Enhancing COVID-19 CT Image Segmentation: A Comparative Study of Attention and Recurrence in UNet Models

Imaging plays a key role in the clinical management of Coronavirus disease 2019 (COVID-19) as the imaging findings reflect the pathological process in the lungs. The visual analysis of High-Resolution Computed Tomography of the chest allows for the differentiation of parenchymal abnormalities of COVID-19, which are crucial to be detected and quantified in order to obtain an accurate disease stratification and prognosis. However, visual assessment and quantification represent a time-consuming task for radiologists. In this regard, tools for semi-automatic segmentation, such as those based on Convolutional Neural Networks, can facilitate the detection of pathological lesions by delineating their contour. In this work, we compared four state-of-the-art Convolutional Neural Networks based on the encoder-decoder paradigm for the binary segmentation of COVID-19 infections after training and testing them on 90 HRCT volumetric scans of patients diagnosed with COVID-19 collected from the database of the Pisa University Hospital. More precisely, we started from a basic model, the well-known UNet, then we added an attention mechanism to obtain an Attention-UNet, and finally we employed a recurrence paradigm to create a Recurrent-Residual UNet (R2-UNet). In the latter case, we also added attention gates to the decoding path of an R2-UNet, thus designing an R2-Attention UNet so as to make the feature representation and accumulation more effective. We compared them to gain understanding of both the cognitive mechanism that can lead a neural model to the best performance for this task and the good compromise between the amount of data, time, and computational resources required. We set up a five-fold cross-validation and assessed the strengths and limitations of these models by evaluating the performances in terms of Dice score, Precision, and Recall defined both on 2D images and on the entire 3D volume. From the results of the analysis, it can be concluded that Attention-UNet outperforms the other models by achieving the best performance of 81.93%, in terms of 2D Dice score, on the test set. Additionally, we conducted statistical analysis to assess the performance differences among the models. Our findings suggest that integrating the recurrence mechanism within the UNet architecture leads to a decline in the model's effectiveness for our particular application.

3D-Vision-Transformer Stacking Ensemble for Assessing Prostate Cancer Aggressiveness from T2w Images

Vision transformers represent the cutting-edge topic in computer vision and are usually employed on two-dimensional data following a transfer learning approach. In this work, we propose a trained-from-scratch stacking ensemble of 3D-vision transformers to assess prostate cancer aggressiveness from T2-weighted images to help radiologists diagnose this disease without performing a biopsy. We trained 18 3D-vision transformers on T2-weighted axial acquisitions and combined them into two- and three-model stacking ensembles. We defined two metrics for measuring model prediction confidence, and we trained all the ensemble combinations according to a five-fold cross-validation, evaluating their accuracy, confidence in predictions, and calibration. In addition, we optimized the 18 base ViTs and compared the best-performing base and ensemble models by re-training them on a 100-sample bootstrapped training set and evaluating each model on the hold-out test set. We compared the two distributions by calculating the median and the 95% confidence interval and performing a Wilcoxon signed-rank test. The best-performing 3D-vision-transformer stacking ensemble provided state-of-the-art results in terms of area under the receiving operating curve (0.89 [0.61-1]) and exceeded the area under the precision-recall curve of the base model of 22% (p < 0.001). However, it resulted to be less confident in classifying the positive class.

Raman spectroscopy and topological machine learning for cancer grading

In the last decade, Raman Spectroscopy is establishing itself as a highly promising technique for the classification of tumour tissues as it allows to obtain the biochemical maps of the tissues under investigation, making it possible to observe changes among different tissues in terms of biochemical constituents (proteins, lipid structures, DNA, vitamins, and so on). In this paper, we aim to show that techniques emerging from the cross-fertilization of persistent homology and machine learning can support the classification of Raman spectra extracted from cancerous tissues for tumour grading. In more detail, topological features of Raman spectra and machine learning classifiers are trained in combination as an automatic classification pipeline in order to select the best-performing pair. The case study is the grading of chondrosarcoma in four classes: cross and leave-one-patient-out validations have been used to assess the classification accuracy of the method. The binary classification achieves a validation accuracy of 81% and a test accuracy of 90%. Moreover, the test dataset has been collected at a different time and with different equipment. Such results are achieved by a support vector classifier trained with the Betti Curve representation of the topological features extracted from the Raman spectra, and are excellent compared with the existing literature. The added value of such results is that the model for the prediction of the chondrosarcoma grading could easily be implemented in clinical practice, possibly integrated into the acquisition system.

Correction: Ambient Assisted Living: Scoping Review of Artificial Intelligence Models, Domains, Technology, and Concerns

[This corrects the article DOI: 10.2196/36553.].

NAVIGATOR: an Italian regional imaging biobank to promote precision medicine for oncologic patients

NAVIGATOR is an Italian regional project boosting precision medicine in oncology with the aim of making it more predictive, preventive, and personalised by advancing translational research based on quantitative imaging and integrative omics analyses. The project’s goal is to develop an open imaging biobank for the collection and preservation of a large amount of standardised imaging multimodal datasets, including computed tomography, magnetic resonance imaging, and positron emission tomography data, together with the corresponding patient-related and omics-related relevant information extracted from regional healthcare services using an adapted privacy-preserving model. The project is based on an open-source imaging biobank and an open-science oriented virtual research environment (VRE). Available integrative omics and multi-imaging data of three use cases (prostate cancer, rectal cancer, and gastric cancer) will be collected. All data confined in NAVIGATOR (i.e., standard and novel imaging biomarkers, non-imaging data, health agency data) will be used to create a digital patient model, to support the reliable prediction of the disease phenotype and risk stratification. The VRE that relies on a well-established infrastructure, called D4Science.org, will further provide a multiset infrastructure for processing the integrative omics data, extracting specific radiomic signatures, and for identification and testing of novel imaging biomarkers through big data analytics and artificial intelligence.

Cardio-metabolic risk modeling and assessment through sensor-based measurements

OBJECTIVE

Cardio-metabolic risk assessment in the general population is of paramount importance to reduce diseases burdened by high morbility and mortality. The present paper defines a strategy for out-of-hospital cardio-metabolic risk assessment, based on data acquired from contact-less sensors.

METHODS

We employ Structural Equation Modeling to identify latent clinical variables of cardio-metabolic risk, related to anthropometric, glycolipidic and vascular function factors. Then, we define a set of sensor-based measurements that correlate with the clinical latent variables.

RESULTS

Our measurements identify subjects with one or more risk factors in a population of 68 healthy volunteers from the EU-funded SEMEOTICONS project with accuracy 82.4%, sensitivity 82.5%, and specificity 82.1%.

CONCLUSIONS

Our preliminary results strengthen the role of self-monitoring systems for cardio-metabolic risk prevention.

Machine and Deep Learning Prediction Of Prostate Cancer Aggressiveness Using Multiparametric MRI

Prostate cancer (PCa) is the most frequent male malignancy and the assessment of PCa aggressiveness, for which a biopsy is required, is fundamental for patient management. Currently, multiparametric (mp) MRI is strongly recommended before biopsy. Quantitative assessment of mpMRI might provide the radiologist with an objective and noninvasive tool for supporting the decision-making in clinical practice and decreasing intra- and inter-reader variability. In this view, high dimensional radiomics features and Machine Learning (ML) techniques, along with Deep Learning (DL) methods working on raw images directly, could assist the radiologist in the clinical workflow. The aim of this study was to develop and validate ML/DL frameworks on mpMRI data to characterize PCas according to their aggressiveness. We optimized several ML/DL frameworks on T2w, ADC and T2w+ADC data, using a patient-based nested validation scheme. The dataset was composed of 112 patients (132 peripheral lesions with Prostate Imaging Reporting and Data System (PI-RADS) score ≥ 3) acquired following both PI-RADS 2.0 and 2.1 guidelines. Firstly, ML/DL frameworks trained and validated on PI-RADS 2.0 data were tested on both PI-RADS 2.0 and 2.1 data. Then, we trained, validated and tested ML/DL frameworks on a multi PI-RADS dataset. We reported the performances in terms of Area Under the Receiver Operating curve (AUROC), specificity and sensitivity. The ML/DL frameworks trained on T2w data achieved the overall best performance. Notably, ML and DL frameworks trained and validated on PI-RADS 2.0 data obtained median AUROC values equal to 0.750 and 0.875, respectively, on unseen PI-RADS 2.0 test set. Similarly, ML/DL frameworks trained and validated on multi PI-RADS T2w data showed median AUROC values equal to 0.795 and 0.750, respectively, on unseen multi PI-RADS test set. Conversely, all the ML/DL frameworks trained and validated on PI-RADS 2.0 data, achieved AUROC values no better than the chance level when tested on PI-RADS 2.1 data. Both ML/DL techniques applied on mpMRI seem to be a valid aid in predicting PCa aggressiveness. In particular, ML/DL frameworks fed with T2w images data (objective, fast and non-invasive) show good performances and might support decision-making in patient diagnostic and therapeutic management, reducing intra- and inter-reader variability.

Discriminating Stress From Cognitive Load Using Contactless Thermal Imaging Devices

This study proposes long wave infrared technology as a contactless alternative to wearable devices for stress detection. To this aim, we studied the change in facial thermal distribution of 17 healthy subjects in response to different stressors (Stroop Test, Mental Arithmetic Test). During the experimental sessions the electrodermal activity (EDA) and the facial thermal response were simultaneously recorded from each subject. It is well known from the literature that EDA can be considered a reliable marker for the psychological state variation, therefore we used it as a reference signal to validate the thermal results. Statistical analysis was performed to evaluate significant differences in the thermal features between stress and non-stress conditions, as well as between stress and cognitive load. Our results are in line with the outcomes of previous studies and show significant differences in the temperature trends over time between stress and resting conditions. As a new result, we found that the mean temperature changes of some less studied facial regions, e.g., the right cheek, are able not only to significantly discriminate between resting and stressful conditions, but also allow to recognize the typology of stressors. This outcome not only directs future studies to consider the thermal patterns of less explored facial regions as possible correlates of mental states, but more importantly it suggests that different psychological states could potentially be discriminated in a contactless manner.

Can Magnetic Resonance Radiomics Analysis Discriminate Parotid Gland Tumors? A Pilot Study

Our purpose is to evaluate the performance of magnetic resonance (MR) radiomics analysis for differentiating between malignant and benign parotid neoplasms and, among the latter, between pleomorphic adenomas and Warthin tumors. We retrospectively evaluated 75 T2-weighted images of parotid gland lesions, of which 61 were benign tumors (32 pleomorphic adenomas, 23 Warthin tumors and 6 oncocytomas) and 14 were malignant tumors. A receiver operating characteristics (ROC) curve analysis was performed to find the threshold values for the most discriminative features and determine their sensitivity, specificity and area under the ROC curve (AUROC). The most discriminative features were used to train a support vector machine classifier. The best classification performance was obtained by comparing a pleomorphic adenoma with a Warthin tumor (yielding sensitivity, specificity and a diagnostic accuracy as high as 0.8695, 0.9062 and 0.8909, respectively) and a pleomorphic adenoma with malignant tumors (sensitivity, specificity and a diagnostic accuracy of 0.6666, 0.8709 and 0.8043, respectively). Radiomics analysis of parotid tumors on conventional T2-weighted MR images allows the discrimination of pleomorphic adenomas from Warthin tumors and malignant tumors with a high sensitivity, specificity and diagnostic accuracy.

An E-Nose for the Monitoring of Severe Liver Impairment: A Preliminary Study

Biologically inspired to mammalian olfactory system, electronic noses became popular during the last three decades. In literature, as well as in daily practice, a wide range of applications are reported. Nevertheless, the most pioneering one has been (and still is) the assessment of the human breath composition. In this study, we used a prototype of electronic nose, called Wize Sniffer (WS) and based it on an array of semiconductor gas sensor, to detect ammonia in the breath of patients suffering from severe liver impairment. In the setting of severely impaired liver, toxic substances, such as ammonia, accumulate in the systemic circulation and in the brain. This may result in Hepatic Encephalopathy (HE), a spectrum of neuro-psychiatric abnormalities which include changes in cognitive functions, consciousness, and behaviour. HE can be detected only by specific but time-consuming and burdensome examinations, such as blood ammonia levels assessment and neuro-psychological tests. In the presented proof-of-concept study, we aimed at investigating the possibility of discriminating the severity degree of liver impairment on the basis of the detected breath ammonia, in view of the detection of HE at its early stage.

Decision Making Concepts for the Remote, Personalized Evaluation of COPD Patients’ Health Status

Introduction: This article is part of the Focus Theme of Methods of Information in Medicine on “Biosignal Interpretation: Advanced Methods for Neural Signals and Images”.

Objectives: This paper presents the main concepts of a decision making approach for the remote management of COPD patients based on the early detection of disease exacerbation episodes.

Methods: An e-diary card is defined to evaluate a number of physiological variables and clinical parameters acquired remotely by means of wearable and environmental sensors deployed in patients’ long-stay settings. The automatic evaluation of the card results in a so-called Chronic Status Index (CSI) whose computation is tailored to patients’ specific manifestation of the disease (i.e., patient’s phenotype). The decision support method relies on a parameterized analysis of CSI variations so as to early detect worsening changes, identify exacerbation severity and track the patterns of recovery.

Results: A preliminary study, carried out in real settings with 30 COPD patients monitored at home, has shown the validity and sensitivity of the method proposed, which was effectively able to timely and correctly identify patients’ critical situation.

Conclusion: The preliminary results showed that the proposed e-diary card, which presents several novel features with respect to other solutions presented in the literature, can be practically used to remotely monitor COPD patients.

Self-monitoring systems for personalised health-care and lifestyle surveillance

The quality of life and individual well-being are universally recognised as key factors in disease prevention. In particular, lifestyle interventions are effective tools for reducing the risk and incidence of major illnesses, such as cardiovascular diseases and metabolic disorders. On the other hand, patient role is progressively shifting from being a passive recipient of care towards being a co-producer of her/his health. In this frame, novel devices and systems able to help individuals in self-evaluation are expected to play a crucial role. In this special issue we focus on innovative methodologies and technologies devoted to individual self-assessment, oriented both to healthy people to maintain their well-being, and to diseased persons to improve their care.

Face morphology: Can it tell us something about body weight and fat?

This paper proposes a method for an automatic extraction of geometric features, related to weight parameters, from 3D facial data acquired with low-cost depth scanners. The novelty of the method relies both on the processing of the 3D facial data and on the definition of the geometric features which are conceptually simple, robust against noise and pose estimation errors, computationally efficient, invariant with respect to rotation, translation, and scale changes. Experimental results show that these measurements are highly correlated with weight, BMI, and neck circumference, and well correlated with waist and hip circumference, which are markers of central obesity. Therefore the proposed method strongly supports the development of interactive, non obtrusive systems able to provide a support for the detection of weight-related problems.

A wize mirror for lifestyle improvement

This paper discusses the problem of fostering lifestyle changes towards healthier habits via tailored user guidance. We present a novel multisensory device, the Wize Mirror, which will be able to detect semeiotic face signs related to cardio-metabolic risk, and encourage users to reduce their risk by improving their lifestyle. Offering a proper user guidance requires solving three main issues: user profiling, definition of a wellness index based on biophysical data, and personalized guidance by means of coaching and supportive messages. For each of these issues, the solutions proposed in the EU FP7 Project SEMEOTICONS are presented, highlighting their advantages with respect to the state-of-the-art.

CHRONIOUS: an open, ubiquitous and adaptive chronic disease management platform for chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD) and renal insufficiency

CHRONIOUS is an highly innovative Information and Communication Technologies (ICT) research Initiative that aspires to implement its vision for ubiquitous health and lifestyle monitoring. The 17 European project partners are strictly working together since February 2008 to realize and open platform to manage and monitor elderly patients with chronic diseases and many difficulties to reach hospital centers for routine controls. The testing activities will be done in Italy and Spain involving COPD (Chronic Obstructive Pulmonary Disease) and CKD (Chronic Kidney Disease) patients, these being widespread and highly expensive in terms of social and economic costs. Patients, equipped by wearable technologies and sensors and interacting with lifestyle interfaces, will be assisted by healthcare personnel able to check the health record and critical conditions through the Chronious platform data analysis and decision support system. Additionally, the new ontology based literature search engine will help the clinicians in the standardization of care delivery process. This paper is to present the main project objectives and its principal components from the intelligent system point of view.

Isonymous structure in the white population of Córdoba, Argentina, in 1813

Surnames analysis is useful for populations in which only written documents remain, as is the case for historical populations. In Córdoba province, Argentina, census records contain nominal data of inhabitants, including information on sex, age, ethnosocial category, civil status, occupation, place of birth, and residence, that can be analyzed using surnames. Relationship indicators within and among ecclesiastic units in Córdoba were estimated by isonymy for the adult white population registered in the 1813 census. The Rii, Rij, and R(ST) coefficients and the surname abundance indicator (a) were calculated. Lasker's distances among categories of population units were used to cluster the 16 provincial population categories. Gradients for kinship within population and for surname diversity were in agreement with the principal areas and waves of original settlement in the province. The main population clusters reflect those areas, whereas minor clusters coincide with the network of roads existing in the territory by 1813. The structure of the white population in Córdoba province was determined by the geographic location of the original waves of settlement, and it followed a pattern of relationships conditioned by the routes connecting population units in the Colonial period.

A pharmacokinetic study of gemcitabine at fixed dose rate infusion in patients with impaired hepatic function

12009

Background: The aim of this study was to evaluate if hepatic dysfunction leads to increased toxicity of gemcitabine (gem) at fixed dose rate, and to characterize the pharmacokinetic (PK) of gem and its major metabolite (2`,2`-difluorodeoxyuridine- 2dFdU) in patients (pts) with normal and altered liver function. Methods: Eight pts with metastatic pancreatic or biliary tract cancer were treated with the followed schedule: gem 1000 mg/m2 at 10/mg/m2/min fixed rate days 1,8, and 15 every 28 days for a maximum of six cycles. Three pts had normal serum bilirubin level and AST level less than two times the upper limit of normal (ULN) (Cohort I); four pts had bilirubin level from 1.6 to 7.0 mg/dL and normal AST level, one pt had serum bilirubin level less than 1.6 mg/dL and AST level greater than two times the ULN (Cohort II). The PK parameters measured were: plasmatic peak concentration (Cmax), area under the plasma concentration-time curve (AUCexp), total plasma clearance (Cl p) and half life (t1/2). Results: Patient characteristics were: median age 62 yrs (range 28–75), male/female 4/4, median cycles cohort I: 6 cycles (3–6), median cycles cohort II: 3 cycles (1–5), median follow-up: 30 weeks (range 3–79) and median weeks of treatment: 14 (1–25). The rate of dose reduction was the same in the two cohorts, as the rate of omitted administrations. Patients with liver dysfunction tolerated gemcitabine without increased toxicity and neither AST nor bilirubin elevation was observed after drug administration. PK parameters were calculated at the first cycle and the results are presented below (see table ). Conclusions: The pharmacokinetics of gemcitabine at fixed dose rate in patients with impaired liver function seems similar to control; no difference between the two cohorts was observed in terms of toxicity and dose reduction.

[Table: see text]

No significant financial relationships to disclose.