Instytut Podstawowych Problemów Techniki

Abstract:
Simple Summary

Thyroid cancer includes several types that differ in how they grow and how they should be treated. Although ultrasound is widely used to examine thyroid nodules, it can be difficult to determine which type of cancer is present using standard imaging alone. In this study, we applied a computer-based method to automatically measure and analyze ultrasound features of thyroid tumors. By using machine learning techniques, we distinguished between three common types of thyroid cancer: papillary, follicular, and medullary. We found that certain features, such as tumor shape, brightness, and internal structure, were helpful in identifying the cancer subtype. This approach could support doctors in making more accurate diagnoses, reduce unnecessary procedures such as biopsies, and guide more personalized treatment decisions.

Keywords:
thyroid cancer, ultrasound imaging, quantitative analysis, machine learning, papillary thyroid carcinoma, follicular thyroid carcinoma, medullary thyroid carcinoma

Abstract:
Breast cancer remains the most frequently diagnosed malignancy among women world- wide, with rising incidence and significant biological heterogeneity influencing treatment strategies. Neoadjuvant chemotherapy (NAC) has become a standard option, particularly for aggressive molecular subtypes, underscoring the need for sensitive tools to monitor early treatment response. Conventional imaging (MRI, CT, mammography, and B-mode ultrasound) primarily captures morphological change, often lagging biological alterations. Quantitative ultrasound (QUS) is an emerging modality that characterizes tumor mi- crostructure and yields reproducible, operator-independent biomarkers. This narrative review synthesizes current evidence, clarifies the conceptual framework (spectral, ampli- tude, and attenuation metrics; parametric maps and texture), highlights clinical applications and limitations, and outlines future directions for integrating QUS into NAC response assessment in breast cancer.

Keywords:
breast cancer, neoadjuvant chemotherapy, quantitative ultrasound (QUS), treatment response monitoring

Abstract:
We numerically investigate the dynamics of polydisperse microbubble clouds driven by ultrasound, focusing on the generation of half-frequency subharmonics (HSH). This study demonstrates that polydispersity in bubble size distributions, in the case of a~cloud with interacting bubbles, greatly expands the frequency range of ultrasound for HSH generation—contrary to identically sized, non-interacting bubbles, where half-frequency subharmonics exist only at a~single excitation frequency corresponding to twice the resonance frequency of an individual bubble. Additionally, we identify a~robust low-frequency subharmonic that is governed by the resonance of the largest bubbles in the cloud. Finally, we thoroughly address issues of numerical instability by conducting stability-aware simulations and independent solver checks, thereby offering practical guidance for the reliable computation of bubble cloud dynamics. These findings provide new insights into the dynamics of bubble clouds, which are relevant to biomedical ultrasound and other phenomena modelled by similar nonlinear differential equations.

Keywords:
nonlinear microbubble dynamics, half-frequency subharmonics, polydispersity, bubble interactions, numerical simulation

Abstract:
This article presents experimental, theoretical, and numerical studies of the propagation of guided ultrasonic waves in a layered epitaxial structure of garnet compounds. A microscopic model, which yields dispersion equations based on material and geometrical properties, is developed. Acoustic microscopy experiments on a YAG:Ce crystal substrate and an epitaxial structure containing LuAG:Ce single crystalline films, grown using the liquid phase epitaxy growth method onto a YAG:Ce crystal substrate, reveal distinct phase velocity behaviors. The YAG substrate exhibits consistent velocities, minimally influenced by frequency, while the epitaxial structure shows dispersion, indicating frequency-dependent phase velocities. Experimental results are compared with numerically calculated dispersion curves, showing high agreement in the low-frequency range and minor deviations at higher frequencies. An optimization procedure is developed and applied, starting with the YAG substrate and extending to the LuAG:Ce film/YAG:Ce crystal epitaxial structure. The procedure allows for the extraction of material properties, offering valuable insights into the mechanical characteristics of the all-solid-state LuAG:Ce film/YAG:Ce crystal structure. This research represents a significant advancement in understanding ultrasonic wave dynamics in layered structures, particularly unveiling previously unexplored elastic properties of LuAG:Ce single crystalline films as a well-known scintillation material.

Abstract:
This paper presents a mathematical and numerical framework for modelling and parametrising tumour evolution dynamics to enhance computer-aided diagnosis and personalised treatment. The model comprises six differential equations describing cancer cell and blood vessel concentrations, tissue stiffness, Ki-67 marker distribution, and the apparent velocity of marker propagation. These equations are coupled through S-functions with adjustable coefficients. An inverse problem approach calibrates the model by fitting adjustable coefficients to patient-specific clinical data, thereby enabling disease progression and treatment response simulations. By integrating historical and prospective patient data supported by machine learning algorithms, this framework holds promise as a robust decision-support tool for optimising therapeutic strategies.

Keywords:
Tumour modelling, Personalised treatment, Breast cancer, Navier–stokes, Evolution simulation, Machine learning

Abstract:
This study introduces a spatially distributed diffusion model based on a Navier–Stokes formulation with a pseudo-velocity field, providing a framework for modelling cellular growth dynamics within diseased tissues. Five coupled partial differential equations describe diseased cell development within a two-dimensional spatial domain over time. A pseudo-velocity field mimics biomarker concentration increasing over time and space, influencing tumour growth dynamics. An

Keywords:
Tumour growth, Cellular growth, Cancer, Navier–stokes, Diffusion, Finite element method

Abstract:
The aim of the article is to highlight the key role of artificial intelligence in modern oncology. The search for scientific publications was carried out through the following web search engines: PubMed, PMC, Web of Science, Scopus, Embase and Ebsco. Artificial intelligence plays a special role in oncology and is considered to be the future of oncology. The largest application of artificial intelligence in oncology is in diagnostics (more than 80%),
particularly in radiology and pathology. This can help oncologists not only detect cancer at an early stage but also forecast the possible development of the disease by using predictive models. Artificial intelligence plays a special role in clinical trials. AI makes it possible to
accelerate the discovery and development of new drugs, even if not necessarily successfully. This is done by detecting new molecules. Artificial intelligence enables patient recruitment by combining diverse demographic and medical patient data to match the requirements of a given research protocol. This can be done by reducing population heterogeneity, or by prognostic and predictive enrichment. The effectiveness of artificial intelligence in oncology
depends on the continuous learning of the system based on large amounts of new data but the development of artificial intelligence also requires the resolution of some ethical and legal issues.

Keywords:
artificial intelligence, intelligent oncology, cancer prediction, cancer screening

Abstract:
Objective: Focused ultrasound (FUS) with intravenously administered microbubbles (MBs) enables different therapeutic effects, e.g. localized opening of the blood-brain barrier (BBB). Acoustic activation of MBs under FUS induces mechanical effects---primarily stable or inertial cavitation - that can reversibly disrupt endothelial tight junctions without permanent tissue damage. MB acoustic emissions are widely used as indicators of cavitation activity and, by extension, treatment efficacy and safety. While some aspects of microbubble behavior under different exposure conditions are known, the overall influence of various parameter combinations on cavitation dose remains incompletely described. Approach: This study examined how MB concentration (0.0008-0.4% V/V), peak negative pressure (61.5-2600 kPa), pulse duration (95-952 µs), and effective sonication time affect cavitation activity in a flow setup. Cavitation was quantified as a cavitation dose which was divided into three types: stable harmonic (SCD_har), ultraharmonic (SCD_ultra), and broadband (ICD) emissions. Results: SCD_har and ICD increased mostly monotonically with pressure and MB concentration, while SCD_ultra peaked at intermediate values suggesting optimal parameters for the control of the ultrasound BBB opening procedure. Cavitation metrics showed 10% reproducibility. Critically, we found that for fixed effective sonication times, increasing the number of pulses led to significantly change the response of cavitation dose in time. To our knowledge, this relationship has not been studied before, change of pulse length was always related to effective sonication time. Our results suggests that pulse number is an important factor of how MB oscillate, introducing a potentially pivotal control parameter for therapeutic ultrasound.Significance: These findings provide new insights into MB dynamics and highlight pulse count as an underrecognized yet potentially important factor in protocol design. This perspective may inform refinements of FUS treatments, contributing to greater safety, consistency, and efficacy, and represents a step toward optimizing ultrasonic BBB opening strategies.

Keywords:
Microbubbles, Nonlinear oscillations, Ultrasound