With the power of parallel processing, large datasets, and fast computational resources, deep neural networks (DNNs) have outperformed highly trained and experienced human experts in medical applications. However, the large global community of healthcare professionals, many of whom routinely face potentially life-or-death outcomes with complex medicolegal consequences, have yet to embrace this powerful technology. The major problem is that most current AI solutions function as a metaphorical black-box positioned between input data and output decisions without a rigorous explanation for their internal processes. With the goal of enhancing trust and improving acceptance of artificial intelligence– (AI) based technology in clinical medicine, there is a large and growing effort to address this challenge using eXplainable AI (XAI), a set of techniques, strategies, and algorithms with an explicit focus on explaining the “hows and whys” of DNNs. Here, we provide a comprehensive review of the state-of-the-art XAI techniques concerning healthcare applications and discuss current challenges and future directions. We emphasize the strengths and limitations of each category, including image, tabular, and textual explanations, and explore a range of evaluation metrics for assessing the effectiveness of XAI solutions. Finally, we highlight promising opportunities for XAI research to enhance the acceptance of DNNs by the healthcare community.

Background: The chemokine (C-C motif) ligand 2 (CCL2) is a monocyte chemoattractant protein that mediates macrophage recruitment and migration during peripheral and central nervous system (CNS) inflammation.Methods: To determine the impact of CCL2 in inflammationin vivoand to elucidate the CCL2-induced polarization of activated brain microglia, we delivered CCL2 into the brains of wild-type mice via recombinant adeno-associated virus serotype 9 (rAAV-9) driven by the chicken β-actin promoter. We measured microglial activation using histological and chemical measurement and recruitment of monocytes using histology and flow cytometry.Results: The overexpression of CCL2 in the CNS induced significant activation of brain resident microglia. CD45 and major histocompatibility complex class II immunoreactivity significantly increased at the sites of CCL2 administration. Histological characterization of the microglial phenotype revealed the elevation of “classically activated” microglial markers, such as calgranulin B and IL-1β, as well as markers associated with “alternative activation” of microglia, including YM1 and arginase 1. The protein expression profile in the hippocampus demonstrated markedly increased levels of IL-6, GM-CSF and eotaxin (CCL-11) in response to CCL2, but no changes in the levels of other cytokines, including TNF-α and IFN-γ. Moreover, real-time PCR analysis confirmed increases in mRNA levels of gene transcripts associated with neuroinflammation following CCL2 overexpression. Finally, we investigated the chemotactic properties of CCL2in vivoby performing adoptive transfer of bone marrow–derived cells (BMDCs) isolated from donor mice that ubiquitously expressed green fluorescent protein. Flow cytometry and histological analyses indicated that BMDCs extravasated into brain parenchyma and colabeled with microglial markers.Conclusion: Taken together, our results suggest that CCL2 strongly activates resident microglia in the brain. Both pro- and anti-inflammatory activation of microglia were prominent, with no bias toward the M1 or M2 phenotype in the activated cells. As expected, CCL2 overexpression actively recruited circulating monocytes into the CNS. Thus, CCL2 expression in mouse brain induces microglial activation and represents an efficient method for recruitment of peripheral macrophages.

Infant cries serve as a primary indicator of distress and pain; however, distinguishing pain-related cries from those triggered by other needs remains a challenging task, even for trained professionals. Timely and accurate pain assessment is essential for appropriate medical intervention, particularly in preverbal infants who cannot express their needs verbally. Recently, Deep Learning (DL) models have demonstrated significant potential in addressing this challenge by enabling automated and efficient pain assessment through audio signal processing. In this survey, we review methods for pain assessment from infant cry sounds, covering deep learning architectures, modern Transformer-based models, and emerging Vision-Language Model (VLM) pipelines. The review includes approaches that integrate Mel-spectrogram representations of cry audio with multimodal model frameworks to improve robustness, interpretability, and cross-modal reasoning in pain detection. By summarizing recent advancements and identifying limitations and open challenges in current methodologies, this review aims to provide insights into future research directions that may enhance the robustness, generalizability, and clinical applicability of automated infant pain assessment tools.