Joana A. Loureiro

Discovering Mammography-based Machine Learning Classifiers for Breast Cancer Diagnosis

This work explores the design of mammography-based machine learning classifiers (MLC) and proposes a new method to build MLC for breast cancer diagnosis. We massively evaluated MLC configurations to classify features vectors extracted from segmented regions (pathological lesion or normal tissue) on craniocaudal (CC) and/or mediolateral oblique (MLO) mammography image views, providing BI-RADS diagnosis. Previously, appropriate combinations of image processing and normalization techniques were applied to reduce image artifacts and increase mammograms details. The method can be used under different data acquisition circumstances and exploits computer clusters to select well performing MLC configurations. We evaluated 286 cases extracted from the repository owned by HSJ-FMUP, where specialized radiologists segmented regions on CC and/or MLO images (biopsies provided the golden standard). Around 20,000 MLC configurations were evaluated, obtaining classifiers achieving an area under the ROC curve of 0.996 when combining features vectors extracted from CC and MLO views of the same case.

Targeting nanoparticles across the blood-brain barrier with monoclonal antibodies

Development of therapeutics for brain disorders is one of the more difficult challenges to be overcome by the scientific community due to the inability of most molecules to cross the blood-brain barrier (BBB). Antibody-conjugated nanoparticles are drug carriers that can be used to target encapsulated drugs to the brain endothelial cells and have proven to be very promising. They significantly improve the accumulation of the drug in pathological sites and decrease the undesirable side effect of drugs in healthy tissues. We review the systems that have demonstrated promising results in crossing the BBB through receptor-mediated endocytic mechanisms for the treatment of neurodegenerative disorders such as Alzheimers and Parkinsons disease.

Cellular uptake of PLGA nanoparticles targeted with anti-amyloid and anti-transferrin receptor antibodies for Alzheimer’s disease treatment

During the last few decades, relevant efforts have been reported to design nanocarriers for drug transport through the blood brain barrier (BBB). New drugs, such as peptide iAβ5, capable to inhibit the aggregates associated with Alzheimeŕs disease (AD) are being tested but the most frequent drawback is to reach the brain in the desired concentrations due to the low BBB permeability-surface area. Our approach, as a proof of concept to improve drug transport through the BBB, is based on poly(lactic-co-glycolic acid) (PLGA) nanoparticles with surface functionalized with anti-transferrin receptor monoclonal antibody (OX26) and anti-Aβ (DE2B4) to deliver encapsulated iAβ5 into the brain. Porcine brain capillary endothelial cells (PBCECs) were used as a BBB model to evaluate the system efficacy and toxicity. The uptake of immune nanoparticles with a controlled delivery of the peptide iAβ5 was substantially increased compared to the nanoparticles (NPs) without monoclonal antibody functionalization.

Resveratrol and Grape Extract-loaded Solid Lipid Nanoparticles for the Treatment of Alzheimer’s Disease

The aggregation of amyloid-β peptide (Aβ) has been linked to the formation of neuritic plaques, which are pathological hallmarks of Alzheimer’s disease (AD). Various natural compounds have been suggested as therapeutics for AD. Among these compounds, resveratrol has aroused great interest due to its neuroprotective characteristics. Here, we provide evidence that grape skin and grape seed extracts increase the inhibition effect on Aβ aggregation. However, after intravenous injection, resveratrol is rapidly metabolized into both glucuronic acid and sulfate conjugations of the phenolic groups in the liver and intestinal epithelial cells (within less than 2 h), which are then eliminated. In the present study, we show that solid lipid nanoparticles (SLNs) functionalized with an antibody, the anti-transferrin receptor monoclonal antibody (OX26 mAb), can work as a possible carrier to transport the extract to target the brain. Experiments on human brain-like endothelial cells show that the cellular uptake of the OX26 SLNs is substantially more efficient than that of normal SLNs and SLNs functionalized with an unspecific antibody. As a consequence, the transcytosis ability of these different SLNs is higher when functionalized with OX-26.

Peptide–surfactant interactions: Consequences for the amyloid-beta structure

The conformation of amyloid-beta peptide (Aβ) determines if toxic aggregates are formed. The peptide structure by its turn depends on the environment and molecule-molecule interactions. We characterized the secondary structure of Aβ-(1-40) in surfactant solutions and interacting with monolayers. The peptide adopts β-sheet structure in solutions of ionic surfactants at sub-micelle concentrations and α-helix in the presence of ionic micelles. Uncharged micelles induce β-sheets. Aβ-(1-40) alters the critical micelle concentration value of the non-ionic surfactant, underlining hydrophobic interactions. At ionic monolayers the peptide forms β-sheets when its concentration at the surface is high enough. These results suggest that only electrostatic interactions of charged micelles that surround completely the peptide are able to induce non-aggregated α-helix structure.

Region of interest demarcation for quantification of the apparent diffusion coefficient in breast lesions and its interobserver variability

PURPOSE We aimed to compare two different methods of region of interest (ROI) demarcation and determine interobserver variability on apparent diffusion coefficient (ADC) in breast lesions. METHODS Thirty-two patients with 39 lesions were evaluated with a 3.0 Tesla scanner using a diffusion-weighted sequence with several b-values. Two observers independently performed the ADC measurements using: 1) a small fixed area of 10 mm2 ROI within the area with highest restriction; 2) a large ROI so as to include the whole lesion. Differences were assessed using the Wilcoxon-rank test. Bland-Altman method and Spearman coefficient were applied for interobserver variability and correlation analysis. RESULTS ADC values measured using the two ROI demarcation methods were significantly different for both observers (P = 0.026; P = 0.033). There was no interobserver variability in ADC values using either method (large ROI, P = 0.21; small ROI, P = 0.64). ADC values of malignant lesions were significantly different between the two methods (P < 0.001). Variability in ADC was ≤0.008×10-3 mm2/s for both methods. When using the same method, ADC values were significantly correlated between the observers (small ROI: r=0.990, P < 0.001; large ROI: r=0.985, P < 0.001). CONCLUSION The choice of ROI demarcation method influences ADC measurements. Small ROIs show less overlap in ADC values and higher ADC reproducibility, suggesting that this method may improve lesion discrimination. Interobserver variability was low for both methods.

Dual ligand immunoliposomes for drug delivery to the brain

Drug delivery systems that can reach brain areas affected by amyloid deposits are still underdeveloped. We propose pegylated liposomes functionalized with two antibodies, the anti-transferrin receptor monoclonal antibody (OX26MAb) and the anti-amyloid beta peptide antibody (19B8MAb), as nanocarriers of drugs for Alzheimers disease therapy. Two distinct conjugation methods are investigated. In one formulation, the OX26MAb is conjugated to the tip of polyethylene glycol molecules through the maleimide group and the 19B8MAb is bound through the streptavidin-biotin complex. In the second system the conjugation reagents are swapped between the antibodies. Fluorescence spectroscopy experiments on porcine brain capillary endothelial cells show that the cellular uptake of the immunoliposomes is substantially more efficient if OX26MAb antibody is conjugated through the streptavidin-biotin complex instead of the maleimide group. The ability of the immunoliposomes to cross the blood brain barrier was established by in vivo studies in wild type rats. Our results demonstrate the importance of the conjugation method used to bind the antibody that targets the blood brain barrier to immunoliposomes for efficient drug delivery to the brain.

PLGA nanoparticles as a platform for vitamin D-based cancer therapy.

Summary Poly(lactic-co-glycolic acid) (PLGA) nanoparticles were studied as drug delivery vehicles for calcitriol, the active form of vitamin D3. In vitro effects of calcitriol encapsulated in PLGA nanoparticles were evaluated with respect to free calcitriol on human pancreatic cell lines, S2-013 and hTERT-HPNE, and the lung cancer cell line A549. Encapsulated calcitriol retained its biological activity, reducing the cell growth. Cytotoxicity assays demonstrated that encapsulation of calcitriol enhanced its inhibitory effect on cell growth at a concentration of 2.4 μM for the S2-013 cells (91%) and for A549 cells (70%) comparared to the free calcitriol results. At this concentration the inhibitory effect on nontumor cells (hTERT-HPNE) decreased to 65%. This study highlights the ability of PLGA nanoparticles to deliver vitamin D3 into cancer cells, with major effects regarding cancer cell cycle arrest and major changes in the cell morphological features.

Improving the performance of machine learning classifiers for Breast Cancer diagnosis based on feature selection

This paper proposed a comprehensive algorithm for building machine learning classifiers for Breast Cancer diagnosis based on the suitable combination of feature selection methods that provide high performance over the Area Under receiver operating characteristic Curve (AUC). The new developed method allows both for exploring and ranking search spaces of image-based features, and selecting subsets of optimal features for feeding Machine Learning Classifiers (MLCs). The method was evaluated using six mammography-based datasets (containing calcifications and masses lesions) with different configurations extracted from two public Breast Cancer databases. According to the Wilcoxon Statistical Test, the proposed method demonstrated to provide competitive Breast Cancer classification schemes reducing the number of employed features for each experimental dataset.

Cooperativity between various types of polar solute–solvent interactions in aqueous media

Partition coefficients of seven low molecular weight compounds were measured in multiple aqueous two-phase systems (ATPSs) formed by pairs of different polymers. The ionic composition of each ATPS was varied to include 0.01M sodium phosphate buffer (NaPB), pH 7.4 and 0.1M Na2SO4, 0.15M NaCl, and 0.15M NaClO4 all in 0.01M NaPB, pH 7.4. The differences between the solvent features of the coexisting phases in all the ATPSs were estimated from partitioning of a homologous series of dinitrophenylated-amino acids and by the solvatochromic method. The solute-specific coefficients for the compounds examined were determined by the multiple linear regression analysis using the modified linear solvation energy relationship equation. It is established that the solute specific coefficients characterizing different types of the solute-water interactions (dipole-dipole, dipole-ion, and H-bonding) for a given solute change in the presence of different salt additives in the solute specific manner. It is also found that these characteristics are linearly interrelated. It is suggested that there is a cooperativity between various types of solute-water interactions governed by the solute structure.