We propose a new method that employs transfer learning techniques to effectively correct sampling selection errors introduced by sparse annotations during supervised learning for automated tumor segmentation. The practicality of current learning-based automated tissue classification approaches is severely impeded by their dependency on manually segmented training databases that need to be recreated for each scenario of application, site, or acquisition setup. The comprehensive annotation of reference datasets can be highly labor-intensive, complex, and error-prone. The proposed method derives high-quality classifiers for the different tissue classes from sparse and unambiguous annotations and employs domain adaptation techniques for effectively correcting sampling selection errors introduced by the sparse sampling. The new approach is validated on labeled, multi-modal MR images of 19 patients with malignant gliomas and by comparative analysis on the BraTS 2013 challenge data sets. Compared to training on fully labeled data, we reduced the time for labeling and training by a factor greater than 70 and 180 respectively without sacrificing accuracy. This dramatically eases the establishment and constant extension of large annotated databases in various scenarios and imaging setups and thus represents an important step towards practical applicability of learning-based approaches in tissue classification.

The video-assisted thoracoscopic surgery (VATS) approach has become a standard for the treatment of early-stage non-small-cell lung cancer (NSCLC). Recently published meta-analyses proved the benefit of VATS versus thoracotomy for overall survival (OS) and reduction of postoperative complications. The aim of this study was to compare early outcomes, long-term survival and rate of postoperative complications of the VATS approach versus thoracotomy. In this retrospective cohort study, we analysed 982 individuals who underwent surgical resection for Stage I–IIA NSCLC between 2007 and 2015. Thirty- and 90-day mortality rates, length of hospital stay, rate of complications and OS were assessed. Propensity score matching was performed to compare 2 groups of patients. Two hundred and twenty-five individuals from the thoracotomy group and 225 patients from the VATS group were matched regarding pTNM, sex, the Charlson comorbidity index, type of resection and histological diagnosis. In the propensity score-matched patient group, the VATS approach was associated with a significant benefit regarding OS (P = 0.042). Although no significant difference was observed (P = 0.14) in the 3-year survival rate of patients who had a thoracotomy versus VATS, the 5-year survival rate among patients with VATS increased significantly (61% vs 78%, P = 0.0081). The adjusted VATS-related hazard ratio for pTNM, sex and age was 0.63 (95% confidence interval 0.40–0.98). The VATS surgical approach also reduced both the rate of postoperative atelectasis (4% for VATS vs 10% for open thoracotomy; P = 0.0052) and the need for blood transfusions (4% vs 12% respectively, P = 0.0054) and significantly shortened the postoperative length of stay (mean 7.25 vs 9.34 days, P < 0.0001). No significant differences in the 30-day mortality (1% vs 1%, P = 0.66) and 90-day mortality (1% vs 1%, P = 0.48) rates were observed. Patients with early-stage NSCLC operated on with VATS had fewer complications, shorter postoperative length of stay and better OS compared to those who were operated on by thoracotomy.

Although there are several data analysis frameworks, both commercial and open source, supporting the detection of tumours on nuclear magnetic resonance (NMR) sequences, none of them gives satisfactory results in the case of low volume tumors. The majority of the frameworks require the detailed analysis of at least two sequences of the examined sample, or give sample specific thresholds distinguishing between the tumor and subtypes of healthy tissue. In this paper, we present a novel algorithm for the automated estimation of tumor specific cut-off values in the domain of the apparent diffusion coefficient (ADC). Once the cut-off characteristics for a particular type of tumor is estimated, their further usage on other independent samples does not require any calculations except for an easy thresholding. The proposed methodology is a combination of classical decomposition of ADC distribution into a Gaussian mixture model (GMM) with k-means clustering subsequently performed on the parameters of mixture model components, leading to the identification of ADC distributions for every tissue type. The maximum conditional probability criterion gives the final threshold estimate. The developed signal analysis pipeline was applied to the problem of Glioblastoma Multiforme grade IV brain tumor segmentation, with a dataset of 119 randomly chosen ADC maps and Leave-One-Out cross-validation procedure for population error estimate. Additionally, a comparison to standard GMM based tumor segmentation algorithms as well as to three other automated segmentation methods was performed and the obtained tumor regions were referenced to the segmentation done by a human expert. The results demonstrate the average MiMSeg similarity to the expert-curated decision measured by the Dice coefficient as equal to 89.2% (with 95% confidence interval 87.7 ÷ 90.6). The MiMSeg algorithm significantly outperforms other techniques in the case of small tumors (of volume less than 10%), obtaining similarity to the expert-curated decision at the level 86.7%, with 44.9% obtained by standard GMM, 79.0% by Self-Organising-Maps algorithm, 68.7% by Murakami’s algorithm, and 78.2% by Kang’s method.

Here, we retrospectively investigate the value of voxel-wisely plotted diffusion tensor-derived (DTI) axial, radial and mean diffusivity for the early detection of malignant transformation (MT) in WHO II glioma compared to contrast-enhanced images. Forty-seven patients underwent brain magnetic resonance imaging follow-up between 2006–2014 after gross-tumor resection of intra-axial WHO II glioma. Axial/Mean/Radial diffusivity maps (AD/MD/RD) were generated from DTI data. AD_min/MD_min/RD_min values were quantified within tumor regions-of-interest generated by two independent readers including tumor contrast-to-noise (CNR). Sensitivity/specificity and area-under-the-curve (AUC) were calculated using receiver-operating-characteristic analysis. Inter-reader agreement was assessed (Cohen’s kappa). Eighteen patients demonstrated malignant transformation (MT) confirmed in 8/18 by histopathology and in 10/18 through imaging follow-up. Twelve of 18 patients (66.6%) with MT showed diffusion restriction timely coincidental with contrast-enhancement (CE). In the remaining six patients (33.3%), the diffusion restriction preceded the CE. The mean gain in detection time using DTI was (0.8±0.5 years, p = 0.028). Compared to MD_min and RD_min, ROC-analysis showed best diagnostic value for AD_min (sensitivity/specificity 94.94%/89.7%, AUC 0.96; p<0.0001) to detect MT. CNR was highest for AD (1.83±0.14), compared to MD (1.31±0.19; p<0.003) and RD (0.90±0.23; p<0.0001). Cohen’s Kappa was 0.77 for AD_min, 0.71 for MD_min and 0.65 for RD_min (p<0.0001, respectively). MT is detectable at the same time point or earlier compared to T1w-CE by diffusion restriction in diffusion-tensor-derived maps. AD demonstrated highest sensitivity/specificity/tumor-contrast compared to radial or mean diffusivity (= apparent diffusion coefficient) to detect MT.

Nuclear Magnetic Resonance (NMR) spectroscopy is a popular medical diagnostic technique. NMR is also the favourite tool of chemists/biochemists to elucidate the molecular structure of small or big molecules; it is also a widely used tool in material science, in food science etc. In the case of medical diagnosis it allows for determining a metabolic composition of analysed tissue which may support the identification of tumour cells. Precession signal, that is a crucial part of MR phenomenon, contains distortions that must be filtered out before signal analysis. One of such distortions is phase error. Five popular algorithms: Automics, Shanon's entropy minimization, Ernst's method, Dispa and eDispa are presented and discussed. A novel adaptive tuning algorithm for Automics method was developed and numerically optimal solutions to automatic tuning of the other four algorithms were proposed. To validate the performance of the proposed techniques, two experiments were performed - the first one was done with the use of in silico generated data. For all presented methods, the fine tuning strategies significantly increased the correction accuracy. The highest improvement was observed for Automics algorithm, independently of noise level, with relative phase error dropping by average from 10.25% to 2.40% for low noise level and from 12.45% to 2.66% for high noise level. The second validation experiment, done with the use of phantom data, confirmed the in silico results. The obtained accuracy of the estimation of metabolite concentration was at 99.5%. The proposed strategies for optimizing the phase correction algorithms significantly improve the accuracy of Nuclear Magnetic Resonance spectroscopy signal analysis.