EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Calibration and Cross-Dataset Behavior in Malaria Microscopy with Parameter-Efficient Fine-Tuning: A Matched LoRA–PiSSA Evaluation

Sorio Boit, Rajvardhan Patil, Tiehang Duan

Diagnostics · 12.11.2026 · 16 (22) · 3749

DOI: 10.3390/diagnostics16223749 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Diagnostics, online erschienen am 12.11.2026. Verfasst von Sorio Boit und 2 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background/Objectives: Reliable evaluation of deep-learning methods for malaria microscopy requires assessing both classification performance and the quality of predicted infection probabilities across datasets. We compared low-rank adaptation (LoRA) and principal singular values and singular vectors adaptation (PiSSA) under matched parameter-efficient fine-tuning (PEFT).

Methods

An ImageNet-pretrained vision transformer (ViT-B/16) was adapted at ranks 2, 4, 8, and 16 using 27,558 National Institutes of Health (NIH) malaria cell images. One NIH grouped partition was held fixed across paired training seeds 42, 123, and 456. Models were evaluated on the NIH test split and, without target-domain adaptation, on BBBC041 P. vivax cell crops. The analysis was also replicated in invasive ductal carcinoma (IDC) histopathology. External probability and threshold-transfer analyses were post hoc.

Results

Across LoRA and PiSSA configurations, NIH area under the receiver operating characteristic curve (ROC–AUC) ranged from 0.9960 to 0.9970. PiSSA had lower rank-level mean Brier scores and expected calibration error (ECE) at every evaluated rank in NIH and IDC, although paired-run differences varied. On BBBC041, PiSSA-r16 achieved the highest PEFT ROC–AUC (0.8451) and area under the precision–recall curve (PR–AUC, 0.1588), while fully trained reference models performed better. Across all 12 matched external comparisons, PiSSA had lower overall Brier scores and ECE but higher infected-cell Brier scores. PiSSA had lower sensitivity at NIH-derived thresholds in 11 of 12 comparisons. All PEFT models overpredicted infection relative to the observed 2.70% prevalence.

Conclusions

Nearly identical internal discrimination concealed differences in external probability behavior and class-specific errors. PiSSA reduced overprediction, but neither adapter achieved adequate external calibration.

Zitieren

APA

Boit, S., Patil, R., & Duan, T. (2026). Calibration and Cross-Dataset Behavior in Malaria Microscopy with Parameter-Efficient Fine-Tuning: A Matched LoRA–PiSSA Evaluation. Diagnostics, 16(22), 3749. https://doi.org/10.3390/diagnostics16223749

Vancouver

Boit S, Patil R, Duan T. Calibration and Cross-Dataset Behavior in Malaria Microscopy with Parameter-Efficient Fine-Tuning: A Matched LoRA–PiSSA Evaluation. Diagnostics. 2026;16(22):3749. doi:10.3390/diagnostics16223749

BibTeX
@article{Boit2026,
  author = {Boit, Sorio and Patil, Rajvardhan and Duan, Tiehang},
  title = {Calibration and Cross-Dataset Behavior in Malaria Microscopy with Parameter-Efficient Fine-Tuning: A Matched LoRA–PiSSA Evaluation},
  journal = {Diagnostics},
  year = {2026},
  volume = {16},
  number = {22},
  pages = {3749},
  doi = {10.3390/diagnostics16223749}
}
Kostenlos starten →