Privacy-Preserving Tampering Detection in Automotive Systems

Modern auto-vehicles are built upon a vast collection of sensors that provide large amounts of data processed by dozens of Electronic Control Units (ECUs). These, in turn, monitor and control advanced technological systems providing a large palette of features to the vehicle’s end-users (e.g., automated parking, autonomous vehicles). As modern cars become more and more interconnected with external systems (e.g., cloud-based services), enforcing privacy on data originating from vehicle sensors is becoming a challenging research topic. In contrast, deliberate manipulations of vehicle components, known as tampering, require careful (and remote) monitoring of the vehicle via data transmissions and processing. In this context, this paper documents an efficient methodology for data privacy protection, which can be integrated into modern vehicles. The approach leverages the Fast Fourier Transform (FFT) as a core data transformation algorithm, accompanied by filters and additional transformations. The methodology is seconded by a Random Forest-based regression technique enriched with further statistical analysis for tampering detection in the case of anonymized data. Experimental results, conducted on a data set collected from the On-Board Diagnostics (OBD II) port of a 2015 EUR6 Skoda Rapid 1.2 L TSI passenger vehicle, demonstrate that the restored time-domain data preserves the characteristics required by additional processing algorithms (e.g., tampering detection), showing at the same time an adjustable level of privacy. Moreover, tampering detection is shown to be 100% effective in certain scenarios, even in the context of anonymized data.

The Role of Information Theory in the Field of Big Data Privacy

Protecting privacy in Big Data is a rapidly growing research area. The first approach towards privacy assurance was the anonymity method. However, recent research indicated that simply anonymized data sets can be easily attacked. Later, differential privacy was proposed, which proved to be the most promising approach. The trade-off between privacy and the usefulness of published data, as well as other problems, such as the availability of metrics to compare different ways of achieving anonymity, are in the realm of Information Theory. Although a number of review articles are available in literature, the information - theoretic methods capacities haven’t been paid due attention. In the current article an overview of state-of-the-art methods from Information Theory to ensure privacy are provided.

Acute and Long-Term Outcomes of ST-Elevation Myocardial Infarction in Cancer Patients, a ‘Real World’ Analysis with 175,000 Patients

Background: Acute myocardial infarction (AMI) and cancer are common and serious diseases. As the prognosis and treatment of both diseases has improved, more cancer patients will suffer an AMI. Unfortunately, data on these “double hit” patients is scarce. Methods: From the largest public German health insurance, anonymized data of all patients with pre-existing cancer who were hospitalized due to ST-elevation MI (STEMI) between 2010 and 2017 were analyzed and followed-up until 2018. Results: Of 175,262 STEMI patients, 27,213 had pre-existing cancer (15.5%). Most frequent were skin (24.9%), prostate (17.0%), colon (11.0%), breast (10.9%), urinary tract (10.6%), and lung cancer (5.2%). STEMI patients with malignancies were older and presented more often with coronary three-vessel disease, atrial arrhythmias, chronic kidney disease, chronic heart failure, cerebrovascular and peripheral artery disease (PAD, each p < 0.001). They showed more often previous AMI, percutaneous coronary interventions (PCI), cardiac surgery, and stroke (all p < 0.001). Acute PCIs were applied between 2 and 6% less frequently compared to those without cancer. In-hospital adverse events occurred more frequently in cancer. Eight-year survival was 57.3% (95% CI 57.0–57.7%) without cancer and ranged between 41.2% and 19.2% in distinct cancer types. Multivariable Cox regression for all-cause mortality found, e.g., lung cancer (HR 2.04), PAD stage 4–6 (HR 1.78), metastasis (HR 1.72), and previous stroke (HR 1.44) to have the strongest impact (all p < 0.001). Conclusion: In this large “real world” data, prognosis after STEMI in cancer patients was markedly reduced but differed widely between cancer types. Of note, no withholding of interventional treatments in cancer patients could be observed.

Learning Realistic Patterns from Visually Unrealistic Stimuli: Generalization and Data Anonymization

Good training data is a prerequisite to develop useful Machine Learning applications. However, in many domains existing data sets cannot be shared due to privacy regulations (e.g., from medical studies). This work investigates a simple yet unconventional approach for anonymized data synthesis to enable third parties to benefit from such anonymized data. We explore the feasibility of learning implicitly from visually unrealistic, task-relevant stimuli, which are synthesized by exciting the neurons of a trained deep neural network. As such, neuronal excitation can be used to generate synthetic stimuli. The stimuli data is used to train new classification models. Furthermore, we extend this framework to inhibit representations that are associated with specific individuals. We use sleep monitoring data from both an open and a large closed clinical study, and Electroencephalogram sleep stage classification data, to evaluate whether (1) end-users can create and successfully use customized classification models, and (2) the identity of participants in the study is protected. Extensive comparative empirical investigation shows that different algorithms trained on the stimuli are able to generalize successfully on the same task as the original model. Architectural and algorithmic similarity between new and original models play an important role in performance. For similar architectures, the performance is close to that of using the original data (e.g., Accuracy difference of 0.56%-3.82%, Kappa coefficient difference of 0.02-0.08). Further experiments show that the stimuli can provide state-ofthe-art resilience against adversarial association and membership inference attacks.

Earning like a Prof: Academic rank wage premia in Ústí nad Labem, Czechia

The article attempts to estimate the size of the wage premia fetched by ranked academics on the academic market in Ústí nad Labem, Czechia. We employed a large (anonymized) data set of contracts and wages of employees of a medium-sized regional public university in Ústí nad Labem. We used OLS regression in various specifications to determine the wage premia of all educational levels (mainly full professors over associate professors/ docents and associate professors/docents over assistant professors/PhDs) while controlling for many attributes (of employees or contracts) possibly affecting wage levels. The local context regarding the topic of this article is discussed as well. The results generally confirm the intuition and show a clear pattern of increasing wages with levels and ranks. Focusing predominantly on the academic ranks, the monthly premium of associate professors (docents) over PhDs seems to be somewhere between 5 and 6 thousand CZK (185 and 220 EUR), and the premium of full professors over associate professors to an average of around 4 thousand CZK (150 EUR). The latter premium, however, exhibits systematic variation across different schools within the university: in some it is insignificant (around 0), while in others it is rather large and averages around 8 thousand CZK (300 EUR).

Safety and preliminary efficacy of the Gam-COVID-Vac vaccine and outcomes of SARS-CoV-2 infection in Russian patients with genitourinary malignancies

Background To our knowledge, there is no clinical data pertaining to COVID-19 outcomes and safety of COVID-19 vaccination in Russian patients with genitourinary (GU) malignancies. Aim of our analysis was to describe the characteristics of the COVID-19 infection course as well as preliminary safety and efficacy of Gam-COVID-Vac vaccine in patients with active GU malignancies. Methods Patients were retrospectively identified at nine cancer centers in different regions. Patients were included if COVID-19 was diagnosed by a polymerase chain reaction. Data from additional patients with GU cancers who had no positive SARS-CoV-2 RT-PCR test before vaccination and who received two doses of Gam-COVID-Vac (Sputnik V) between 11 February and 31 August 2021 were collected for safety assessment. Anonymized data were collected through an online registry covering demographics, treatments, and outcomes. Results The Gam-COVID-Vac vaccine was well tolerated; no grade 3–5 toxicities were reported in 112 vaccinated metastatic GU cancer patients. The most common grade 1 adverse events (81%) were injection site reactions (76%), flu-like illness (68%), and asthenia (49%). Five patients experienced grade 2 chills (4.5%) and 3 patients had grade 2 fever (2.7%). With median follow-up of 6.2 months, two COVID-19 cases were confirmed by RT-PCR test in the vaccine group (of 112 participants; 1.8%). Eighty-eight patients with COVID-19 disease were included in the analysis. The average age as of the study enrollment was 66 (range 39–81) and the majority of patients were male with renal cell carcinoma (RCC). Thirty-six patients (41%) had evidence of metastatic disease, of these 22 patients were receiving systemic therapy. More than half of patients required hospitalization. Fifty-four patients (61%) experienced complications. Sixteen patients who developed COVID-19 pneumonia required mechanical ventilator support. Sixteen patients (18%) died in a median of 23.5 days after the date of COVID-19 diagnosis was established. The 3-month survival rate was 82%. Clinical and/or radiographic progression of cancer during COVID-19 infection or the subsequent 3 months was observed in 10 patients (11.4%). Conclusion Patients with GU malignancies are at increased risk of mortality from COVID-19 infection when compared to the general population. Vaccination could be safe in GU cancer patients. Trial registration: retrospectively registered.

Survival of patients with metastatic bladder cancer in the Russian Federation: results of a multicenter registry study URRU

Background. Data on the overall survival (OS) of patients with metastatic bladder cancer (BCa) is rarely published.The objective of the URRU register study is to assess OS and collect information on the administration of different treatments in patients with metastatic BCa in routine clinical practice in Russia.Materials and methods. Patients were retrospectively identified in 9 oncology centers in different regions of Russia and included in the study if they were diagnosed with metastatic BCa between January 2017 and January 2018. We collected anonymized data online, including demographic characteristics of patients, details of their therapy, and outcomes.Results. This study included 246 patients. Their mean age upon the diagnosis of metastatic BCa was 72 years with 60.6 % of patients over 70 years of age. The proportion of males was 74.8 %. The histological subtype of BCa (urothelial carcinoma, etc.) was identified in 70.3 % of cases. Ninety-two patients (37.4 %) received pharmacotherapy. The most common treatment option was chemotherapy (76 %); the most common drug combination was gemcitabine and cisplatin (41.3 %). Immunotherapy was used in 19.6 % of patients; 13.6 % of participants received more than two lines of therapy. Three-year OS rate was 10.6 %; median OS was 7 months (95 % confidence interval (CI) 5.4-8.6). Patients receiving systemic therapy demonstrated significantly longer survival than those receiving no therapy (21 months; 95 % CI 17.38-24.62 vs 3 months; 95 % CI 1.79-4.22; p <0.0001). Patients receiving immunotherapy had better survival than individuals receiving chemotherapy (median OS 34.5 months vs 18 months; p = 0.003).Conclusion. The survival rates in the URRU study were relatively low, which can be attributed to the fact that only one-third of patients received pharmacotherapy and very few patients received immunotherapy. Second and subsequent lines of therapy were rarely used in patients with progressive disease. The implementation of novel treatments, including immune checkpoint inhibitors, will increase the survival of BCa patients.

Inferring the effect of interventions on COVID-19 transmission networks

Countries around the world implement nonpharmaceutical interventions (NPIs) to mitigate the spread of COVID-19. Design of efficient NPIs requires identification of the structure of the disease transmission network. We here identify the key parameters of the COVID-19 transmission network for time periods before, during, and after the application of strict NPIs for the first wave of COVID-19 infections in Germany combining Bayesian parameter inference with an agent-based epidemiological model. We assume a Watts–Strogatz small-world network which allows to distinguish contacts within clustered cliques and unclustered, random contacts in the population, which have been shown to be crucial in sustaining the epidemic. In contrast to other works, which use coarse-grained network structures from anonymized data, like cell phone data, we consider the contacts of individual agents explicitly. We show that NPIs drastically reduced random contacts in the transmission network, increased network clustering, and resulted in a previously unappreciated transition from an exponential to a constant regime of new cases. In this regime, the disease spreads like a wave with a finite wave speed that depends on the number of contacts in a nonlinear fashion, which we can predict by mean field theory.

Ethics appraisal procedure in 79,670 Marie Skłodowska-Curie proposals from the entire European HORIZON 2020 research and innovation program (2014–2020): A retrospective analysis

Introduction Horizon 2020 was the most significant EU Research and Innovation programme ever implemented and included the Marie Skłodowska-Curie Actions (MSCA). Proposals submitted to the MSCA actions awere subject to the Ethics Appraisal Procedure. In this work we explored the ethics appraisal procedure in MSCA H2020. Methods Using a retrospective analysis of pooled anonymized data, we explored the ethics appraisal procedure on proposals submitted to Marie Skłodowska-Curie Actions (MSCA) during the entire Horizon 2020 program period (N = 79,670). Results Our results showed that one of the most frequently identified ethics categories was Data protection. We also detected slight differences between applicants’ and the ethics reviewers’ awareness of ethical issues. Trajectory analysis of all ethics screened proposals appearing on main lists showed that a minimal portion of all screened submissions required additional ethics checks in the project implementation phase. Conclusion Personal data protection is one of the most represented ethics categories indicated among MSCA actions which exhaust ethics assessment efforts and may lead to “overkills” in ethics requirements. Excluding the majority of personal data protection assessment from the ethics assessment, except for parts which are directly related to ethics like “Informed consent procedures”, might be necessary in the future. A gap in understanding of ethics issues between applicants and reviewers’ points to the necessity to further educate researchers on research ethics issues.

L-Diversity for Data Analysis: Data Swapping with Customized Clustering

Data anonymization should support the analysts who intend to use the anonymized data. Releasing datasets that contain personal information requires anonymization that balances privacy concerns while preserving the utility of the data. This work shows how choosing anonymization techniques with the data analyst requirements in mind improves effectiveness quantitatively, by minimizing the discrepancy between querying the original data versus the anonymized result, and qualitatively, by simplifying the workflow for querying the data.