Martynas Maciulevičius

Background and objectives: This study presents an interpretable two-level machine-learning framework for localizing neoplastic tissue in canine and feline H&E-stained histological images and examining how the predicted cancerous tissue fraction affects tumor classification.,Methods: A sliding-window k-nearest-neighbor (KNN) classifier generated cancerous-tissue heatmaps and estimated the cancerous tissue fraction in each region of interest (ROI). ROIs were sorted into 14 fraction groups. Color, texture, and statistical descriptors were used to classify mastocytoma and soft tissue sarcoma with linear discriminant analysis (LDA), support vector machines (SVM), and random forests (RF), evaluated by nested 10-fold ROI-level cross-validation.,Results: KNN distinguished cancerous from non-cancerous patches with 95.4 ± 0.3% accuracy (AUC 0.982) in sarcoma and 89.3 ± 0.5% accuracy (AUC 0.950) in mastocytoma. Lipomas showed distinct feature distributions, whereas mastocytoma and sarcoma features overlapped. Visual feature analysis additionally demonstrated that color-domain descriptors provided greater discrimination between histological tissue structures than conventional statistical and textural features. Importantly, stable classification performance was achieved when approximately 15–20% of the analyzed tissue was cancerous.,Conclusions: The framework combines interpretable spatial localization with feature-based tumor classification and reveals fraction-dependent performance. Independent patient-level validation is needed before clinical application.

Study advances current diagnostic efficiency of canine/feline (sub-)cutaneous tumors using machine learning and multimodal imaging data. White light (WL), fluorescence (FL) and ultrasound (US) imaging were combined into hybrid approaches to differentiate between malignant mastocytomas, soft tissue sarcomas and benign lipomas. Support Vector Machine and Ensemble classifiers were optimized via sequential feature selection. US radio-frequency signals were quantitatively analyzed to derive the colormaps of six US estimates, corresponding to spectral and temporal domains of the acoustic field. This resulted in the quantification of 72 morphological features for US; as well as 24 and 12 - for WL and FL data, respectively. Resulting classification efficiency for mastocytoma and sarcoma using US data was >75%; US+FL - 75-80%; US+WL - 85-90% and US+OPTICS - 90-95%. ∼100% classification efficiency was achieved for the differentiation between benign and malignant tumors even using single WL feature for Ensemble classifier. US features, resulting in inferior classification efficiency, were competitive to superior optical, as they were selected during optimization to be added to or replace optical counterparts. Additional tissue differentiation was performed on z-stacks of US colormaps, obtained using 3D arrays of US radio-frequency signals. This resulted in ∼70% differentiation efficiency for mastocytoma and sarcoma as well as >95% for benign and malignant tissues. The obtained additional metric of classification efficiency provides complementary diagnostic support, which for Support Vector Machine can be expressed as: 90.3 ± 1.9% (US+WL)×71.2 ± 0.6% (US_{Depth Profile}). This hybrid criterion adds robustness to diagnostic model and may be very beneficial to characterize heterogeneous tissues.

Canine/feline (sub-)cutaneous tumors, which include lipomas, mastocytomas and soft tissue sarcomas, introduce diagnostic challenges due to inherent tissue heterogeneity, accompanied by diverse clinical pathogenesis. Current study integrates conventional imaging techniques optical (white light and autofluorescence) as well as high frequency ultrasound imaging to train machine learning classifiers: linear discriminant analysis, support vector machine and random forest. Study resulted in ~ 100% classification efficiency between benign lipoma and combined mastocytoma and sarcoma tissues for all the classifiers. For the differentiation between mastocytoma and sarcoma tumors, both support vector machine and random forest outperformed conventional linear discriminant analysis classifier. Support vector machine displayed the highest classification efficiency for bimodal groups: (i) ultrasound + fluorescence and (ii) ultrasound + white light as well as (iii) fluorescence + white light. However, it failed for trimodal ultrasound + optics combination, indicating possible upper limit for imaging mode addition. The multimodal effect was obtained using both statistically significant set of features as well as optimal set of features, determined using sequential feature addition. Resulting classification efficiency for combined ultrasound + fluorescence approach was > 85% and even higher for ultrasound + white light or ultrasound + optics multimodal approaches reaching ~ 95%. In the classification of mastocytoma and sarcoma, support vector machine classifier was able to detect significant (p < 0.05) multimodal effect for bimodal groups of: (i) fluorescence + white light, (ii) ultrasound + fluorescence and (iii) ultrasound + white light. On the contrary, random forest demonstrated relevant increment only for the combination of fluorescence and white light. Inferior features of ultrasound or fluorescence have been evaluated to be competitive with the features of highly-efficient white light as they were automatically selected during the process of feature optimization. In addition, another phenomenon of manifestation of multimodality has been observed: in multimodal groups, ultrasound features tended to substitute the features of white light, not just simply be added to them. Multimodal approach was determined to be highly-required for the classification of heterogeneous mastocytoma and sarcoma tumors, which display more similar morphological characteristics. However, when differentiating very distinct lipomas from mastocytomas or sarcomas, the multimodal approach was not a requisite.

Microbubble (MB)- and ultrasound (US)-facilitated intracellular Ca²⁺ delivery, known as sonoporation (SP), is a promising anticancer treatment modality, since it allows a spatio-temporally controllable and side-effect-free alternative to conventional chemotherapy. The current study provides extensive evidence that a 5 mM concentration of Ca²⁺ in combination with US alone or US and Sonovue MBs can be an alternative to the conventional 20 nM concentration of the anticancer drug bleomycin (BLM). Ca²⁺ application together with SP induces a similar level of death in Chinese hamster ovary cells to the combination of BLM and SP but does not cause systemic toxicity, as is inherent to conventional anticancer drugs. In addition, Ca²⁺ delivery via SP alters three vital characteristics essential for viable cells: membrane permeability, metabolic activity and proliferation ability. Most importantly, Ca²⁺ delivery via SP elicits sudden cell death—occurring within 15 min—which remains similar during 24–72 h and 6 d periods. The extensive study of US waves side-scattered by MBs led to the quantification of the cavitation dose (CD) separately for subharmonics, ultraharmonics, harmonics and broadband noise (up to 4 MHz). The CD was suitable for the prognostication of the cytotoxic efficiency of both anticancer agents, Ca²⁺ and BLM, as was indicated by an overall high (R² ≥ 0.8) correlation (22 pairs in total). These extensive analytical data imply that a broad range of frequencies are applicable for the feedback-loop control of the process of US-mediated Ca²⁺ or BLM delivery, successively leading to the eventual standardization of the protocols for the sonotransfer of anticancer agents as well as the establishment of a universal cavitation dosimetry model.

Gene electrotransfer (GET) is recognized as a promising technique for the development of an efficient tool for gene therapy. Such a therapy would have applications in the treatment of a variety of genetic diseases, including cancer. However, despite its wide applicability, the technique is limited by the lack of understanding of the fundamental mechanism of electroporation as well as other important factors that directly or indirectly influence its success rate. In the current study, we analyzed the impact of low concentrations (0–1 mM) of Ca²⁺ on the process of DNA electrotransfer using flow cytometry. The results revealed that the presence of a CaCl₂ concentration as low as 0.25 mM decreased the efficiency of GET by ~1.5-fold and cell viability decreased by ~2–3-fold. In addition, we determined that the observed phenomenon of the decrease in pDNA electrotransfer due to the influence of Ca²⁺ was not the consequence of cell death but rather should be attributed to secondary mechanisms. The data presented in this study provide an insight into the importance of Ca²⁺ in the process of gene electrotransfer that may be directly applicable to in vivo settings.

Intracellular calcium ion delivery via sonoporation has been validated to be a substitute for classical chemotherapy. However, the mechanism behind calcium sonoporation remains unclear to this day. To elucidate the role of calcium in the process of sonoporation, we aimed to investigate the influence of different calcium concentration on cell membrane permeabilization and cell viability after sonoporation. In this study, we present experimental evidence that extracellular calcium plays a major role in cell membrane molecular transport after applying ultrasound pulses. Ultrasound-microbubble cavitation in the presence of different calcium concentration affects fundamental cell bio-physio-chemical conditions: cell membrane integrity, metabolic activity, and colony formation. Corresponding vital characteristics were evaluated using three independent viability tests: propidium iodide assay (20 min–3 h), MTT assay (48 h), and cell clonogenic assay (6 d). The results indicate instant cell death, as the level of cell viability was determined to be similar within a 20 min–48 h–6 d period. Inertial cavitation activities have been determined to be directly involved in calcium delivery via sonoporation according to high correlation (R² > 0.85, p < 0.01) of inertial cavitation dose with change in either cell membrane permeabilization, metabolic activity, and colony formation efficiency. In general, calcium delivery via sonoporation induces rapid cell death, occurring within 20 min after treatment, that is the result of ultrasound mediated microbubble cavitation.