Background: Automated classification of thyroid Fine Needle Aspiration Biopsy (FNAB) images faces challenges in limited data, inter-observer variability, and computational cost. Efficient, interpretable models are crucial for clinical support. Objective: To develop and externally validate a deep learning system for multi-class thyroid FNAB image classification into three key categories directly guiding post-biopsy treatment in Vietnam: Benign (Bethesda II), Indeterminate/Suspicious (BI, III, IV, V), and Malignant (BVI), achieving high diagnostic accuracy with low computational overhead. Methods: Our pipeline features: (1) YOLOv10 cell cluster detection for informative sub-region extraction/noise reduction; (2) curriculum learning sequencing localized crops to full images for multi-scale capture; (3) adaptive lightweight EfficientNetB0 (4M parameters) balancing performance/efficiency; and (4) a Transformer-inspired module for multi-scale/multi-region analysis. External validation used 1,015 independent FNAB images. Results: ThyroidEffi Basic achieved macro F1 of 89.19% and AUCs of 0.98 (Benign), 0.95 (Indeterminate/Suspicious), 0.96 (Malignant) on the internal test set. External validation yielded AUCs of 0.9495 (Benign), 0.7436 (Indeterminate/Suspicious), 0.8396 (Malignant). ThyroidEffi Premium improved macro F1 to 89.77%. Grad-CAM highlighted key diagnostic regions, confirming interpretability. The system processed 1000 cases in 30 seconds, demonstrating feasibility on widely accessible hardware. Conclusions: This work demonstrates that high-accuracy, interpretable thyroid FNAB image classification is achievable with minimal computational demands.

A cytopathologist at SRL Diagnostics' Central Reference Laboratory in Mumbai, screens a Pap smear sample for the screening of cervical cancer under his microscope. His trained eyes work with an apparent effortlessness. However, there is an unspoken urgency in his actions as he strives to complete the set of samples for the day. Along with his team of five members, he screens about 200 slides for cervical cancer every day, apart from another 100 slides for diagnosing other types of cancers. SRL Diagnostics, the largest diagnostics laboratory company in India, has been witnessing an increase in the demand for cervical cancer screening. According to estimates by the World Health Organization (WHO), cervical cancer is the fourth most frequent cancer among women worldwide.

SRL Diagnostics (the largest diagnostics laboratory company in India), which has been working with Microsoft to develop digital pathology services, has begun assessing its Cervical Cancer Image Detection application programming interface (API). The Cervical Cancer Image Detection API operates on Microsoft's Azure, and the research has been conducted in Mumbai. According to TechCrunch the system is capable of rapidly screening liquid-based cytology slide images in order to detect cervical cancer at the early stages. Digital data is then transmitted back to pathologists in the originating laboratory. Detecting cervical cancer early is key to increasing survival rates.

More women in India die from cervical cancer than in any other country. This preventable disease kills around 67,000 women in India every year, more than 25% of the 260,000 deaths worldwide. Effective screening and early detection can help reduce its incidence, but part of the challenge -- and there are several parts -- today is that the testing process to detect the onset of the disease is unbearably time-consuming. This is because the existing methodology that cytopathologists use is time consuming to begin with, but also because there are very few of them in the nation. Could AI speed this up?

We consider preoperative prediction of thyroid cancer based on ultra-high-resolution whole-slide cytopathology images. Inspired by how human experts perform diagnosis, our approach first identifies and classifies diagnostic image regions containing informative thyroid cells, which only comprise a tiny fraction of the entire image. These local estimates are then aggregated into a single prediction of thyroid malignancy. Several unique characteristics of thyroid cytopathology guide our deep-learning-based approach. While our method is closely related to multiple-instance learning, it deviates from these methods by using a supervised procedure to extract diagnostically relevant regions. Moreover, we propose to simultaneously predict thyroid malignancy, as well as a diagnostic score assigned by a human expert, which further allows us to devise an improved training strategy. Experimental results show that the proposed algorithm achieves performance comparable to human experts, and demonstrate the potential of using the algorithm for screening and as an assistive tool for the improved diagnosis of indeterminate cases.