Bio


I am a physician from Chile, passionate about medical education and educational technologies. My research focuses on the use of artificial intelligence, virtual reality, and augmented reality in medical education. I am actively developing tools that leverage artificial intelligence to enhance and evaluate clinical reasoning among healthcare professionals.

Honors & Awards


  • Community Impact Award, Stanford Alumni Association (2024)
  • Stanford Graduate Fellowships in Science and Engineering, Stanford University (2023)
  • Enhancing Diversity in Graduate Education Fellowship, Stanford University (2022)
  • Patil Teaching Innovation Award, Association for Medical Education in Europe (2016)

Professional Affiliations and Activities


  • Member Technology Enhanced Learning Committee, AMEE (2024 - Present)
  • Member, ASME (2024 - Present)
  • Member, AMEE (2023 - Present)
  • Founder - Member, LatinSOTL (2019 - Present)

Education & Certifications


  • M.D., University of Chile, Medicine (2019)

Personal Interests


Medical education, Higher Education, Educational Technologies, Artificial Intelligence

Research Interests


  • Higher Education
  • Technology and Education

Lab Affiliations


All Publications


  • Reclaiming focus in medical education: Learning, not AI. Medical teacher Rojas, M. 2024: 1

    View details for DOI 10.1080/0142159X.2024.2394171

    View details for PubMedID 39167925

  • Exploring the Performance of ChatGPT Versions 3.5, 4, and 4 With Vision in the Chilean Medical Licensing Examination: Observational Study. JMIR medical education Rojas, M., Rojas, M., Burgess, V., Toro-Pérez, J., Salehi, S. 2024; 10: e55048

    Abstract

    The deployment of OpenAI's ChatGPT-3.5 and its subsequent versions, ChatGPT-4 and ChatGPT-4 With Vision (4V; also known as "GPT-4 Turbo With Vision"), has notably influenced the medical field. Having demonstrated remarkable performance in medical examinations globally, these models show potential for educational applications. However, their effectiveness in non-English contexts, particularly in Chile's medical licensing examinations-a critical step for medical practitioners in Chile-is less explored. This gap highlights the need to evaluate ChatGPT's adaptability to diverse linguistic and cultural contexts.This study aims to evaluate the performance of ChatGPT versions 3.5, 4, and 4V in the EUNACOM (Examen Único Nacional de Conocimientos de Medicina), a major medical examination in Chile.Three official practice drills (540 questions) from the University of Chile, mirroring the EUNACOM's structure and difficulty, were used to test ChatGPT versions 3.5, 4, and 4V. The 3 ChatGPT versions were provided 3 attempts for each drill. Responses to questions during each attempt were systematically categorized and analyzed to assess their accuracy rate.All versions of ChatGPT passed the EUNACOM drills. Specifically, versions 4 and 4V outperformed version 3.5, achieving average accuracy rates of 79.32% and 78.83%, respectively, compared to 57.53% for version 3.5 (P<.001). Version 4V, however, did not outperform version 4 (P=.73), despite the additional visual capabilities. We also evaluated ChatGPT's performance in different medical areas of the EUNACOM and found that versions 4 and 4V consistently outperformed version 3.5. Across the different medical areas, version 3.5 displayed the highest accuracy in psychiatry (69.84%), while versions 4 and 4V achieved the highest accuracy in surgery (90.00% and 86.11%, respectively). Versions 3.5 and 4 had the lowest performance in internal medicine (52.74% and 75.62%, respectively), while version 4V had the lowest performance in public health (74.07%).This study reveals ChatGPT's ability to pass the EUNACOM, with distinct proficiencies across versions 3.5, 4, and 4V. Notably, advancements in artificial intelligence (AI) have not significantly led to enhancements in performance on image-based questions. The variations in proficiency across medical fields suggest the need for more nuanced AI training. Additionally, the study underscores the importance of exploring innovative approaches to using AI to augment human cognition and enhance the learning process. Such advancements have the potential to significantly influence medical education, fostering not only knowledge acquisition but also the development of critical thinking and problem-solving skills among health care professionals.

    View details for DOI 10.2196/55048

    View details for PubMedID 38686550

  • Characterization of content associated with lesbian, gay, bisexual, transgender, intersex, and queer individuals in Chilean medical schools: a cross-sectional survey. BMC medical education Rojas, M., Cánepa González, J., Ortiz-López, N. 2024; 24 (1): 167

    Abstract

    Lesbian, gay, bisexual, transgender, intersex, queer, and other sexual and gender identities (LGBTIQ+) individuals face health inequities. Additionally, medical students report a lack of confidence in providing specific health care to LGBTIQ + individuals, and medical schools do not offer the breadth and depth of coverage needed to fully prepare and make them comfortable in caring for these individuals. This study aims to characterize the teaching of curricular content related to LGBTIQ + health issues in medical schools in Chile.This was a cross-sectional descriptive mixed-methods study based on a 15-question survey sent to school directors of the 24 medical schools in Chile, conducted between October 2020 and July 2021. The questions included in the study were mostly based on two pre-existing questionnaires covering content, assessment methods, and identification of barriers to teaching this content.The validated questionnaire was answered by 14 of 24 Chilean medical schools, with 11 schools (78.9%) declaring that they included some training in their curriculum. The predominant range of time allocated to LGBTIQ + training in medical programs was between 1 and 5 h. The most addressed topics were HIV (92.85%), sexual orientation (78.57%), and chronic disease risk in LGBTIQ + populations (78.57%). Most schools, accounting for 71.5%, considered the content they delivered to be "moderately insufficient" or "insufficient". Regarding the teaching methodologies, the most used were lectures (92.8%), clinical cases (42.9%), and clinical simulation (28.6%).Most surveyed medical schools reported curricular spaces dedicated to teaching health issues of LGBTIQ + individuals, primarily during the pre-internship training period. However, the time allocated is insufficient, and there is little approach to topics beyond the patient's sexual history or sexual orientation. Given the crucial role of medical schools, they must adopt both local and national strategies to enrich training focused on the care of LGBTIQ + patients.

    View details for DOI 10.1186/s12909-024-05150-6

    View details for PubMedID 38383416

    View details for PubMedCentralID 4102129

  • Introducing the Problem-Solving Template as a Tool for Equity: Addressing Incoming Preparation Disparities J. Chem. Educ. Schwartz Poehlmann, J. K., Nardo, J. E., Rojas, M., Salehi, S. 2024; 101 (3): 1332–1340
  • Well-being variations on students of health sciences related to their learning opportunities, resources, and daily activities in an online and on-crisis context: a survey-based study. BMC medical education Perez-Villalobos, C., Ventura-Ventura, J., Spormann-Romeri, C., Paredes-Villarroel, X., Rojas-Pino, M., Jara-Reyes, C., Lopez, M., Castillo-Rabanal, I., Schilling-Norman, M. J., Baquedano-Rodriguez, M., Parra-Ponce, P., Toirkens-Niklitschek, J., Briede-Westermeyer, J. C., Alvarado-Figueroa, D. 2023; 23 (1): 37

    Abstract

    BACKGROUND: Universities' training process intensely relies on face-to-face education. The COVID-19 pandemic interrupted it and forced them to reinvent their process online. But this crisis seems not to be the last we will face, and we take it as a lesson to prepare for future crises. These critical contexts are especially challenging because they imply changing teaching strategies, and students may not have the technology access or the living conditions to connect as they need. They also lived through a pandemic where the virus and the life changes added stress to their learning process and threatened their well-being. So, this study aims to analyze how well-being variations reported by Health sciences students relate to their learning opportunities, access conditions, and daily activities.METHOD: We surveyed 910 Health sciences students from six different Chilean universities at the end of the first semester of 2020, the first in pandemic conditions. Respondents answered online questionnaires about 1) Remote teaching activities, 2) Learning resources availability, 3) Daily life activities, and 4) Well-being changes. We performed descriptive analysis and Structural Equation Modelling.RESULTS: Live videoconference classes were the most frequent teaching activity; only a third of the students had quiet spaces to study online, and most had to housekeep daily. More than two third reported some well-being deterioration. The structural equation model showed a good fit.CONCLUSION: Results show an online learning scenario that tries to emulate traditional learning focusing on expositive strategies. Most students reported that their well-being deteriorated during the semester, but tutorials, workplace availability, and social support were protective factors.

    View details for DOI 10.1186/s12909-023-04011-y

    View details for PubMedID 36653767

  • Satisfaction with remote teaching during the first semester of the COVID-19 crisis: Psychometric properties of a scale for health students PLOS ONE Perez-Villalobos, C., Ventura-Ventura, J., Spormann-Romeri, C., Melipillan, R., Jara-Reyes, C., Paredes-Villarroel, X., Rojas-Pino, M., Baquedano-Rodriguez, M., Castillo-Rabanal, I., Parra-Ponce, P., Bastias-Vega, N., Alvarado-Figueroa, D., Matus-Betancourt, O. 2021; 16 (4): e0250739

    Abstract

    Due to the health crisis caused by the COVID-19 pandemic, 220 million college students in the world had to halt face-to-face teaching and migrate to what has been called Emergency Remote Teaching, using virtual media, but without adequate preparation. The way this has impacted the student body and its satisfaction with the training process is unknown and there are no instruments backed by specific validity and reliability studies for this teaching context. This is why this study aims to analyze the psychometric properties of the Remote Teaching Satisfaction Scale applied to Chilean health sciences students.Quantitative study by means of surveys. We surveyed 1,006 health careers undergraduates chosen by convenience sampling. They came from six Chilean universities, located over a distance of 3,020 kilometers and followed 7 different careers. Women comprised the 78.53%. They answered the Remote Teaching Satisfaction Scale online to evaluate their perception of the first Emergency Remote Teaching term in 2020.A descriptive analysis of the items showed a moderate to positive evaluation of the teaching. The Confirmatory Factorial Analysis showed an adequate adjustment of the theoretical four factors model to the data obtained (CFI = 0.959; TLI = 0.953; RMSEA = 0.040). Correlations among factors oscillated from r = 0.21 to r = 0.69. The measurement invariance analysis supported the Configural, Metric and a partial Scalar model. Differences were found in three of the four factors when comparing the first-year students with those of later years. Finally, the Cronbach's α and McDonald's ω coefficients were over 0.70.The results display initial psychometric evidence supporting the validity and reliability of the Remote Teaching Satisfaction Scale to assess academic satisfaction in Chilean health careers students. Likewise, it is seen that first-year students show higher satisfaction levels about the implemented teaching.

    View details for DOI 10.1371/journal.pone.0250739

    View details for Web of Science ID 000662174400072

    View details for PubMedID 33909704

    View details for PubMedCentralID PMC8081238

  • Music Training and Education Slow the Deterioration of Music Perception Produced by Presbycusis in the Elderly FRONTIERS IN AGING NEUROSCIENCE Moreno-Gomez, F. N., Veliz, G., Rojas, M., Martinez, C., Olmedo, R., Panussis, F., Dagnino-Subiabre, A., Delgado, C., Delano, P. H. 2017; 9: 149

    Abstract

    The perception of music depends on the normal function of the peripheral and central auditory system. Aged subjects without hearing loss have altered music perception, including pitch and temporal features. Presbycusis or age-related hearing loss is a frequent condition in elderly people, produced by neurodegenerative processes that affect the cochlear receptor cells and brain circuits involved in auditory perception. Clinically, presbycusis patients have bilateral high-frequency hearing loss and deteriorated speech intelligibility. Music impairments in presbycusis subjects can be attributed to the normal aging processes and to presbycusis neuropathological changes. However, whether presbycusis further impairs music perception remains controversial. Here, we developed a computerized version of the Montreal battery of evaluation of amusia (MBEA) and assessed music perception in 175 Chilean adults aged between 18 and 90 years without hearing complaints and in symptomatic presbycusis patients. We give normative data for MBEA performance in a Latin-American population, showing age and educational effects. In addition, we found that symptomatic presbycusis was the most relevant factor determining global MBEA accuracy in aged subjects. Moreover, we show that melodic impairments in presbycusis individuals were diminished by music training, while the performance in temporal tasks were affected by the educational level and music training. We conclude that music training and education are important factors as they can slow the deterioration of music perception produced by age-related hearing loss.

    View details for DOI 10.3389/fnagi.2017.00149

    View details for Web of Science ID 000401658300001

    View details for PubMedID 28579956

    View details for PubMedCentralID PMC5437118