A Survey on Uncertainty Estimation in Deep Learning Classification Systems from a Bayesian Perspective

Decision-making based on machine learning systems, especially when this decision-making can affect human lives, is a subject of maximum interest in the Machine Learning community. It is, therefore, necessary to equip these systems with a means of estimating uncertainty in the predictions they emit in order to help practitioners make more informed decisions. In the present work, we introduce the topic of uncertainty estimation, and we analyze the peculiarities of such estimation when applied to classification systems. We analyze different methods that have been designed to provide classification systems based on deep learning with mechanisms for measuring the uncertainty of their predictions. We will take a look at how this uncertainty can be modeled and measured using different approaches, as well as practical considerations of different applications of uncertainty. Moreover, we review some of the properties that should be borne in mind when developing such metrics. All in all, the present survey aims at providing a pragmatic overview of the estimation of uncertainty in classification systems that can be very useful for both academic research and deep learning practitioners.

Clinical relevance of proteomic profiling in de novo pediatric acute myeloid leukemia: a Children’s Oncology Group study

Pediatric acute myeloid leukemia (AML) remains a fatal disease for at least 30% of patients, stressing the need for improved therapies and better risk stratification. As proteins are the unifying feature of (epi)genetic and environmental alterations, and are often targeted by novel chemotherapeutic agents, we studied the proteomic landscape of pediatric AML. Protein expression and activation levels were measured in 500 bulk leukemic patient samples and 30 control CD34+ samples, using the reverse phase protein arrays with 296 strictly validated antibodies. The multi-step “MetaGalaxy” analysis methodology was applied and identified nine protein expression signatures (PrSIG), based on strong recurrent protein expression patterns. PrSIGs were associated with cytogenetics and mutational state, and with both favorable or unfavorable prognosis. Analysis based on treatment (i.e., ADE vs. ADE plus bortezomib (ADEB)) identified three PrSIGs that did better with ADEB vs. ADE. When PrSIGs were studied in the context of genetic subgroups, PrSIGs were independently prognostic after multivariate analysis, suggesting a potential value for proteomics in combination with current classification systems. Proteins with universally increased (n=7) or decreased (n=17) expression were observed across PrSIGs. Expression of certain proteins significantly differentially expressed from normal could be identified, forming a hypothetical platform for personalized medicine.

External fixation in distal radius fractures

Background: Distal radius fractures are frequent in emergency departments, treatment is generally conservative, but there are patients who meet certain criteria of instability who need surgical treatment, including external fixation. Aim: To update on the most important elements regarding the use of external fixation in unstable fractures of the distal radius. Methods: The search and analysis of the information was carried out in a period of 30 days (from April 1 to April 30, 2021) and the following words were used: distal radius fractures, distal radius fractures and external fixation and unstable radius fractures from the information obtained, a bibliographic review of a total of 809 articles published in the PubMed databases, Hinari, SciELO and Medline was carried out using the search manager and EndNote reference manager, of which 44 selected citations were used to perform the review, 37 from the last five years. Development: The criteria to be taken into account for the instability described by various authors are indicated, as well as the imaging parameters. The two most used classification systems are mentioned. Reference is made to general surgical indications, external fixation, and placement of additional wires. Complications are discussed and a comparison is made between external fixation and blocked volar plates.

Latent subtypes of manic and/or irritable episode symptoms in two population-based cohorts

Background Mood disorders are characterised by pronounced symptom heterogeneity, which presents a substantial challenge both to clinical practice and research. Identification of subgroups of individuals with homogeneous symptom profiles that cut across current diagnostic categories could provide insights in to the transdiagnostic relevance of individual symptoms, which current categorical diagnostic systems cannot impart. Aims To identify groups of people with homogeneous clinical characteristics, using symptoms of manic and/or irritable mood, and explore differences between groups in diagnoses, functional outcomes and genetic liability. Method We used latent class analysis on eight binary self-reported symptoms of manic and irritable mood in the UK Biobank and PROTECT studies, to investigate how individuals formed latent subgroups. We tested associations between the latent classes and diagnoses of psychiatric disorders, sociodemographic characteristics and polygenic risk scores. Results Five latent classes were derived in UK Biobank (N = 42 183) and were replicated in the independent PROTECT cohort (N = 4445), including ‘minimally affected’, ‘inactive restless’, active restless’, ‘focused creative’ and ‘extensively affected’ individuals. These classes differed in disorder risk, polygenic risk score and functional outcomes. One class that experienced disruptive episodes of mostly irritable mood largely comprised cases of depression/anxiety, and a class of individuals with increased confidence/creativity reported comparatively lower disruptiveness and functional impairment. Conclusions Findings suggest that data-driven investigations of psychopathological symptoms that include sub-diagnostic threshold conditions can complement research of clinical diagnoses. Improved classification systems of psychopathology could investigate a weighted approach to symptoms, toward a more dimensional classification of mood disorders.