Biomedical Imaging

Biomedical imaging encompasses a range of non-invasive techniques for visualising anatomical structures and physiological function. Modalities such as X-ray radiography, computed tomography (CT), magnetic resonance imaging (MRI), ultrasound and nuclear medicine (PET/SPECT) differ in their sources of energy and image-formation physics, but all now acquire data in digital form. Anatomic imaging prioritises high spatial resolution to depict fine morphological detail, whereas functional imaging assesses dynamic or molecular processes—blood flow, metabolic activity or tissue composition—with lower spatial but higher contrast and temporal sensitivity. Recent advances span molecular imaging agents that target specific biomarkers, super-resolution optical methods that localise single molecules beyond the diffraction limit, and artificial-intelligence algorithms that enhance image reconstruction and automate interpretation. Together, these innovations extend the reach of imaging from gross anatomy to cellular and even sub-molecular scales, transforming diagnosis, treatment planning and therapeutic monitoring across oncology, neurology, cardiology and beyond.

Research from Nature Portfolio

Ångström-resolution fluorescence microscopy has been achieved by a DNA-barcoding approach called resolution enhancement by sequential imaging (RESI). By sparsely labelling subsets of targets and precisely localising them in successive rounds, researchers have demonstrated sub-nanometre accuracy in intact cells, resolving individual protein complexes and even base-pair distances in DNA origami constructs. This method bridges conventional microscopy and structural biology, enabling intramolecular imaging under ambient conditions.

Radiomic analysis of corticospinal tractography in high-grade glioma patients has uncovered that conventional measures such as streamline count lack functional specificity. Instead, radiomic metrics—fractional anisotropy energy, surface–volume ratio and kurtosis—correlate strongly with preoperative motor deficits and clinical paresis. These quantitative markers promise objective, non-invasive assessment of white matter integrity to guide surgical planning and prognostication.

Research from all publishers

A comprehensive review on machine-learning-aided thermal imaging for diabetic foot management synthesises advances in two-dimensional signal processing and deep-learning classifiers. It highlights how subtle plantar temperature gradients—often imperceptible to the human eye—can predict ulceration risk. The authors stress the importance of large annotated thermogram datasets and robust algorithms to enable early warning systems in home and clinical settings.

Deep-STORM applies convolutional neural networks to reconstruct super-resolution images from sparsely activated fluorophores. Trained on simulated and experimental data, the network delivers localisation precision comparable to standard STORM but with reduced acquisition time and lower light exposure. This accelerates single-molecule microscopy workflows and diminishes phototoxicity in live-cell imaging.

A phase-transition nanodroplet review in cancer theranostics describes how perfluorocarbon-filled droplets convert into contrast-enhancing microbubbles under ultrasound or laser stimuli. These constructs enable real-time dual-modality imaging (ultrasound and photoacoustics) and potentiate photothermal or cavitation-enhanced therapies. Combining image guidance with triggered drug or gas release illustrates the growing synergy between diagnostics and treatment.

Biomedical Imaging publication trend

The graph below shows the total number of articles in biomedical imaging across all publications each year (not limited to Nature Index journals).

Technical terms

Spatial resolution: The smallest object size that can be distinguished in an image, typically measured in millimetres or micrometres.

Contrast resolution: The ability to discern small differences in image intensity, crucial for distinguishing adjacent tissues with similar properties.

Functional imaging: Techniques that reveal physiological or molecular processes—such as perfusion, metabolism or gene expression—rather than purely anatomical structure.

Radiomics: The extraction of large numbers of quantitative features from medical images, capturing shape, texture and signal intensity patterns for diagnostic or prognostic modelling.

Super-resolution: A set of optical or computational methods that achieve image resolutions beyond the diffraction limit of light microscopy, often by localising individual fluorescent molecules.

Resolution enhancement by sequential imaging (RESI): A fluorescence microscopy strategy that sparsely labels and sequentially images barcoded targets to achieve ångström-scale localisation in intact cells.

References

1. Ångström-resolution fluorescence microscopy. Nature (2023).
2. Radiomic white matter parameters of functional integrity of the corticospinal tract in high-grade glioma. Scientific Reports (2024).
3. Advances in Machine Learning-Aided Thermal Imaging for Early Detection of Diabetic Foot Ulcers: A Review. Biosensors (2024).
4. Phase-Transition Nanodroplets for Real-Time Photoacoustic/Ultrasound Dual-Modality Imaging and Photothermal Therapy of Sentinel Lymph Node in Breast Cancer. Scientific Reports (2017).