidence Raymond, Lowami Uwimana, Ahmed Tahiru Issah

Organizations: Carnegie Mellon University Africa, Kigali, Rwanda · Department of Neurology, University of Rwanda, Kigali, Rwanda · Department of Physics, Federal University of Technology, Minna, Nigeria · Department of Biomedical Engineering, McGill University, Montreal, Canada

Abstract

White blood cells (WBCs) present on every Giemsa-stained thick blood smear share visual properties with early-stage Plasmodium falciparum ring-form trophozoites: small size, round morphology, and intense purple staining. They are a plausible but untested source of false positives in parasite-only detectors. We trained two YOLOv12s models on the Lacuna Malaria Detection dataset (8,000 images from Uganda and Ghana): Model A with parasite labels only, and Model B with both parasite and WBC labels. Seven independent spatial and statistical analyses tested whether false positive (FP) predictions cluster near WBC locations. All seven refute the hypothesis. In both models, 95% of FPs are pure background detections (IoU below 0.10 against any ground-truth box); zero are WBC class confusions. Ripley's Cross-K analysis shows spatial repulsion between FP centroids and WBC positions at every radius tested. Model B outperforms Model A overall (mAP50 0.859 vs. 0.755), and the advantage is uniform across all WBC-proximity bands, pointing to multi-task representation learning rather than WBC suppression as the cause. False positives arise from Giemsa stain debris and preparation artifacts. Effective mitigation requires staining artifact augmentation and annotation of unannotated early-stage ring forms rather than WBC labeling alone.

Figures & tables

Metric	Model A	Model B	B (parasite)	B (WBC)
mAP50	0.755	0.859	0.748	0.970
mAP50-95	0.302	0.429	0.299	0.558
Precision	0.679	0.801	0.686	0.917
Recall	0.734	0.846	0.728	0.965
F1	0.705	0.823	0.706	0.940

Table 1: Test-set detection performance. Best per-row values in bold.

Error	Definition	A	A%	B	B%
Bkg	IoU $<$ 0.10 with all GT	1,883	95.1	2,335	95.3
Loc	$0.10\leq$ IoU $<$ 0.50, correct cls	96	4.9	114	4.7
Cls	IoU $\geq$ 0.50, wrong cls	0	0.0	0	0.0
Dup	IoU $\geq$ 0.50, correct, GT consumed	0	0.0	0	0.0
Both	$0.10\leq$ IoU $<$ 0.50, wrong cls	0	0.0	0	0.0
Miss (FN)	GT matched by no prediction	294	12.4% GT	285	10.9% GT

Table 2: TIDE error decomposition for FP predictions.

Band (px)	A FPs	A rate	A% FPs	B FPs	B rate	B% FPs
0–30	3	42.9%	0.4%	3	17.6%	0.3%
30–75	14	34.2%	1.8%	18	23.1%	1.7%
75–150	40	38.5%	5.2%	50	29.2%	4.7%
150–300	99	35.7%	13.0%	127	29.3%	12.0%
${>}300$	608	32.9%	79.6%	864	27.8%	81.4%

Table 3: FP proportions and per-band FP rates by WBC-proximity band.

Model	$N$	Mean FP	DI	NB AIC	Pois. AIC
Model A	306	2.79	9.52	1329	2140
Model B	312	3.80	7.83	1536	2338

Table 4: Overdispersion test results for false-positive (FP) counts per image. Negative binomial models achieved lower AIC than Poisson models for both detectors.

Band (px)	$N$	A $\checkmark$ B $\times$	A $\times$ B $\checkmark$	BH-significant	Favours
0–30	22	0	1	No ( $n$ too small)	Neither
30–75	231	2	12	Yes	Model B
75–150	584	4	26	Yes	Model B
150–300	776	14	37	Yes	Model B
$>300$	423	10	30	Yes ( $p=0.003$ )	Model B

Table 5: Stratified McNemar test by WBC-proximity band (BH-corrected).

Explore similar work

MalariAI: A Label-Resilient Decoupled Framework for Universal Cell Segmentation and Explainable Stage Classification in Dense Malaria Blood Smears
Jul 1, 2026Kaysarul Anas Apurba, Md Hasibul Hasan, Mohammed Ali +1Explainable Medical Image Analysis Microscopy Image Analysis
Department of Electrical and Computer Engineering, North South University, Dhaka 1229, Bangladesh · Department of Computer Science and Engineering, International University of Business Agriculture and Technology (IUBAT), Dhaka 1230, Bangladesh
CoSWA-YOLOv12: Scale-Invariant Tiny Object Detection and Segmentation of Malaria Parasites
Aug 25, 2026Ahmed Tahiru Issah, Carine MukamakuzaSmall Object Detection Object Detection
Carnegie Mellon University, Kigali, Rwanda
EMFE: A lightweight, explainable machine learning framework for malaria cell classification
Aug 25, 2026Md Abdullah Al Kafi, Walayat Hussain, Mousumi Karmakar +2Medical Image Classification Classification
Multidisciplinary Action Research (MARS) Lab, Daffodil International University, Dhaka, Bangladesh · Artificial Intelligence for Decision Excellence (AIDX) Lab, Peter Faber Business School, Australian Catholic University, North Sydney, Australia · Department of Computer Science and Engineering, Alliance University, Bangalore, India +1

Jul 1, 2026, eess.IVJ/K move · Enter open · S save

MalariAI: A Label-Resilient Decoupled Framework for Universal Cell Segmentation and Explainable Stage Classification in Dense Malaria Blood Smears

Kaysarul Anas Apurba, Md Hasibul Hasan, Mohammed Ali, Tanzilur Rahman

Department of Electrical and Computer Engineering, North South University, Dhaka 1229, Bangladesh · Department of Computer Science and Engineering, International University of Business Agriculture and Technology (IUBAT), Dhaka 1230, Bangladesh

Explainable Medical Image Analysis Microscopy Image Analysis

Open paper AskPDF arXiv

Automated malaria diagnosis from blood smear microscopy is a critical challenge in global health AI; in resource-limited settings, the scarcity of expert microscopists remains the primary bottleneck to timely and accurate diagnosis. Three compounding failure modes prevent reliable clinical deployment of existing deep learning systems. First, end-to-end detectors treat unannotated cells as background during training, producing recall figures that are strongly influenced by annotation completeness rather than reflecting true cell recovery. Second, Non-Maximum Suppression tends to suppress valid detections in dense smear regions where infection counts matter most. Third, existing whole-slide detection pipelines lack per-cell spatial evidence for clinical audit, despite image-level explainability methods such as Grad-CAM having been applied to malaria image classification tasks. We present MalariAI, a two-stage decoupled framework that addresses all three failure modes in a unified pipeline. Stage 1 applies an annotation-agnostic distance-transform guided watershed algorithm to isolate every cell in a full 1600x1200 blood smear image, recovering 75.95% of ground-truth cells by centroid localisation across the 120-image NIH BBBC041 test set without any ground-truth input. Stage 2 fine-tunes EfficientNet-B0 with Focal Loss (gamma = 2.0, per-class inverse-frequency weights) on 64x64 crops, achieving 98.36% overall classification accuracy with 87.5% and 75.0% per-class accuracy on the rare schizont and gametocyte stages, compared to only 24.57% and 25.95% AP for a Faster R-CNN baseline on the same classes. Grad-CAM++ heatmaps generated per detected cell provide instance-level spatial evidence for clinical audit, enabling microscopists to verify model predictions at the individual parasite level without sacrificing classification performance.

Abstract

Explore similar work

MalariAI: A Label-Resilient Decoupled Framework for Universal Cell Segmentation and Explainable Stage Classification in Dense Malaria Blood Smears

CoSWA-YOLOv12: Scale-Invariant Tiny Object Detection and Segmentation of Malaria Parasites

EMFE: A lightweight, explainable machine learning framework for malaria cell classification