Influence of Artificial Intelligence on Canadian Medical Students' Preference for Radiology Specialty: A National Survey Study

Publication date: Available online 11 November 2018

Source: Academic Radiology

Author(s): Bo Gong, James P. Nugent, William Guest, William Parker, Paul J. Chang, Faisal Khosa, Savvas Nicolaou

Rationale and Objectives

Artificial intelligence (AI) has the potential to transform the clinical practice of radiology. This study investigated Canadian medical students' perceptions of the impact of AI on radiology, and their influence on the students' preference for radiology specialty.

Materials and Methods

In March 2018, an anonymous online survey was distributed to students at all 17 Canadian medical schools.

Results

Among 322 respondents, 70 students considered radiology as the top specialty choice, and 133 as among the top three choices. Only a minority (29.3%) of respondents agreed AI would replace radiologists in foreseeable future, but a majority (67.7%) agreed AI would reduce the demand for radiologists. Even among first-choice respondents, 48.6% agreed AI caused anxiety when considering the radiology specialty. Furthermore, one-sixth of respondents who would otherwise rank radiology as the first choice would not consider radiology because of the anxiety about AI. Prior significant exposure to radiology and high confidence in understanding of AI were shown to decrease the anxiety level. Interested students valued the opinions of local radiologists, radiology conferences, and journals. Students were most interested in "expert opinions on AI" and "discussing AI in preclinical radiology lectures" to understand the impact of AI.

Conclusion

Anxiety related to "displacement" (not "replacement") of radiologists by AI discouraged many medical students from considering the radiology specialty. The radiology community should educate medical students about the potential impact of AI, to ensure radiology is perceived as a viable long-term career choice.

from Imaging via a.sfakia on Inoreader https://ift.tt/2QBguxP

PET/TC con 18F-FDG en cáncer de cérvix localmente avanzado

Publication date: Available online 11 November 2018

Source: Revista Española de Medicina Nuclear e Imagen Molecular

Author(s): A.P. Caresia-Aróztegui, R.C. Delgado-Bolton, S. Alvarez-Ruiz, M. del Puig Cózar-Santiago, J. Orcajo-Rincon, M. de Arcocha-Torres, M.J. García-Velloso, en nombre del Grupo de Trabajo de Oncología de la Sociedad Española de Medicina Nuclear e Imagen Molecular

Resumen

El cáncer de cérvix es el segundo cáncer ginecológico en frecuencia a nivel mundial. En tumores localmente avanzados la PET/TC con ¹⁸F-FDG tiene un papel relevante en la detección de enfermedad ganglionar y a distancia, factores en los que se basan el tratamiento y el pronóstico de estas pacientes.

El objetivo de este trabajo es revisar las indicaciones actuales de la PET/TC con ¹⁸F-FDG en el cáncer de cérvix para cada una de las principales sociedades científicas (FIGO, NCCN, SEGO, SEOM, ESGO, ESMO) y la rentabilidad diagnóstica de la prueba comparada con las técnicas radiológicas convencionales, así como el procedimiento y su utilidad en la planificación de la radioterapia, en la valoración de respuesta y en la detección de recidiva.

Abstract

Cervical cancer is the second most common gynecological cancer worldwide. In locally advanced cervical cancer, ¹⁸F-FDG PET/CT has become important in the initial staging, particularly in the detection of nodal and distant metastasis, aspects with treatment implications and prognostic value.

The aims of this study were to review the role of ¹⁸F-FDG PET/CT in uterine cervical cancer, according to the guidelines of the main scientific institutions (FIGO, NCCN, SEGO, SEOM, ESGO, and ESMO) and its diagnostic accuracy compared to conventional radiological techniques, as well as to review the acquisition protocol and its utility in radiotherapy planning, response assessment and detection of recurrence.

from Imaging via a.sfakia on Inoreader https://ift.tt/2z78jlW

The role of experience for abstract concepts: Expertise modulates the electrophysiological correlates of mathematical word processing

Publication date: January 2019

Source: Brain and Language, Volume 188

Author(s): Laura Bechtold, Christian Bellebaum, Sophie Egan, Marco Tettamanti, Marta Ghio

Abstract

Embodied theories assign experience a crucial role in shaping conceptual representations. Supporting evidence comes mostly from studies on concrete concepts, where e.g., motor expertise facilitated action concept processing. This study examined experience-dependent effects on abstract concept processing. We asked participants with high and low mathematical expertise to perform a lexical decision task on mathematical and nonmathematical abstract words, while acquiring event-related potentials. Analyses revealed an interaction of expertise and word type on the amplitude of a fronto-central N400 and a centro-parietal late positive component (LPC). For mathematical words, we found a trend for a lower N400 and a significantly higher LPC amplitude in experts compared to nonexperts. No differences between groups were found for nonmathematical words. The results suggest that expertise affects the processing stages of semantic integration and memory retrieval specifically for expertise-related concepts. This study supports the generalization of experience-dependent conceptual processing mechanisms to the abstract domain.

from Speech via a.sfakia on Inoreader https://ift.tt/2OEPPOG

Spoken language coding neurons in the Visual Word Form Area: Evidence from a TMS adaptation paradigm

Publication date: Available online 12 November 2018

Source: NeuroImage

Author(s): Chotiga Pattamadilok, Samuel Planton, Mireille Bonnard

Abstract

While part of the left ventral occipito-temporal cortex (left-vOT), known as the Visual Word Form Area, plays a central role in reading, the area also responds to speech. This cross-modal activation has been explained by three competing hypotheses. Firstly, speech is converted to orthographic representations that activate, in a top-down manner, written language coding neurons in the left-vOT. Secondly, the area contains multimodal neurons that respond to both language modalities. Thirdly, the area comprises functionally segregated neuronal populations that selectively encode different language modalities. A transcranial magnetic stimulation (TMS)-adaptation protocol was used to disentangle these hypotheses. During adaptation, participants were exposed to spoken or written words in order to tune the initial state of left-vOT neurons to one of the language modalities. After adaptation, they performed lexical decisions on spoken and written targets with TMS applied over the left-vOT. TMS showed selective facilitatory effects. It accelerated lexical decisions only when the adaptors and the targets shared the same modality, i.e., when left-vOT neurons had initially been adapted to the modality of the target stimuli. Since this within-modal adaptation was observed for both input modalities and no evidence for cross-modal adaptation was found, our findings suggest that the left-vOT contains neurons that selectively encode written and spoken language rather than purely written language coding neurons or multimodal neurons encoding language regardless of modality.

from Imaging via a.sfakia on Inoreader https://ift.tt/2T3Ljg1

Age-related changes in brain deactivation but not in activation after motor learning

Publication date: Available online 12 November 2018

Source: NeuroImage

Author(s): K.M.M. Berghuis, S. Fagioli, N.M. Maurits, I. Zijdewind, J.B.C. Marsman, T. Hortobágyi, G. Koch, M. Bozzali

Abstract

It is poorly understood how healthy aging affects neural mechanisms underlying motor learning. We used blood-oxygen-level dependent (BOLD) contrasts to examine age-related changes in brain activation after acquisition and consolidation (24 h) of a visuomotor tracking skill. Additionally, structural magnetic resonance imaging and diffusion tensor imaging were used to examine age-related structural changes in the brain. Older adults had reduced gray matter volume (628 ± 57 ml) and mean white matter anisotropy (0.18 ± 0.03) compared with young adults (741 ± 59 ml and 0.22 ± 0.02, respectively). Although motor performance was 53% lower in older (n = 15, mean age 63.1 years) compared with young adults (n = 15, mean age 25.5 years), motor practice improved motor performance similarly in both age groups. While executing the task, older adults showed in general greater and more widespread activation compared with young adults. BOLD activation decreased in parietal and occipital areas after skill acquisition but activation increased in these areas after consolidation in both age groups, indicating more efficient visuospatial processing immediately after skill acquisition. Changes in deactivation in specific areas were age-dependent after consolidating the motor skill into motor memory. Young adults showed greater deactivations from post-test to retention in parietal, occipital and temporal cortices whereas older adults showed smaller deactivation in the frontal cortex. Since learning rate was similar between age groups, age-related changes in activation patterns may be interpreted as a compensatory mechanism for age-related structural decline.

from Imaging via a.sfakia on Inoreader https://ift.tt/2zVw4NJ

Differentiating guilt and shame in an interpersonal context with univariate activation and multivariate pattern analyses

Publication date: Available online 12 November 2018

Source: NeuroImage

Author(s): Ruida Zhu, Chunliang Feng, Shen Zhang, Xiaoqin Mai, Chao Liu

Abstract

Guilt and shame are usually evoked during interpersonal interactions. However, no study has compared guilt and shame processing under such circumstances. In the present study, we investigated guilt and shame in an interpersonal context using functional magnetic resonance imaging (fMRI). Behaviorally, participants reported more "guilt" when their wrong advice caused a confederate's economic loss, whereas they reported more "shame" when their wrong advice were correctly refused by the confederate. The fMRI results showed that both guilt and shame activated regions related to the integration of theory of mind and self-referential information (dorsal medial prefrontal cortex, dmPFC) and to the emotional processing (anterior insula). Guilt relative to shame activated regions linked with theory of mind (supramarginal gyrus and temporo-parietal junction) and cognitive control (orbitofrontal cortex/ventrolateral prefrontal cortex and dorsolateral prefrontal cortex). Shame relative to guilt revealed no significant results. Using multivariate pattern analysis, we demonstrated that in addition to the regions found in the univariate activation analysis, the ventral anterior cingulate cortex and dmPFC could also distinguish guilt and shame. These results do not only echo previous studies of guilt and shame using recall and imagination paradigms but also provide new insights into the psychological and neural mechanisms of guilt and shame.

from Imaging via a.sfakia on Inoreader https://ift.tt/2T2wdrc

Diagnosis of osteoporotic vertebral fractures in children

Abstract

Osteoporosis is a generalised disorder of the skeleton with reduced bone density and abnormal bone architecture. It increases bone fragility and renders the individual susceptible to fractures. Fractures of the vertebrae are common osteoporotic fractures. Vertebral fractures may result in scoliosis or kyphosis and, because they may be clinically silent, it is imperative that vertebral fractures are diagnosed in children accurately and at an early stage, so the necessary medical care can be implemented. Traditionally, diagnosis of osteoporotic vertebral fractures has been from lateral spine radiographs; however, a small number of studies have shown that dual energy x-ray absorptiometry is comparable to radiographs for identifying vertebral fractures in children, while allowing reduced radiation exposure. The diagnosis of vertebral fractures from dual energy x-ray absorptiometry is termed vertebral fracture assessment. Existing scoring systems for vertebral fracture assessment in adults have been assessed for use in children, but there is no standardisation and observer reliability is variable. This literature review suggests the need for a semiautomated tool that (compared to the subjective and semiquantitative methods available) will allow more reliable and precise detection of vertebral fractures in children.

from Imaging via a.sfakia on Inoreader https://ift.tt/2Pl2BXX